Not a member of Pastebin yet?
Sign Up,
it unlocks many cool features!
- Your Telegram History in "EOS Validation (EMLG)"
- .....28.05.2018 03:18:24, Sharif Bouktila - EOS Dublin [@sharifbouktila]: >>upgraded the group to a supergroup<<
- 28.05.2018 03:20:38, Sharif Bouktila - EOS Dublin [@sharifbouktila]: This purpose of this Group is to agree how we validate EOS Mainnet.
- Once we have agreed, we will look for BPs to issue a statement to state that they are in agreement.
- .....28.05.2018 03:23:07, Sharif Bouktila - EOS Dublin [@sharifbouktila]: >>changed group photo<<
- 28.05.2018 03:27:22, Sharif Bouktila - EOS Dublin [@sharifbouktila]: For now if you can please share this group with BPs who we know will be positive.
- and
- If you can post here what is the Minimum which is needed for an EOS Chain to be validated as Mainnet
- I will compile into a Draft Document and then host a call to try reach agreement.
- 28.05.2018 03:29:01, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Lets aim to have the Conference call tomorrow and Agreement Statement issued by all BPs on May 30th.
- 28.05.2018 03:29:17, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Does that sound achievable?
- 28.05.2018 03:29:21, Eric - sw/eden [@xebb82]: Sounds good to me.
- 28.05.2018 03:29:42, Sam - EOS Dublin [@samnoble]: 👍
- 28.05.2018 03:30:02, Igor Lins e Silva - EOS Rio [@igorls]: Agreed
- .....28.05.2018 03:30:51, jem [@jemxpat]: >>unsupported service message type: messageActionChatJoinedByLink<<
- 28.05.2018 03:31:30, jem [@jemxpat]: Yup
- .....28.05.2018 03:39:34, Sharif Bouktila - EOS Dublin [@sharifbouktila]: >>changed group name to "EOS Validation (TEMP Group)"<<
- .....28.05.2018 03:45:32, Marcin @eosgreen.io [@marcin_eosgreen]: >>unsupported service message type: messageActionChatJoinedByLink<<
- 28.05.2018 03:47:14, Marcin @eosgreen.io [@marcin_eosgreen]: Hi Guys!
- 28.05.2018 03:48:26, Marcin @eosgreen.io [@marcin_eosgreen]: what time are you planing to have this call?
- 28.05.2018 03:49:48, Eric - sw/eden [@xebb82]: What do we have to agree upon to do a successful launch?
- Validation
- * ERC20 distribution (already agreed)
- * Extra accounts and contracts? (eosio.disco for example)
- * Also <producer-name>.p accounts (?)
- * Can we agree that all of us will burn 0.1 EOS or something for those accounts?
- * How do we pay for ram?
- * Early created account will pay more for same amount of ram than last created account
- * Extra producer accounts (EOS.Dublin.p)? Other name standard?.
- * Is it ok to use multisig in future to create more accounts for newly registered producers?
- * How do we want to (s)elect eosio?
- * How do we want to (s)elect ABPs?
- *
- I'll update this message as we come up with more things.
- 28.05.2018 03:51:12, Sharif Bouktila - EOS Dublin [@sharifbouktila]: 21:00 UTC Y/N ?
- .....28.05.2018 03:53:06, Basit Raza - EOSVibes [@boltxyz]: >>unsupported service message type: messageActionChatJoinedByLink<<
- .....28.05.2018 03:54:30, Kyle - EOSVibes [@TheAwakenment]: >>unsupported service message type: messageActionChatJoinedByLink<<
- .....28.05.2018 03:56:38, Sukesh Tedla (EOSGreen) [@Sukeshtedla]: >>unsupported service message type: messageActionChatJoinedByLink<<
- 28.05.2018 03:57:54, Eric - sw/eden [@xebb82]: Evening time in US means night time in asia, right? Better to run 09:00 UTC - or am I missing something?
- 28.05.2018 03:59:31, Sharif Bouktila - EOS Dublin [@sharifbouktila]: [[Photo]]
- 28.05.2018 03:59:47, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Basically add 12 hours to this on 30th May.
- .....28.05.2018 04:00:45, Devin Lee KOREOS.IO [@indend007]: >>unsupported service message type: messageActionChatJoinedByLink<<
- .....28.05.2018 04:08:13, Tyee Noprom | EOSBlock.co | EOSpace.io | EOS Committer [@tyeenoprom]: >>unsupported service message type: messageActionChatJoinedByLink<<
- .....28.05.2018 04:10:36, Khosi Morafo eosAfrica [@khosimorafo]: >>unsupported service message type: messageActionChatJoinedByLink<<
- .....28.05.2018 04:14:39, An Byung Chan - EOSYS [@Anbyungchan]: >>unsupported service message type: messageActionChatJoinedByLink<<
- 28.05.2018 04:15:01, Khosi Morafo eosAfrica [@khosimorafo]: Hi Guys, does the conference call mean a freeze on the discussion? I hope not. Erics list seems rather large and trying to fit that into a call could be difficult.
- .....28.05.2018 04:17:41, Robin [email protected] [@robinwen]: >>unsupported service message type: messageActionChatJoinedByLink<<
- 28.05.2018 04:18:32, Eric - sw/eden [@xebb82]: I think we'll need to increase the list further. And I'm open for doing several calls with different people multiple times during the coming hours if anyone thinks they can benefit from that.
- .....28.05.2018 04:19:54, Ross (EOSphere) [@rossco99]: >>unsupported service message type: messageActionChatJoinedByLink<<
- 28.05.2018 04:20:23, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Do as many calls as is needed , but we have a hard deadline this weekend.
- 28.05.2018 04:22:56, Eric - sw/eden [@xebb82]: Can we make it into a super group so I can link to my message above?
- 28.05.2018 04:23:05, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Try to get to a point that we have a checklist and once all items are complete you will be happy to publicly state you are happy to call that chain EOS Mainnet
- Once we have that we should try agree how many others need to also agree.
- 28.05.2018 04:24:05, Sharif Bouktila - EOS Dublin [@sharifbouktila]: I’ve made you an admin.
- 28.05.2018 04:24:48, Khosi Morafo eosAfrica [@khosimorafo]: [[Webpage]]For those who may want to add the list, here is a doc: -
- https://docs.google.com/document/d/1HG40PwrN0fL62TfwbOrrCjUpoQqlVhGtQb7WJe7kulA/edit?usp=sharing
- Based on Eric's original list
- 28.05.2018 04:26:16, Eric - sw/eden [@xebb82]: I can't seem to make it super.
- .....28.05.2018 04:27:36, Ravi (Blockpro.one) [@ravincal]: >>unsupported service message type: messageActionChatJoinedByLink<<
- 28.05.2018 04:28:01, Eric - sw/eden [@xebb82]: nice, that's better
- .....28.05.2018 04:50:19, Vahid Toosi - sw/eden [@vahidtoosi1]: >>unsupported service message type: messageActionChatJoinedByLink<<
- .....28.05.2018 04:52:51, mao EOSREAL [@maoyifeng]: >>unsupported service message type: messageActionChatJoinedByLink<<
- .....28.05.2018 04:53:58, Lee - EOSTEA [@strahe]: >>unsupported service message type: messageActionChatJoinedByLink<<
- 28.05.2018 04:55:31, mao EOSREAL [@maoyifeng]: hi bro and sis. It is a urgent and necessary task to do. Glad you cover it!
- .....28.05.2018 04:56:27, Ian(Yahuang Wu) - MEET.ONE [@wuyahuang]: >>unsupported service message type: messageActionChatJoinedByLink<<
- .....28.05.2018 05:06:49, Dafeng Guo - EOS Asia [@dftle]: >>unsupported service message type: messageActionChatJoinedByLink<<
- .....28.05.2018 05:07:01, Dafeng Guo - EOS Asia [@dftle]: >>unsupported service message type: messageActionChatAddUsers<<
- 28.05.2018 05:07:22, aza eosar [@p17er0bl3do]: about pausing the chain / altering base contracts i believe we should avoid that, we can validate the state of the chain at some specific block, even though it will be a little harder
- .....28.05.2018 05:08:07, Tim Weston - GenerEOS (BP Sydney) [@timweston]: >>unsupported service message type: messageActionChatJoinedByLink<<
- 28.05.2018 05:08:12, Sharif Bouktila - EOS Dublin [@sharifbouktila]: How would we agree on what block ?
- 28.05.2018 05:08:27, Dafeng Guo - EOS Asia [@dftle]: Then I think that’s 3 am beijing time
- .....28.05.2018 05:09:09, Dafeng Guo - EOS Asia [@dftle]: >>invited Alexandre Bourget - EOS Canada [@abourget]<<
- 28.05.2018 05:10:02, aza eosar [@p17er0bl3do]: after all system contracts have been set and stake holders account have been created, basically once the boot finishes
- .....28.05.2018 05:10:06, Charles H - EOS 42 [@ankh2054]: >>unsupported service message type: messageActionChatJoinedByLink<<
- 28.05.2018 05:10:17, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Would you prefer earlier or later? Or is that workable ?
- 28.05.2018 05:11:02, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Can you add the Google Doc
- 28.05.2018 05:13:07, aza eosar [@p17er0bl3do]: will see where it fits
- .....28.05.2018 05:13:10, Bean@EOS Cannon [@Luckybean]: >>unsupported service message type: messageActionChatJoinedByLink<<
- 28.05.2018 05:16:14, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Maybe add it as an open question “can we agree a specific block upon which to validate?”
- 28.05.2018 05:18:26, Charles H - EOS 42 [@ankh2054]: 09:00 UTC?
- 28.05.2018 05:20:02, Sharif Bouktila - EOS Dublin [@sharifbouktila]: One question in the Doc is about the launch process , let’s keep this discussion agnostic to the approach and aim for a checklist which can be used regardless.
- Think , could Deloitte (or similar) pick up and use this to independently validate
- 28.05.2018 05:20:14, Sharif Bouktila - EOS Dublin [@sharifbouktila]: 21:00 UTC
- 28.05.2018 05:21:17, Charles H - EOS 42 [@ankh2054]: Okay cool thanks
- 28.05.2018 05:22:36, Eric - sw/eden [@xebb82]: +1
- I added about Alexs yaml file,because it's the easiest way to understand the launch process right now.
- .....28.05.2018 05:23:50, Lucas @EOS Beijing [@LucasEosBeijing]: >>unsupported service message type: messageActionChatJoinedByLink<<
- .....28.05.2018 05:30:06, Jorge Ferrando - EOSMetal [@jorfermo]: >>unsupported service message type: messageActionChatJoinedByLink<<
- .....28.05.2018 05:30:48, LOME eosfans [@lomect]: >>unsupported service message type: messageActionChatJoinedByLink<<
- .....28.05.2018 05:31:05, Sharif Bouktila - EOS Dublin [@sharifbouktila]: >>unsupported service message type: messageActionPinMessage<<
- 28.05.2018 05:32:53, Sharif Bouktila - EOS Dublin [@sharifbouktila]: I’ve pinned the Google Document
- Please read and add any questions / requirements you feel need to be included in our EOS Validation Test Plan
- 28.05.2018 05:33:16, aza eosar [@p17er0bl3do]: Added, would be the same block at which we would "pause" it
- .....28.05.2018 05:39:12, Zhao Yu@ EOSLaoMao.com [@zhaoyu]: >>unsupported service message type: messageActionChatJoinedByLink<<
- 28.05.2018 05:42:12, Dafeng Guo - EOS Asia [@dftle]: Considering the large number of BPs in Asia, I propose 13:00 UTC.
- .....28.05.2018 05:43:10, Larry Ma @EOS CANNON [@larrymanysh]: >>unsupported service message type: messageActionChatJoinedByLink<<
- 28.05.2018 05:44:05, mao EOSREAL [@maoyifeng]: Please just let us know whatever time it would be. I will stay for the rest of the month...
- 28.05.2018 05:45:40, Eric - sw/eden [@xebb82]: Stated before, stating again;
- Several meeting during all of today could allow us to come to consensus quicker.
- 28.05.2018 05:46:25, Dafeng Guo - EOS Asia [@dftle]: Agree
- 28.05.2018 05:46:51, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Agreed - but let’s try make sure we have good participation and a circulation of minutes from each call.
- 28.05.2018 05:47:55, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Maybe have a minimum of 5 teams on each call?
- 28.05.2018 05:49:09, Sharif Bouktila - EOS Dublin [@sharifbouktila]: And on each one agree who will document actions / decisions and publish them
- .....28.05.2018 06:07:30, Michael Yeates [@michaelyeates]: >>unsupported service message type: messageActionChatJoinedByLink<<
- 28.05.2018 06:09:17, Eric - sw/eden [@xebb82]: We could log calls as comments on each point to discuss to make it as simple to follow as possible.
- I won't read any of the minutes and I reckon most of us don't have time tondo that anyways.
- .....28.05.2018 06:09:18, Kevin Rose - EOS New York [@eosnewyork]: >>unsupported service message type: messageActionChatJoinedByLink<<
- 28.05.2018 06:09:44, Kevin Rose - EOS New York [@eosnewyork]: Hey everyone
- .....28.05.2018 06:09:56, Kevin Rose - EOS New York [@eosnewyork]: >>unsupported service message type: messageActionChatAddUsers<<
- .....28.05.2018 06:10:54, Kevin Rose - EOS New York [@eosnewyork]: >>invited Warrick FitzGerald - EOS New York [@WarrickF]<<
- .....28.05.2018 06:15:14, Matthew Bryce 马修·布赖斯 - BitSpace [@matthewbryce]: >>unsupported service message type: messageActionChatJoinedByLink<<
- .....28.05.2018 06:17:10, Cryptosufi - 威杰 - BitSpace [@KryptoSufi]: >>unsupported service message type: messageActionChatJoinedByLink<<
- 28.05.2018 06:18:53, Eric - sw/eden [@xebb82]: Feel free to hit me up for a vc at any time if anyone needs explanations or clarifications regarding anything in the document.
- .....28.05.2018 06:20:41, Orchid K - EOSYS [@appletreehill]: >>unsupported service message type: messageActionChatJoinedByLink<<
- .....28.05.2018 06:25:54, Bohdan CryptoLions 🦁 [@BohdanCryptoLions]: >>unsupported service message type: messageActionChatJoinedByLink<<
- 28.05.2018 06:25:58, Bohdan CryptoLions 🦁 [@BohdanCryptoLions]: Hello all
- .....28.05.2018 06:28:01, David P - EOS42 [@GunnisonCap]: >>unsupported service message type: messageActionChatJoinedByLink<<
- 28.05.2018 06:36:03, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Hi All
- I’ve pinned the Google Document
- Please read and add any questions / requirements you feel need to be included in our EOS Validation Test Plan
- The end goal here is an agreement on how a chain should be validated as mainnet.
- Once the agreed Validation is complete each team should be comfortable to publicly declare the EOS Mainnet.
- .....28.05.2018 06:36:07, Jae Chung - HKEOS [@jaechung]: >>unsupported service message type: messageActionChatJoinedByLink<<
- 28.05.2018 06:41:22, Kyle - EOSVibes [@TheAwakenment]: Glad we’re doing this.
- Agreeing and declaring the process is equally as important as agreeing to the intent of launching one main net.
- .....28.05.2018 06:42:50, Roman CryptoLions [@romanskaskiw]: >>unsupported service message type: messageActionChatJoinedByLink<<
- 28.05.2018 06:55:30, Nathan Rempel | GenerEOS Sydney [@nsrempel]: whats the best way to add to the document so it doesn't get messy?
- 28.05.2018 06:55:33, David P - EOS42 [@GunnisonCap]: Fine by me
- 28.05.2018 06:56:37, David P - EOS42 [@GunnisonCap]: Middle of the night in Asia isn’t it, actually two calls would work best for global coverage
- 28.05.2018 06:57:01, Nathan Rempel | GenerEOS Sydney [@nsrempel]: Ok guys I just want to check in here
- 28.05.2018 06:57:32, Michael Yeates [@michaelyeates]: i think adding comments is the best way and then the owner can modify the doc
- 28.05.2018 06:57:32, Eric - sw/eden [@xebb82]: I'm trying to keep an eye on it and will try to help us clean it.
- 28.05.2018 06:57:44, Eric - sw/eden [@xebb82]: +1
- 28.05.2018 06:57:46, Nathan Rempel | GenerEOS Sydney [@nsrempel]: Everyone is aware that blocks are produced for free, forever, until 15% token supply is staked
- 28.05.2018 06:57:54, Nathan Rempel | GenerEOS Sydney [@nsrempel]: but blocks are being produced
- 28.05.2018 06:58:22, Eric - sw/eden [@xebb82]: yes
- 28.05.2018 06:58:30, Charles H - EOS 42 [@ankh2054]: Yes I believe so and the 21 BPs booting the chain cannot be changed either.
- 28.05.2018 06:58:36, Eric - sw/eden [@xebb82]: true
- 28.05.2018 06:58:53, Michael Yeates [@michaelyeates]: i think safest it to boot and then freeze for validation
- 28.05.2018 06:59:01, Nathan Rempel | GenerEOS Sydney [@nsrempel]: ok great, ive seen some misinformation around where somehow blocks wouldn't be produced at all heh
- 28.05.2018 06:59:07, Michael Yeates [@michaelyeates]: we need the producers to stick around to unfreeze
- 28.05.2018 06:59:23, Eric - sw/eden [@xebb82]: we are discussing freezing the chain as a validation step
- 28.05.2018 06:59:44, Bohdan CryptoLions 🦁 [@BohdanCryptoLions]: I think I know how to freeze chain
- 28.05.2018 06:59:49, Eric - sw/eden [@xebb82]: =)
- 28.05.2018 06:59:52, Bohdan CryptoLions 🦁 [@BohdanCryptoLions]: but didnt tested yet
- 28.05.2018 07:00:02, Eric - sw/eden [@xebb82]: We've tested quite a bit, so that's no problem.
- 28.05.2018 07:00:10, Igor Lins e Silva - EOS Rio [@igorls]: can't we just call /producer/pause ?
- 28.05.2018 07:00:13, Eric - sw/eden [@xebb82]: yes
- 28.05.2018 07:00:15, Charles H - EOS 42 [@ankh2054]: So we freeze and then validate, then unfreeze and allow votes.
- 28.05.2018 07:00:27, Igor Lins e Silva - EOS Rio [@igorls]: yes sir
- 28.05.2018 07:00:40, Eric - sw/eden [@xebb82]: That's one way of doing it that I could absolutely agree on.
- 28.05.2018 07:01:01, Charles H - EOS 42 [@ankh2054]: So we need to call producer pause on all or only one?
- 28.05.2018 07:01:06, Eric - sw/eden [@xebb82]: yes
- 28.05.2018 07:01:08, Eric - sw/eden [@xebb82]: all
- 28.05.2018 07:01:11, Michael Yeates [@michaelyeates]: +1
- 28.05.2018 07:01:18, Eric - sw/eden [@xebb82]: You control your own node via that api
- 28.05.2018 07:01:35, Michael Yeates [@michaelyeates]: or depending on firewall, other peoples nodes
- 28.05.2018 07:01:36, Bohdan CryptoLions 🦁 [@BohdanCryptoLions]: why we need to freeze?
- 28.05.2018 07:01:37, Eric - sw/eden [@xebb82]: We can jump on a call if u want more info Charles. Welcome back btw.
- 28.05.2018 07:01:44, Eric - sw/eden [@xebb82]: Validate state.
- 28.05.2018 07:01:47, Charles H - EOS 42 [@ankh2054]: Thanks :)
- 28.05.2018 07:01:53, Michael Yeates [@michaelyeates]: to prevent tokens moving while we validate
- 28.05.2018 07:02:04, Igor Lins e Silva - EOS Rio [@igorls]: 🙊
- 28.05.2018 07:02:10, Nathan Rempel | GenerEOS Sydney [@nsrempel]: @BohdanCryptoLions you have experience generating blocksnapshots, can you specify which block to produce the snapshot against
- 28.05.2018 07:03:02, Bohdan CryptoLions 🦁 [@BohdanCryptoLions]: but I think we can validate till when eosio resign .
- 28.05.2018 07:03:46, Kevin Rose - EOS New York [@eosnewyork]: Can our goal also be how to choose one from multiple valid and correct chains to be called the EOS mainnet?
- 28.05.2018 07:04:11, Eric - sw/eden [@xebb82]: Not if some of us wants to be able to say loop all existing accounts, and check balance and compare with file.
- 28.05.2018 07:05:00, Michael Yeates [@michaelyeates]: if people start transacting and then we have a problem with validation, we would have to create another chain and all those transactions would be reversed
- 28.05.2018 07:05:06, Roman CryptoLions [@romanskaskiw]: ❤️
- 28.05.2018 07:05:54, Charles H - EOS 42 [@ankh2054]: Yes and I wonder how we do this. Assuming somewhere else in the world they are discussing this right now too. How too boot and validate a chain.
- 28.05.2018 07:06:07, Nathan Rempel | GenerEOS Sydney [@nsrempel]: (Especially in China....)
- 28.05.2018 07:06:43, Charles H - EOS 42 [@ankh2054]: Does block.one have any official capacity to choose?
- 28.05.2018 07:06:57, Eric - sw/eden [@xebb82]: Been in contact with some of the BPCs from China. Some have joined in here already, other I rely messages to in private.
- 28.05.2018 07:06:58, Kevin Rose - EOS New York [@eosnewyork]: No
- 28.05.2018 07:07:25, Kevin Rose - EOS New York [@eosnewyork]: Chinese Block Producers are on the same page. We want one mainnet.
- 28.05.2018 07:07:31, Eric - sw/eden [@xebb82]: +1
- 28.05.2018 07:08:07, Nathan Rempel | GenerEOS Sydney [@nsrempel]: I appreciate that, but there may be some doing other boot tests?
- 28.05.2018 07:08:16, Kevin Rose - EOS New York [@eosnewyork]: [[Webpage]]As a thought starter, here is a proposal on how to randomly pick a chain. Feel free to tear it apart. https://docs.google.com/document/d/1Ya2l-Yimc1meyOBoSpE0QWX8ITaAYUwKvBMpRT-rInQ/edit?usp=sharing
- 28.05.2018 07:08:27, Eric - sw/eden [@xebb82]: I hope so. I hope most of us are / have been. Both publicly and in private.
- 28.05.2018 07:09:32, Kevin Rose - EOS New York [@eosnewyork]: Not perfect, not even final. Just a draft of thoughts.
- 28.05.2018 07:10:16, Bohdan CryptoLions 🦁 [@BohdanCryptoLions]: my thoughts are that we need to validate all chain from start until eosio user chnage rights to multisig (eosio.prods). after that boot is finished and changes to system can be done only by 2/3 producers. So you just would like just pause chaine for validation after eosio resign, am I right?
- 28.05.2018 07:11:27, Eric - sw/eden [@xebb82]: Yes, that's the plan. Well, atleast the proposition.
- 28.05.2018 07:11:32, David P - EOS42 [@GunnisonCap]: Fully agree with this
- 28.05.2018 07:12:31, Bohdan CryptoLions 🦁 [@BohdanCryptoLions]: I think its ok.
- 28.05.2018 07:13:06, Nathan Rempel | GenerEOS Sydney [@nsrempel]: How do you prevent third party transactions in the interim between network start and the freeze? The moment the system, bios, token transactions and snapshot are loaded third parties could publish transactions
- 28.05.2018 07:13:09, Eric - sw/eden [@xebb82]: 👍
- 28.05.2018 07:13:18, Nathan Rempel | GenerEOS Sydney [@nsrempel]: contracts*
- 28.05.2018 07:13:34, Eric - sw/eden [@xebb82]: We might need to start the chain with only eosio+abps connected from start.
- 28.05.2018 07:13:43, Eric - sw/eden [@xebb82]: Until freeze is active.
- 28.05.2018 07:14:29, David P - EOS42 [@GunnisonCap]: However this works, random section is the only way to do this - assuming multiple, identical term bios chains. That way there cannot become arguments between BPs regarding the kudos of being part of a launch 21. That will happen without this being randomised.
- 28.05.2018 07:14:47, Michael Yeates [@michaelyeates]: can we set it all up with only specified peers
- 28.05.2018 07:14:55, Michael Yeates [@michaelyeates]: should keep unwanted nodes out
- .....28.05.2018 07:26:05, Chitty 🅑 - EOS Argentina [@JChitty]: >>unsupported service message type: messageActionChatJoinedByLink<<
- 28.05.2018 07:26:37, Nathan Rempel | GenerEOS Sydney [@nsrempel]: hmm, if all producers in a boot process block the http server and delay launching "public exposed nodes" it should work... as long as the nominated producers actually follow those guidelines
- 28.05.2018 07:27:19, Eric - sw/eden [@xebb82]: absolutely
- 28.05.2018 07:27:58, Eric - sw/eden [@xebb82]: +1 :)
- 28.05.2018 07:28:21, Igor Lins e Silva - EOS Rio [@igorls]: works.. but its a very uncontrollabe solution
- 28.05.2018 07:28:52, Nathan Rempel | GenerEOS Sydney [@nsrempel]: yeah, most things in the document require 21 people to all follow the rules
- 28.05.2018 07:29:31, Eric - sw/eden [@xebb82]: At this point it's not dangerous though.
- We throw away the chain and restart if someone fucks up.
- 28.05.2018 07:29:47, Eric - sw/eden [@xebb82]: The danger starts coming when people have started voting. Which is why I like this solution.
- .....28.05.2018 07:29:48, CLains - BitSpace [@CLains]: >>unsupported service message type: messageActionChatJoinedByLink<<
- 28.05.2018 07:29:58, Eric - sw/eden [@xebb82]: If we loose peoples votes - that's bad PR.
- 28.05.2018 07:33:53, Nathan Rempel | GenerEOS Sydney [@nsrempel]: can we easily confirm via log that all transactions taking place have been legitimate by the boot node and also have been the only transactions? I'm assuming the transaction log you could easily spot if an unspecified account or a rogue boot producer intiated a transfer/etc
- 28.05.2018 07:36:22, Michael Yeates [@michaelyeates]: the transfer would have gone somewhere, so genesis balances would not validate and the chain would be tossed
- 28.05.2018 07:36:23, Eric - sw/eden [@xebb82]: We have a few different ideas for validation scripts.
- In eos-bios Alex validates each and every transaction bit-by-bit.
- Igor has a quite advanced script building up a database.
- An I'm planning to; loop all accounts, validate against ERC20 and make sure their code = 000000000. Loop ERC20 - validate against accounts. Validate contract hashes, and so forth.
- 28.05.2018 07:36:41, Michael Yeates [@michaelyeates]: the more methods for validation the better
- 28.05.2018 07:36:49, Eric - sw/eden [@xebb82]: +1
- 28.05.2018 07:36:55, Michael Yeates [@michaelyeates]: we all agree on final state so they should all match
- 28.05.2018 07:37:03, Eric - sw/eden [@xebb82]: Then we have the human layer if indeed something doesn't validate.
- 28.05.2018 07:37:34, Bohdan CryptoLions 🦁 [@BohdanCryptoLions]: +++
- 28.05.2018 07:37:37, Michael Yeates [@michaelyeates]: [[GIF, size 81'601 bytes]]
- 28.05.2018 07:38:24, Nathan Rempel | GenerEOS Sydney [@nsrempel]: alright cool, im happy to keep the chain private until validation, makes the most sense
- 28.05.2018 07:39:00, Nathan Rempel | GenerEOS Sydney [@nsrempel]: impetus would exist to validate as rapidly as possible and go public as early as possible to be "first to market"
- 28.05.2018 07:39:24, Bohdan CryptoLions 🦁 [@BohdanCryptoLions]: anyone did some validation tests, how long can it last?
- 28.05.2018 07:40:33, Eric - sw/eden [@xebb82]: Ping @igorls
- 28.05.2018 07:41:30, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Is the 3 Eric lists sufficient?
- If so then no need for “more” as part of the agreed validation plan.
- Let’s try name those approaches so they aren’t “Alex’s , Eric’s” etc...
- 28.05.2018 07:42:14, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Cycle 1 , Cycle 2 , Cycle 3
- 28.05.2018 07:42:58, Sharif Bouktila - EOS Dublin [@sharifbouktila]: And define what the approach is and how they are executed
- 28.05.2018 07:43:33, Michael Yeates [@michaelyeates]: we would have our own method, but im sure independent people will write their own
- 28.05.2018 07:44:34, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Sure but as the end goal here we want one agreed approach that once passed everyone is confident declaring a Mainnet
- 28.05.2018 07:46:06, Eric - sw/eden [@xebb82]: We have one more from @winlineosbixin
- 28.05.2018 07:46:34, Eric - sw/eden [@xebb82]: We also need to try to get more people to generate the snapshots probably, because.. maybe that tool can't be trusted? 🤷🏻♂️
- 28.05.2018 07:46:44, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Cool , cycle 4 ...
- 28.05.2018 07:46:51, Sharif Bouktila - EOS Dublin [@sharifbouktila]: +1
- 28.05.2018 07:48:46, Eric - sw/eden [@xebb82]: The plan with my validation isn't really first and foremost to do everything, or even to do everything as quick as possible.
- The plan is to make it as easy to read as possible with as few dependencies as possible. So anyone with a linux machine who has compiled the latest version can run it.
- 28.05.2018 07:48:52, Eric - sw/eden [@xebb82]: I will also be running at least Igors tool.
- 28.05.2018 07:50:43, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Try to think about what tests you would be comfortable for someone else to run and confirm back.
- 28.05.2018 07:51:13, Eric - sw/eden [@xebb82]: Sure thing.
- I'll be away for an hour trying to get some shut eye so I can be alive on interview tonight.
- .....28.05.2018 07:53:13, Marc-Antoine Ross - EOS Canada [@marcantoineross]: >>unsupported service message type: messageActionChatJoinedByLink<<
- 28.05.2018 07:55:27, Alexandre Bourget - EOS Canada [@abourget]: Yes, that'd be the end block for freeze.
- 28.05.2018 07:55:36, Michael Yeates [@michaelyeates]: I would say that any chain passing validation should be declared a Mainnet candidate
- 28.05.2018 07:56:43, Alexandre Bourget - EOS Canada [@abourget]: WinLin also wrote a validation script, taking into account what people have in eosio.stake
- 28.05.2018 07:56:48, Alexandre Bourget - EOS Canada [@abourget]: Is he here ?
- 28.05.2018 07:56:58, Michael Yeates [@michaelyeates]: The most important goal is to make the token holders confident that the chain represents the snapshot and that nobody has any special rights or code
- 28.05.2018 07:57:50, Michael Yeates [@michaelyeates]: Multiple independent scripts all confirming that is the gold standard
- 28.05.2018 07:59:43, Alexandre Bourget - EOS Canada [@abourget]: I have a program that can filter the p2p protocol, do deep packet or transaction inspection and could stop propagation of blocks after a given action (eosio::updateauth aka eosio resignation)
- 28.05.2018 08:00:07, Alexandre Bourget - EOS Canada [@abourget]: People could use it to do validation
- 28.05.2018 08:00:26, Michael Yeates [@michaelyeates]: We can pause production
- 28.05.2018 08:00:33, Alexandre Bourget - EOS Canada [@abourget]: That won't work
- 28.05.2018 08:01:06, Alexandre Bourget - EOS Canada [@abourget]: Unless you manage to tell everyone to stop their production at the same time. You only need one transaction in a block post-resignation for the process to fail.
- 28.05.2018 08:01:34, Alexandre Bourget - EOS Canada [@abourget]: Pausing production is only instructing your node to not produce when it's on schedule
- 28.05.2018 08:01:39, Michael Yeates [@michaelyeates]: You can do that on a closed network, then open for validation
- 28.05.2018 08:02:36, Alexandre Bourget - EOS Canada [@abourget]: The moment it is "open for validation", it is also open for transactions.
- 28.05.2018 08:03:00, Michael Yeates [@michaelyeates]: If a chain allows any transactions then it fails validation
- 28.05.2018 08:03:13, Michael Yeates [@michaelyeates]: How about we add that to the validation script?
- .....28.05.2018 08:07:43, Boy Maas - EOS Amsterdam [@monetashi]: >>unsupported service message type: messageActionChatJoinedByLink<<
- 28.05.2018 08:07:56, Sharif Bouktila - EOS Dublin [@sharifbouktila]: First past the post ?
- 28.05.2018 08:08:03, WinLin-EOSBIXIN [@hiwinlin]: [[Photo]]
- 28.05.2018 08:08:27, WinLin-EOSBIXIN [@hiwinlin]: eos_onchain_validator.py just test the 1000 account onchain validation
- 28.05.2018 08:08:54, WinLin-EOSBIXIN [@hiwinlin]: the speed is 851 accounts/sec for single node
- 28.05.2018 08:09:07, Michael Yeates [@michaelyeates]: validation is going to take a couple of days i think
- 28.05.2018 08:09:45, Sharif Bouktila - EOS Dublin [@sharifbouktila]: I don’t see any need to rush.
- 28.05.2018 08:10:07, Michael Yeates [@michaelyeates]: not at all
- 28.05.2018 08:10:19, Michael Yeates [@michaelyeates]: i would like at least 2 weeks for validation and testing
- 28.05.2018 08:10:42, WinLin-EOSBIXIN [@hiwinlin]: @abourget I test the eos-bios boot --single with the eos_validator_tools.py
- 28.05.2018 08:10:50, Michael Yeates [@michaelyeates]: if we lose 7 producers before the chain is opened then it is dead and we have to start again
- 28.05.2018 08:11:51, Bohdan CryptoLions 🦁 [@BohdanCryptoLions]: setprods now is in system contract, so probably will work even with eosio as multisig
- 28.05.2018 08:13:08, Michael Yeates [@michaelyeates]: the setprods wouldnt be irreversible though?
- 28.05.2018 08:13:18, Alexandre Bourget - EOS Canada [@abourget]: Is it ?? It wasn't when I checked... I submitted a PR to add it and it was refused.
- 28.05.2018 08:13:44, Michael Yeates [@michaelyeates]: [[Webpage]]https://github.com/EOSIO/eos/blob/master/contracts/eosio.system/eosio.system.abi#L490
- 28.05.2018 08:14:06, Bohdan CryptoLions 🦁 [@BohdanCryptoLions]: hehe you were faster
- 28.05.2018 08:14:14, Alexandre Bourget - EOS Canada [@abourget]: It's on the ABI so it doesn't break if you fetch blocks where eosio.bios was present. It's not in the code
- 28.05.2018 08:14:33, Alexandre Bourget - EOS Canada [@abourget]: It's only is eosio.bios
- 28.05.2018 08:15:16, Alexandre Bourget - EOS Canada [@abourget]: That's why we're doing that dance in the bootseq (system to bios, setprods, bios to system)
- 28.05.2018 08:15:54, Michael Yeates [@michaelyeates]: can we try to keep to validation requirements
- 28.05.2018 08:16:04, Michael Yeates [@michaelyeates]: what do we need to validate a particular chain
- 28.05.2018 08:16:14, Alexandre Bourget - EOS Canada [@abourget]: Also, schedule from setprods isn't altered until 15% votes threshold
- 28.05.2018 08:17:06, Sharif Bouktila - EOS Dublin [@sharifbouktila]: @abourget try think of what you would need to do to validate a chain which you weren’t involved in launching.
- 28.05.2018 08:17:37, Sharif Bouktila - EOS Dublin [@sharifbouktila]: We are trying to come up with a launch agnostic approach here
- 28.05.2018 08:19:30, Nathan Rempel | GenerEOS Sydney [@nsrempel]: Is this a soft setting from your boot process or a hard setting in the eosio code?
- 28.05.2018 08:21:11, WinLin-EOSBIXIN [@hiwinlin]: by the validation, I have found something:
- 1. TESTNET_TRUNCATE_SNAPSHOT: 1000 will inject 1001 indeed
- 2. do we need to create eosio.unregd account? and why issue 758877.2060 to eosio.unregd?
- @abourget
- 28.05.2018 08:22:41, WinLin-EOSBIXIN [@hiwinlin]: eos-bios should export a csv file which contains all the account was created in the boot progress
- 28.05.2018 08:24:09, WinLin-EOSBIXIN [@hiwinlin]: in the first time when I validate the chain, I got failed just because I don't know the create producer or eosio.uregd accounts
- .....28.05.2018 08:26:31, Patryk @perduta from Tokenika [@perduta]: >>unsupported service message type: messageActionChatJoinedByLink<<
- .....28.05.2018 08:28:25, Craig Murray - EOS BlockSmith [@craigmurray100]: >>unsupported service message type: messageActionChatJoinedByLink<<
- .....28.05.2018 08:32:57, Ricardo Marlière - EOS Brazil [@rbmarliere]: >>unsupported service message type: messageActionChatJoinedByLink<<
- .....28.05.2018 08:33:10, heipacker eosunion [@heipacker]: >>unsupported service message type: messageActionChatJoinedByLink<<
- 28.05.2018 08:38:34, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Hi All
- I’ve pinned the Google Document we are using.
- Please read and add any questions / requirements you feel need to be included in our EOS Validation Test Plan
- The end goal here is an agreement on how a chain should be validated as mainnet.
- Once the agreed Validation is complete each team should be comfortable to publicly declare the EOS Mainnet.
- .....28.05.2018 08:41:57, Shaheen Counts - EOS BlockSmith [@SeanPaul19]: >>unsupported service message type: messageActionChatJoinedByLink<<
- .....28.05.2018 08:42:51, Chitty 🅑 - EOS Argentina [@JChitty]: >>invited mini - EOS Argentina [@moonxor]<<
- 28.05.2018 09:03:01, Alexandre Bourget - EOS Canada [@abourget]: yes!! that's what I want as questions.. what can I provide that helps others validate ? I'll put that on the list !
- 28.05.2018 09:03:57, Alexandre Bourget - EOS Canada [@abourget]: what not export all the actions ? we could inspect the actions ?
- 28.05.2018 09:04:00, Alexandre Bourget - EOS Canada [@abourget]: wouldn't that be easier ?
- 28.05.2018 09:04:15, Alexandre Bourget - EOS Canada [@abourget]: if the actions are spit in a .jsonl file.. and we have confirmation that these actions are on chain..
- 28.05.2018 09:04:29, Alexandre Bourget - EOS Canada [@abourget]: checking against the action list might be also a good solution
- 28.05.2018 09:04:44, Alexandre Bourget - EOS Canada [@abourget]: I propose a video call to go through boot sequence proposals..
- .....28.05.2018 09:09:18, Eugene Luzgin - EOS Tribe [@eluzgin]: >>unsupported service message type: messageActionChatJoinedByLink<<
- 28.05.2018 09:10:05, Eugene Luzgin - EOS Tribe [@eluzgin]: Hello, just joined, catching up on messages here.
- 28.05.2018 09:12:18, mao EOSREAL [@maoyifeng]: sharif, do we fix the call time now?
- 28.05.2018 09:12:33, mao EOSREAL [@maoyifeng]: it's 9pm in BJ/HK/SG
- .....28.05.2018 09:13:51, Issay@OracleChain [@issayTseng]: >>unsupported service message type: messageActionChatJoinedByLink<<
- 28.05.2018 09:14:56, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Let’s leave it for now , it might be better to do calls / deep dives in smaller groups and then combine notes.
- What does @everyone think ??
- 28.05.2018 09:16:11, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Again the goal is to agree a Validation approach and test plan which an independent team could use to validate Mainnet
- It should work regardless of the boot approach.
- 28.05.2018 09:16:46, Michael Yeates [@michaelyeates]: it would be easier but a validation script cannot rely on anything provided by the boot
- 28.05.2018 09:39:48, Alexandre Bourget - EOS Canada [@abourget]: @winlineosbixin is demo'ing his validation in the Zoom meeting right now!
- 28.05.2018 09:40:11, Alexandre Bourget - EOS Canada [@abourget]: [[Webpage]]https://zoom.us/j/717170235
- .....28.05.2018 09:52:31, Leonard Dorlöchter - EOS network by Advanced Blockchain AG [@EOSnetwork]: >>unsupported service message type: messageActionChatAddUsers<<
- .....28.05.2018 09:56:39, Leonard Dorlöchter - EOS network by Advanced Blockchain AG [@EOSnetwork]: >>invited Anthony Lhériau | EOS Network by Advanced Blockchain AG [@alheriau]<<
- 28.05.2018 10:15:49, Alexandre Bourget - EOS Canada [@abourget]: woops :) next version will have 1000 equal 1000 :)
- 28.05.2018 10:16:24, Alexandre Bourget - EOS Canada [@abourget]: I think it makes sense to review the unregd pattern together.
- What about doing a videoconf with those involved in validation ?
- 28.05.2018 10:18:22, WinLin-EOSBIXIN [@hiwinlin]: ok
- 28.05.2018 10:19:42, WinLin-EOSBIXIN [@hiwinlin]: @abourget I have send you some suggestion please check to make the validate tool works w ell
- 28.05.2018 10:30:38, Alexandre Bourget - EOS Canada [@abourget]: thanks, I'll have eos-bios export all actions into a .jsonl file
- 28.05.2018 10:42:01, Alexandre Bourget - EOS Canada [@abourget]: ALL:
- Can we have a videoconf around 1700 UTC ?
- 28.05.2018 10:44:20, WinLin-EOSBIXIN [@hiwinlin]: it's 11:00 AM in beijing, I think it's ok
- 28.05.2018 10:45:09, Sharif Bouktila - EOS Dublin [@sharifbouktila]: I can join at 17:00 UTC today to facilitate.
- @abourget what is the agenda?
- I suggest we take some time to review the Pinned Document also.
- 28.05.2018 10:45:27, Alexandre Bourget - EOS Canada [@abourget]: Agenda would be review of bootseq
- 28.05.2018 10:45:41, Alexandre Bourget - EOS Canada [@abourget]: And how to validate it.
- 28.05.2018 11:00:23, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Can we agree that we need the same Validation approach to work regardless of the boot process ?
- 28.05.2018 11:35:46, Eric - sw/eden [@xebb82]: Is this 25minutes from now?
- 28.05.2018 11:36:12, Shaheen Counts - EOS BlockSmith [@SeanPaul19]: non stop video conferences this morning. Lol
- 28.05.2018 12:38:49, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Yes that’s the plan , who can Join??
- 28.05.2018 12:40:02, Eric - sw/eden [@xebb82]: I can join in. Had to run away to buy a mic, but I'm back now.
- 28.05.2018 12:40:42, Shaheen Counts - EOS BlockSmith [@SeanPaul19]: Craig and I are available
- 28.05.2018 12:41:37, Eric - sw/eden [@xebb82]: Do we have a link for that call?
- 28.05.2018 12:42:15, Alexandre Bourget - EOS Canada [@abourget]: didn't put any link yet.. it's still in 15 minutes ?
- 28.05.2018 12:44:44, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Yes, can you provide the Zoom Alex?
- 28.05.2018 12:50:43, Sharif Bouktila - EOS Dublin [@sharifbouktila]: For 100% clarity here we are joining this call to discuss how we agree an validation approach which could be used regardless of the Boot/Launch process. - Agreed ?
- 28.05.2018 12:50:57, Shaheen Counts - EOS BlockSmith [@SeanPaul19]: agreed
- 28.05.2018 12:51:17, Alexandre Bourget - EOS Canada [@abourget]: Yes
- 28.05.2018 12:51:41, Shaheen Counts - EOS BlockSmith [@SeanPaul19]: would be counterproductive to fall into a "which boot process is best" conversation
- 28.05.2018 12:52:22, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Ok lets keep it to 30 Mins
- 28.05.2018 12:52:26, Alexandre Bourget - EOS Canada [@abourget]: no no.. that's not the point.. I might use a yaml used in the process for the sake of discussion here.. hope it won't offend :) but it's more about the process.. and the actions..
- 28.05.2018 12:52:27, Alexandre Bourget - EOS Canada [@abourget]: 8 minutes now
- 28.05.2018 12:52:38, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Have you got a link
- 28.05.2018 12:52:45, Ricardo Marlière - EOS Brazil [@rbmarliere]: will it be in zoom or something else web based?
- 28.05.2018 12:54:43, Eric - sw/eden [@xebb82]: I can launch a zoom if needed.
- 28.05.2018 12:54:56, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Go for it
- 28.05.2018 12:55:09, Alexandre Bourget - EOS Canada [@abourget]: [[Webpage]]https://zoom.us/j/849742198
- 28.05.2018 12:56:40, Alexandre Bourget - EOS Canada [@abourget]: Someone might want to pin it..
- .....28.05.2018 13:14:59, Marshall - SaltBlock [@marshallslc]: >>unsupported service message type: messageActionChatJoinedByLink<<
- .....28.05.2018 13:38:56, Adrian - SaltBlock [@adrian_saltblock]: >>unsupported service message type: messageActionChatJoinedByLink<<
- 28.05.2018 15:25:54, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Great call, thanks to everyone that joined.
- For those who didn't much was discussed. I believe everyone needs to work together from here on in.
- I like the proposal of pausing the chain for a period (24 hours) to propose a "Validation Candidate" and let everyone at it.
- If we can agree this approach I don't think it matters what approach we use to launch it.
- 28.05.2018 15:26:36, Michael Yeates [@michaelyeates]: 👍 sorry i could not make it but it sounded like it was productive
- 28.05.2018 15:26:51, Sharif Bouktila - EOS Dublin [@sharifbouktila]: We should just aim to publish guidelines to how the community at large can validate for themselves
- 28.05.2018 15:26:53, Eric - sw/eden [@xebb82]: If it's not agreed that the chain validated correctly we could switch launch software tool
- 28.05.2018 15:27:34, Michael Yeates [@michaelyeates]: how long do we think it would take to validate, we would need a lot of individual confirmations over time
- 28.05.2018 15:28:14, Sharif Bouktila - EOS Dublin [@sharifbouktila]: We were suggesting 24 hours, and then open it up for voting / staking and see if it gets to 15%
- 28.05.2018 15:28:21, Sharif Bouktila - EOS Dublin [@sharifbouktila]: If no issues
- 28.05.2018 15:28:51, Michael Yeates [@michaelyeates]: what about software testing?
- 28.05.2018 15:28:52, Charles H - EOS 42 [@ankh2054]: [[Webpage]]http://producer.vote
- 28.05.2018 15:28:57, Roman CryptoLions [@romanskaskiw]: +
- 28.05.2018 15:29:19, Charles H - EOS 42 [@ankh2054]: Sharing the above - proxy voting pool. Hopefully this will help get to 15% faster.
- 28.05.2018 15:29:32, Sharif Bouktila - EOS Dublin [@sharifbouktila]: One suggestion I have on the weighting of BPs in BIOS that rather than 1-100 , it should be that a BP flags the ones that they know vs the ones that they don't. 1 or 0
- 28.05.2018 15:30:16, Charles H - EOS 42 [@ankh2054]: +1 on that idea. You know them plus you vouch for their technical ability.
- 28.05.2018 15:30:33, Charles H - EOS 42 [@ankh2054]: Or trust their technical ability.
- 28.05.2018 15:31:09, Michael Yeates [@michaelyeates]: 24 hours seems very rushed to me
- 28.05.2018 15:31:55, Eric - sw/eden [@xebb82]: +1, but I think we'll have time before its active 🤷♂
- 28.05.2018 15:31:57, Sharif Bouktila - EOS Dublin [@sharifbouktila]: We could say 24-72 hours, and make a call. If all validation checks are passing we can push on
- 28.05.2018 15:32:33, Eric - sw/eden [@xebb82]: We basically need to test all functionality that we can think of.
- 28.05.2018 15:32:57, Michael Yeates [@michaelyeates]: 15% is basically 2 people voting. i think validation will be very quick
- 28.05.2018 15:33:11, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Lets try get those test conditions in a document before
- 28.05.2018 15:33:47, Eric - sw/eden [@xebb82]: Sure. I'll help contribute to it.
- I have one started on one of my computers...
- 28.05.2018 15:34:05, Sharif Bouktila - EOS Dublin [@sharifbouktila]: throw it into the google doc pinned as a start
- 28.05.2018 15:35:54, Sharif Bouktila - EOS Dublin [@sharifbouktila]: The other questions we had was:
- 1. How to validate the snapshot?
- 2. Who can and is taking snapshots over the coming days as dry runs
- 3. Once we have a snapshot, what time do we start working at on June 2nd ??
- My vote is sometime around 2pm UTC
- .....28.05.2018 15:55:48, Craig Murray - EOS BlockSmith [@craigmurray100]: >>invited Syed Jafri | EOS Cafe / Calgary [@syed_jafri]<<
- 28.05.2018 15:56:46, Michael Yeates [@michaelyeates]: just a hash of the file should be ok for the snapshot
- 28.05.2018 15:59:58, Ricardo Marlière - EOS Brazil [@rbmarliere]: this
- 28.05.2018 16:00:03, Eric - sw/eden [@xebb82]: 1. Compare with each other. Maybe publish online to allow people to look for their keys and balances...?
- 2. I will look into running the snapshot creation process on multiple machines. Should be able to do it on at least 5...
- 28.05.2018 16:00:11, Craig Murray - EOS BlockSmith [@craigmurray100]: just want to say that that last meeting was pretty outstanding in my view. A great deal of ground was covered and I think we're getting MUCH closer to the point where the very top technical groups are going to be ABPs on the same chain.
- 28.05.2018 16:00:23, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: Syncing parity takes days, better get started soon
- 28.05.2018 16:00:28, Eric - sw/eden [@xebb82]: +1
- 28.05.2018 16:00:40, Michael Yeates [@michaelyeates]: think eosauthority are doing this
- 28.05.2018 16:00:58, Eric - sw/eden [@xebb82]: Have it synced already. Thinking of starting from that synced chain but on multiple machines.
- The chain should be the same... Right? :)
- 28.05.2018 16:02:12, Ricardo Marlière - EOS Brazil [@rbmarliere]: 1 yes definitely
- 28.05.2018 16:02:34, Sam - EOS Dublin [@samnoble]: Is there a guid to getting this up and running? We can spin up a box to run this on
- 28.05.2018 16:03:06, Igor Lins e Silva - EOS Rio [@igorls]: @EOSIOSnapshot
- 28.05.2018 16:03:27, Sam - EOS Dublin [@samnoble]: Thanks
- 28.05.2018 16:54:24, Sharif Bouktila - EOS Dublin [@sharifbouktila]: [[Webpage]]Just posted this - URGENT & AWESOME - https://medium.com/@eosdublin/urgent-awesome-f3e310a4cd42
- .....28.05.2018 17:40:54, Roshan - EOS Authority [@telepound]: >>unsupported service message type: messageActionChatJoinedByLink<<
- 28.05.2018 17:42:51, Roshan - EOS Authority [@telepound]: Yes. We have been running the snapshot code for quite some time now. The new fallback method works well too.
- 28.05.2018 17:46:38, Michael Yeates [@michaelyeates]: has anyone thought about the requirements for external tools such as wallets before we launch?
- 28.05.2018 17:46:45, Michael Yeates [@michaelyeates]: is block one making one?
- 28.05.2018 17:46:59, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: Yes scatter is making one, I do not think block.one will release one right now
- 28.05.2018 17:47:22, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: I have been working on a paper wallet
- 28.05.2018 17:47:37, Michael Yeates [@michaelyeates]: we need to have something ready or there will be so many scam wallets filling the void
- 28.05.2018 17:47:45, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Any idea when after the freeze when a snapshot will be available ?
- Trying to piece together a rough timeline for launch.
- 28.05.2018 17:48:11, Ricardo Marlière - EOS Brazil [@rbmarliere]: i believe scatter will be ready by then
- 28.05.2018 17:48:36, Michael Yeates [@michaelyeates]: is it something else? nathan keeps saying scatter isnt a wallet
- 28.05.2018 17:49:08, Ricardo Marlière - EOS Brazil [@rbmarliere]: I thought we could sign transactions with it?
- 28.05.2018 17:49:11, Sharif Bouktila - EOS Dublin [@sharifbouktila]: I think they built a wallet
- 28.05.2018 17:49:31, Roshan - EOS Authority [@telepound]: The snapshot run can take anywhere from 3 hours to 6 hours. Assuming you have a fully synced parity node. I assume we will have it within 6 hours of freeze max.
- .....28.05.2018 17:50:35, Roshan - EOS Authority [@telepound]: >>invited Rohan - EOSauthority [@RohanWillAnswer]<<
- 28.05.2018 17:51:17, Sharif Bouktila - EOS Dublin [@sharifbouktila]: I can probably find it but what time is freeze on the 1st ?
- 28.05.2018 17:51:55, Michael Yeates [@michaelyeates]: lets be careful with expectations on a time frame here
- 28.05.2018 17:52:03, Michael Yeates [@michaelyeates]: better to over deliver
- 28.05.2018 17:52:15, Sharif Bouktila - EOS Dublin [@sharifbouktila]: I’m just try suggest the start
- 28.05.2018 17:52:41, Roshan - EOS Authority [@telepound]: ERC20 EOS tokens are frozen 23 hours after June 1, 2018 at 22:59:59 UTC. So June 2nd 21:59:59
- 28.05.2018 17:52:46, Roshan - EOS Authority [@telepound]: 👍
- 28.05.2018 17:54:06, Sharif Bouktila - EOS Dublin [@sharifbouktila]: So snapshots ready circa 04:00 UTC 3rd June ?
- 28.05.2018 17:55:54, Roshan - EOS Authority [@telepound]: Let's say latest 8 am UTC
- 28.05.2018 17:56:18, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Ok cool
- 28.05.2018 17:58:47, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: Have we agreed on an onchain validation from these snapshots
- 28.05.2018 17:59:01, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: Last time I checked, I believe the different tools were not matching up
- 28.05.2018 17:59:22, Sharif Bouktila - EOS Dublin [@sharifbouktila]: On which chain?
- 28.05.2018 17:59:46, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: Like the chain that is going to be launched
- 28.05.2018 17:59:52, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: I never tested myself but I heard concerns
- 28.05.2018 17:59:56, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: of no consensus between tools
- 28.05.2018 18:00:30, Sharif Bouktila - EOS Dublin [@sharifbouktila]: nothing is agreed yet
- 28.05.2018 18:02:09, Rohan - EOSauthority [@RohanWillAnswer]: I’ve synced up to period 345. If anyone has a snapshot.json to compare with mine? I will send mine tomorrow.
- 28.05.2018 18:12:55, Michael Yeates [@michaelyeates]: i think we should avoid consensus with which tool to use, they should all agree that the tokens on-chain match those in the snapshot
- .....28.05.2018 18:15:23, Robert DeWilder - Worbli [@BrownDonahue]: >>unsupported service message type: messageActionChatJoinedByLink<<
- 28.05.2018 18:22:13, Rohan - EOSauthority [@RohanWillAnswer]: Bodhan and I compared our versions few days ago. Period 342 and ran a simple diff to get quite a few discrepancies including in the snapshot.json files
- 28.05.2018 18:24:42, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: Hmm, any idea why?
- 28.05.2018 18:30:40, Rohan - EOSauthority [@RohanWillAnswer]: There were quite a lot of differences. When I run it a couple of times after a truncate for example, the output matches exactly what I got both times.
- 28.05.2018 18:33:05, Rohan - EOSauthority [@RohanWillAnswer]: We seem to be missing @sandw1ch here. Any reason we can’t have him here? His expertise will be invaluable.
- 28.05.2018 18:35:21, Sharif Bouktila - EOS Dublin [@sharifbouktila]: I’ve no issue
- 28.05.2018 18:35:58, Sharif Bouktila - EOS Dublin [@sharifbouktila]: The more positive contribution the better!
- 28.05.2018 18:38:11, David P - EOS42 [@GunnisonCap]: Block.one cannot have anything to do with the launch
- 28.05.2018 18:39:05, David P - EOS42 [@GunnisonCap]: I would suggest they will ask to not be included. Referencing them directly for views is probably the best course.
- 28.05.2018 18:40:53, Rohan - EOSauthority [@RohanWillAnswer]: I suspected that might be the case and wanted to check.
- 28.05.2018 18:42:59, David P - EOS42 [@GunnisonCap]: Yep well worth it, there’s a few we would all benefit from being here right now - we’re on our own to work through all this now. We need daily calls like the one today going forward.
- 28.05.2018 18:48:05, Michael Yeates [@michaelyeates]: discussion about validating snapshot.csv files should be in the snapshot channel. we can check that given a consistent input, we can provide consistent validation of the on chain balances. this should happen regardless of the boot method
- 28.05.2018 18:49:09, Kevin Rose - EOS New York [@eosnewyork]: Hey all. Cypherglass asked me to add them. In the spirit of including those who wish to contribute, I'm going to oblige.
- 28.05.2018 19:10:30, aza eosar [@p17er0bl3do]: been thinking, both methods of validation (freeze/snap) that have been proposed have the same basic requirements, agree on a block, check the state there, avoid before that block any transaction that is not part of the boot sequence. This last part is what i'm more unsure about how to enforce. But if we share beforehand they pubs of all BPCs we can use the peer-key param to exclude everyone who is not a BPC
- .....28.05.2018 19:22:05, Kevin Rose - EOS New York [@eosnewyork]: >>invited James Sutherland - Cypherglass.com [@JamesSutherland]<<
- .....28.05.2018 19:22:26, Kevin Rose - EOS New York [@eosnewyork]: >>invited Adrianna Mendez - Cypherglass [@AdriMendez]<<
- .....28.05.2018 20:33:51, James Sutherland - Cypherglass.com [@JamesSutherland]: >>invited Ross - Cypherglass [@Ross_Cypherglass]<<
- .....28.05.2018 20:34:26, Adrianna Mendez - Cypherglass [@AdriMendez]: >>unsupported service message type: messageActionChatAddUsers<<
- 28.05.2018 20:40:11, Eric - sw/eden [@xebb82]: Yes, this was discussed as one of the possible solutions to that problem. Works very well in combination with the pause command.
- 28.05.2018 20:45:54, Michael Yeates [@michaelyeates]: all of the ricardian contracts is something we need to think about for chain validation. they need to be in the codebase and agreed before launch
- 28.05.2018 20:46:29, Kevin Rose - EOS New York [@eosnewyork]: My understanding is that there is no constitution ratification period now. What we load up will be the constitution.
- 28.05.2018 20:46:50, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: Really..
- 28.05.2018 20:47:45, Kevin Rose - EOS New York [@eosnewyork]: Tokens are unlocked after staking period. Constitutional Hash (CH) will start signing all transactions immediately. Basically ratified at chain activation which means what we load is the constitution.
- 28.05.2018 20:48:06, Kevin Rose - EOS New York [@eosnewyork]: Correct me if I'm wrong, please.
- 28.05.2018 20:49:53, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: I think you are right..which means voters have to agree to it or get off the only EOS chain
- 28.05.2018 20:50:00, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: Just seems wrong
- 28.05.2018 20:51:44, Nathan Rempel | GenerEOS Sydney [@nsrempel]: Nothing is locked... only block pay, based on reading the eosio contracts
- 28.05.2018 20:52:43, Eric - sw/eden [@xebb82]: The schedule.
- 28.05.2018 20:52:57, Kevin Rose - EOS New York [@eosnewyork]: I know the recent launch conference call discussed not staking all tokens by default but what you're saying is contrary to what I understood. I thought all tokens are staked for bandwidth by default at the moment of issuance. They are only available for reclamation after the 3 day staking period has ended. Am I wrong?
- 28.05.2018 20:54:22, Nathan Rempel | GenerEOS Sydney [@nsrempel]: Wow, I missed that call (thanks Aussie time). If that's true, that would seem to be up to the boot process. I'd be pretty pissed if my personal tokens were already staked
- 28.05.2018 20:54:50, Nathan Rempel | GenerEOS Sydney [@nsrempel]: What if i wanted to use those to create a new account and deploy a smart contract right away? I'd be furious
- 28.05.2018 20:56:05, Eric - sw/eden [@xebb82]: I'd be furious if someone decided that I can't create whichever account name I wanted, but that they themselves reserved "b1".
- 28.05.2018 20:56:15, Michael Yeates [@michaelyeates]: lol
- 28.05.2018 20:57:00, Nathan Rempel | GenerEOS Sydney [@nsrempel]: Create your own blockchain and you can have all the names you want :)
- 28.05.2018 20:57:38, Eric - sw/eden [@xebb82]: 😂
- 28.05.2018 20:58:10, Kevin Rose - EOS New York [@eosnewyork]: We said one mainnet Nathan! One! 😂
- 28.05.2018 20:58:19, Ross (EOSphere) [@rossco99]: hah!
- 28.05.2018 20:58:31, Nathan Rempel | GenerEOS Sydney [@nsrempel]: I have heard the schedule is locked, though I didn't hear whether that was by the boot bios or if its built in? Can someone link the github code?
- 28.05.2018 20:59:39, Nathan Rempel | GenerEOS Sydney [@nsrempel]: One EOS mainnet, but sw/eden can create their own blockchain core code if they wanted ;)
- 28.05.2018 20:59:40, Michael Yeates [@michaelyeates]: its built in
- 28.05.2018 21:00:17, Michael Yeates [@michaelyeates]: or in the system contract, sorry.. late
- 28.05.2018 21:00:42, Nathan Rempel | GenerEOS Sydney [@nsrempel]: @michaelyeates that makes sense. i suppose theirs no provision that "forces" the original producers out? they can still be voted for at the moment?
- 28.05.2018 21:02:08, Michael Yeates [@michaelyeates]: no, and i think they have to keep producing irreversible blocks
- 28.05.2018 21:02:25, Michael Yeates [@michaelyeates]: otherwise voting will not work and we have a dead chain
- 28.05.2018 21:02:54, Michael Yeates [@michaelyeates]: thats why i would prefer to keep many chains open during the validation phase
- 28.05.2018 21:06:01, Nathan Rempel | GenerEOS Sydney [@nsrempel]: yeah i agree the original ABP will keep producing blocks... just wasn't sure if their is a distinction between "boot producers" and "voted producers" in the codebase
- 28.05.2018 21:06:18, Nathan Rempel | GenerEOS Sydney [@nsrempel]: I would also prefer to keep the chains open and active if possible
- .....28.05.2018 22:32:53, Marc-Antoine Ross - EOS Canada [@marcantoineross]: >>invited Patrick B - Aloha EOS<<
- .....28.05.2018 22:41:38, Patrick B - Aloha EOS: >>invited Nate D - Aloha EOS [@n8dee]<<
- .....28.05.2018 22:43:42, Jae Chung - HKEOS [@jaechung]: >>invited Santhosh EOSNodeOne [@kamsa74]<<
- .....28.05.2018 23:37:00, Santhosh EOSNodeOne [@kamsa74]: >>invited Robinson EOS NodeOne [@aweekago]<<
- .....29.05.2018 00:21:28, Joe EOS Asia [@gaijinJOE]: >>unsupported service message type: messageActionChatJoinedByLink<<
- .....29.05.2018 00:59:24, Heejae - EOSYS.io [@heejaejoo]: >>unsupported service message type: messageActionChatJoinedByLink<<
- .....29.05.2018 02:39:25, nahu [@naueh]: >>unsupported service message type: messageActionChatJoinedByLink<<
- 29.05.2018 03:26:52, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Morning / Afternoon / Evening @everyone
- To the new guys in here, we have a Google Doc pinned , please review and add / comment as you see fit.
- We will be organising another call today.
- How many can make a call at 17:00 UTC ??
- 29.05.2018 04:27:13, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Some questions
- We said yesterday that anyone should be able to validate the snapshot and we should allow anyone to run their own tests to validate the chain.
- 1. How should someone outside the launch group make contact to flag an issue?
- 2. What channel should we use to communicate progress / Milestones to the outside world? eg. Chain is validated / Open for Staking / Mainnet ??
- 3. Where do we publish endpoints?
- 29.05.2018 04:36:49, Eric - sw/eden [@xebb82]: Please explain why you think pausing is a bad idea.
- 29.05.2018 04:44:00, Adrian - SaltBlock [@adrian_saltblock]: So many new names and faces. :)
- .....29.05.2018 05:15:30, Pete BlockMatrix [@blockmatrixpete]: >>unsupported service message type: messageActionChatJoinedByLink<<
- 29.05.2018 05:18:56, aza eosar [@p17er0bl3do]: i don't see any improvement on taking a snapshot and we can use that time it runs to test its stable and clear any doubts on its stability to other potencial boot groups. The idea is to call pause on all producers?
- What's really a bad idea (to me) is modifying the contracts to fit our needs as BPC at boot, we should keep EOS as vanilla as possible, but dont know if thats still on the table.
- 29.05.2018 05:27:26, Eric - sw/eden [@xebb82]: 1. How do we take a snapshot?
- 2. Do we really want to test stability on the live production chain before it has been accepted as validated?
- We have to run multiple chains to test all functionality anyways? Isn't it at this point better to focus on that?
- 3. I agre on the contracts. It's not needed.
- 29.05.2018 05:28:19, Eric - sw/eden [@xebb82]: Yea, the idea is to call pause on all producers to allow 24? 48? 72? hours to validate the chain.
- Allow anyone to connect in and validate the state of the blockchain before we start accepting votes and other transactions.
- 29.05.2018 05:29:20, David P - EOS42 [@GunnisonCap]: How long is required for validation? I don’t understand in my simple perspective why we would talk about up to 3 days for validation?
- 29.05.2018 05:30:46, Eric - sw/eden [@xebb82]: We want everyone in the community to get the chance to validate everything.
- We aren't even sure that the snapshot generation tool will work. We want to know if we missed some account or balance from the ERC20 distribution. It's not just us active producer candidates that should do validation.
- 29.05.2018 05:32:09, Eric - sw/eden [@xebb82]: Possibly we need more than 72 hours, since ... if people haven't pre-synced parity, it takes 3 days just to sync that to actually be able to create their own snapshots to validate against.
- 29.05.2018 05:32:59, Eric - sw/eden [@xebb82]: We are launching a multi billion dollar network. This can't be stressed enough.
- Normally the test phase should be weeks or months for basically any fix that's added.
- 29.05.2018 05:33:43, Adrian - SaltBlock [@adrian_saltblock]: This.
- 29.05.2018 05:34:30, Sharif Bouktila - EOS Dublin [@sharifbouktila]: But it needs to be agreed and clearly communicated
- 29.05.2018 05:34:38, Adrian - SaltBlock [@adrian_saltblock]: Given the sheer amount of eyes on this blockchain project, we really can't afford to get it wrong based on something we could have sorted out through proper validation (which at this point, does indeed seem a bit impossible)
- 29.05.2018 05:34:46, Sharif Bouktila - EOS Dublin [@sharifbouktila]: And we need to have a forum / method for that communication
- 29.05.2018 05:35:03, Eric - sw/eden [@xebb82]: 👍
- 29.05.2018 05:35:54, Issay@OracleChain [@issayTseng]: http://blogs.360.cn/blog/eos-node-remote-code-execution-vulnerability/
- 29.05.2018 05:36:18, Issay@OracleChain [@issayTseng]: 360 Says EOS Epic Vulnerability Can Be Fully Controlled by Trading Detonation System
- .....29.05.2018 05:37:54, Matthew Bryce 马修·布赖斯 - BitSpace [@matthewbryce]: >>invited James David Tandy - 詹姆士 - Bitspace [@jimmythesaint]<<
- 29.05.2018 05:43:28, David P - EOS42 [@GunnisonCap]: So in essence we need to communicate out clearly to the global community that there will be a complete freeze on all EOS token trading for nearly a week in that case. That's not going to down well with them.
- 29.05.2018 05:43:40, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Can I propose we have 4 phases to communicate
- Launch - Snapshot Verification & BIOS Boot
- Pause - Validation
- Enable - Open for staking until 15%
- Release - Declaration of Mainnet
- 29.05.2018 05:44:20, David P - EOS42 [@GunnisonCap]: I don't really get it - if a number of trusted parties validate, then what's to stop the chain launching? If others want to retrospectively validate as well then they can.
- 29.05.2018 05:44:31, David P - EOS42 [@GunnisonCap]: I don't understand why we're talking about such a long time period.
- 29.05.2018 05:44:49, Eric - sw/eden [@xebb82]: Until recent changes to the code this is how we had been told it should work.
- 29.05.2018 05:45:24, Eric - sw/eden [@xebb82]: 👏
- 29.05.2018 05:45:32, David P - EOS42 [@GunnisonCap]: That makes complete sense - we need a joint statement from as many BP's as possible on this approach if agreed.
- 29.05.2018 05:45:34, Michael Yeates [@michaelyeates]: before launch we will need snapshot validation
- 29.05.2018 05:45:41, Michael Yeates [@michaelyeates]: otherwise great :)
- 29.05.2018 05:46:10, Eric - sw/eden [@xebb82]: +1
- 29.05.2018 05:46:18, Charles H - EOS 42 [@ankh2054]: have we figures out a way to pause reliably?
- 29.05.2018 05:46:22, Eric - sw/eden [@xebb82]: yes
- 29.05.2018 05:46:22, Igor Lins e Silva - EOS Rio [@igorls]: 👍
- 29.05.2018 05:46:26, Charles H - EOS 42 [@ankh2054]: great :)
- 29.05.2018 05:46:27, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Launch - Snapshot Verification & BIOS Boot
- Pause - Validation
- Enable - Open for staking until 15%
- Release - Declaration of Mainnet
- 29.05.2018 05:46:34, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Lets keep it as 4 phases
- 29.05.2018 05:46:44, Adrian - SaltBlock [@adrian_saltblock]: I think that sounds good Sharif.
- 29.05.2018 05:47:40, Eric - sw/eden [@xebb82]: I'm all for that. Great idea Sharif.
- 29.05.2018 05:47:49, Michael Yeates [@michaelyeates]: does anyone have any thoughts on generation and validation of the snapshot?
- 29.05.2018 05:48:01, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Rollback and restart is possible at any phase before RELEASE - that should be the message.
- 29.05.2018 05:48:39, Eric - sw/eden [@xebb82]: +1 on that
- 29.05.2018 05:50:44, Charles H - EOS 42 [@ankh2054]: +1
- 29.05.2018 05:50:54, Sharif Bouktila - EOS Dublin [@sharifbouktila]: We don't then have to give time lines. If we agree what Validation is needed, once its complete we can move to Enable
- 29.05.2018 05:51:05, Eric - sw/eden [@xebb82]: Open for "voting", all will be staked already. right?=
- 29.05.2018 05:51:41, Sharif Bouktila - EOS Dublin [@sharifbouktila]: I believe so
- 29.05.2018 06:51:02, aza eosar [@p17er0bl3do]: i guess we can both take the snap and pause it, just lets make a test of pausing it to see we can recover ok from it
- 29.05.2018 06:54:24, Eric - sw/eden [@xebb82]: We have tested pausing chains many times already. But yes. There are a lot we need to test, and the pause time is perfect for that.
- The ABPs just fork up another chain. And test on that one.
- 29.05.2018 06:56:12, David P - EOS42 [@GunnisonCap]: So I'm clear, when and how do we agree that "validation is complete" as opposed to "completed by some"?
- 29.05.2018 06:56:37, Kevin Rose - EOS New York [@eosnewyork]: We need to very clearly set expectations to the world. Sharif you should draft another joint statement that each BP can publish.
- 29.05.2018 06:56:48, Michael Yeates [@michaelyeates]: i guess we would create a window and then if no bad reports come in we move on
- 29.05.2018 06:57:02, Kevin Rose - EOS New York [@eosnewyork]: It needs to be communicated as soon as possible
- 29.05.2018 06:57:10, David P - EOS42 [@GunnisonCap]: And my view is that we have a responsibility to clearly set expectations for the community - we can't sit here saying "we don't have to give timelines". I think we do need to communicate something clearly and soon.
- 29.05.2018 06:57:16, David P - EOS42 [@GunnisonCap]: Ah, yes - same as Kevin says.
- 29.05.2018 06:57:32, Michael Yeates [@michaelyeates]: igor is preparing a flowchart on mainnet launch
- 29.05.2018 06:57:39, Michael Yeates [@michaelyeates]: he will share in a moment
- 29.05.2018 06:58:17, David P - EOS42 [@GunnisonCap]: That seems the only logical way - nothing reported after say 12hrs (or whatever is agreed), considered validated and move on
- 29.05.2018 06:58:27, David P - EOS42 [@GunnisonCap]: Great
- 29.05.2018 07:06:22, Sharif Bouktila - EOS Dublin [@sharifbouktila]: We’ll aim to get a statement ready for tomorrow
- @GunnisonCap @eosnewyork can you round the groups so they are prepared like the one which went out on Friday ?
- 29.05.2018 07:09:59, Sharif Bouktila - EOS Dublin [@sharifbouktila]: We also need to decide how we communicate Endpoint(s) available for validation - don’t we??
- 29.05.2018 07:11:20, Kevin Rose - EOS New York [@eosnewyork]: Just so we’re saying the same thing the purpose of this statement is to set expectations of launch timing with explanation. We should keep it hyper focused.
- 29.05.2018 07:11:46, Kevin Rose - EOS New York [@eosnewyork]: Unless you weren’t suggesting that be added :)
- 29.05.2018 07:13:17, David P - EOS42 [@GunnisonCap]: Grand, that should be an aim of today then. I'll talk to @eosnewyork about coordinating this.
- 29.05.2018 07:14:59, David P - EOS42 [@GunnisonCap]: Here's a key point, we can give estimated time ranges for the different phases by all means. We don't need to tie ourselves to best case or worst, but setting community expectations is our collective responsibility on this too.
- 29.05.2018 07:16:14, Sharif Bouktila - EOS Dublin [@sharifbouktila]: I think just the 4 phases , How we will announce each one and a commitment that we will publish our validation approach.
- 29.05.2018 07:17:19, Khosi Morafo eosAfrica [@khosimorafo]: 👏👏👏👏👏👏
- 29.05.2018 07:19:32, Roman CryptoLions [@romanskaskiw]: +1
- 29.05.2018 07:22:02, Khosi Morafo eosAfrica [@khosimorafo]: 👍👍👍
- 29.05.2018 07:23:47, Charles H - EOS 42 [@ankh2054]: If we are to pause the chain and we also need to decide who are the ABP’s to to boot and whether we use wiregaurd VPN OR simply setup firewall rules and OR # peer-key.
- 29.05.2018 07:27:02, Kevin Rose - EOS New York [@eosnewyork]: +1
- 29.05.2018 07:28:02, Adrian - SaltBlock [@adrian_saltblock]: I was kind of hoping it would show all the +1 in that quote 😉
- 29.05.2018 07:28:59, Kevin Rose - EOS New York [@eosnewyork]: Yea me too. Boo telegram
- 29.05.2018 07:29:10, Adrian - SaltBlock [@adrian_saltblock]: Feature request.
- 29.05.2018 07:29:32, Adrian - SaltBlock [@adrian_saltblock]: (useless, but could get exciting if enough people leveraged the feature...)
- 29.05.2018 07:29:35, Kevin Rose - EOS New York [@eosnewyork]: +1
- 29.05.2018 07:29:40, Charles H - EOS 42 [@ankh2054]: Should we all nominate some people on our teams to provide community updates
- 29.05.2018 07:30:18, Charles H - EOS 42 [@ankh2054]: When we reach each phase we do a collective tweet and publish on a website.
- 29.05.2018 07:31:56, Kevin Rose - EOS New York [@eosnewyork]: Have the person with the idea write the original draft every time. Have @GunnisonCap and I execute on it.
- 29.05.2018 07:33:23, Khosi Morafo eosAfrica [@khosimorafo]: Well, who are the best communicators? Keeping in mind that should something go wrong, these will be the go-to people for the press.
- 29.05.2018 07:33:51, Charles H - EOS 42 [@ankh2054]: +1 for Kevin and David P.
- 29.05.2018 07:34:16, Kevin Rose - EOS New York [@eosnewyork]: What I mean is that @GunnisonCap and I work to disseminate the statement to all BPs who agree for publishing. Every BP is an access point for the press. We can’t designate someone for that
- 29.05.2018 07:34:48, Charles H - EOS 42 [@ankh2054]: Enable - Open for staking until 15% - on this release would be good to add some links where people can vote.
- 29.05.2018 07:36:47, Khosi Morafo eosAfrica [@khosimorafo]: I do not think its a bad thing for the community to have one or two people who will tell a single story. More people may risk more FUD. But I say that accepting that in a distributed system like EOS, some may sulk at the idea.
- 29.05.2018 07:38:06, Kevin Rose - EOS New York [@eosnewyork]: Normally I prefer majority buy in on ideas and nominations but we don’t have the luxury of time. Let @GunnisonCap and I worry about taking a drafted statement from the idea creator and getting it out there to every BP. Every BP can publish the statements we edit. We only ever publish statements that the group approves of.
- 29.05.2018 07:38:42, Kevin Rose - EOS New York [@eosnewyork]: No one BPs name will ever be on them. They are signed by “EOS Mainnet Launch Group” or something to that effect.
- 29.05.2018 07:38:47, Charles H - EOS 42 [@ankh2054]: +1
- 29.05.2018 07:38:54, Kevin Rose - EOS New York [@eosnewyork]: And let’s return to the important task of detailing how launch will go.
- 29.05.2018 07:41:18, Khosi Morafo eosAfrica [@khosimorafo]: I would have thought properly communicating that very launch is as important. But cool, let's move on.
- 29.05.2018 07:42:15, Igor Lins e Silva - EOS Rio [@igorls]: [[Webpage]]https://zoom.us/j/427431126
- 29.05.2018 07:42:19, Kevin Rose - EOS New York [@eosnewyork]: No it is, absolutely. Sorry I didn’t mean to sound that way.
- 29.05.2018 07:42:25, Igor Lins e Silva - EOS Rio [@igorls]: Join now
- .....29.05.2018 07:43:38, Craig Murray - EOS BlockSmith [@craigmurray100]: >>invited Jiri Jetmar | EOS Germany | [email protected] [@jirijetmar]<<
- 29.05.2018 07:44:12, Khosi Morafo eosAfrica [@khosimorafo]: ☺😉😍
- .....29.05.2018 07:44:16, Santhosh EOSNodeOne [@kamsa74]: >>invited Hahn - EOS NodeOne [@hahnryu]<<
- 29.05.2018 07:50:12, Kevin Rose - EOS New York [@eosnewyork]: @deckw
- 29.05.2018 07:50:22, Eric - sw/eden [@xebb82]: He's in
- 29.05.2018 07:51:00, Jiri Jetmar | EOS Germany | [email protected] [@jirijetmar]: Thx!
- 29.05.2018 07:54:00, Kevin Rose - EOS New York [@eosnewyork]: Thank you
- 29.05.2018 08:30:58, David P - EOS42 [@GunnisonCap]: Interesting call, I'll brief you later
- 29.05.2018 08:45:58, Craig Murray - EOS BlockSmith [@craigmurray100]: another amazing meeting. Nothing like deadline pressure to bring everyone together like one big happy family :) props to everyone for their coordination and cooperation
- 29.05.2018 09:25:05, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Re: Communications out. I've just nabbed @eoslaunch on Twitter.
- Could we use that for official communications for Launch and then all the other BPs RT from it?
- Could the Asian BPs pick it up and recast on WeChat or similar ?
- If its useless I'm sure some team would fancy launching another BP under that brand 🙂
- 29.05.2018 09:26:14, Kevin Rose - EOS New York [@eosnewyork]: I kind of like the idea. Let’s think about this.
- 29.05.2018 09:28:09, David P - EOS42 [@GunnisonCap]: I'm certainly hoping for at least another 50 BP's launching between now and June 3, we don't have enough
- 29.05.2018 09:35:52, Kevin Rose - EOS New York [@eosnewyork]: Sharif the difficulty I’m seeing is that twitter isn’t the most popular platform for EOS by a long shot and other platforms like WeChat only support org specific channels. It might just be another layer of management rather than helping.
- 29.05.2018 09:37:43, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Probably right.
- 29.05.2018 09:38:34, David P - EOS42 [@GunnisonCap]: I think the only way is the coordinated BP 'thunderclap' style announcement across multiple of their social media channels globally
- 29.05.2018 09:39:06, Michael Yeates [@michaelyeates]: network of communities
- 29.05.2018 09:42:26, Kevin Rose - EOS New York [@eosnewyork]: And I know everyone is feverishly working but I want to bring up a very important component to the launch. Even if we have multiple chains that are verified correct by many BPs many of us have agreed to one mainnet. This means we need a method of choosing from the mainnet chain candidates that we have vetted using the tools and processes we have established here. I propose we decide on as simple of a process as possible which randomly chooses a verified correct chain. From there we present it to the community for 3rd party verification and ultimately voting validation and activation.
- 29.05.2018 09:45:32, Rohan - EOSauthority [@RohanWillAnswer]: Update on the snapshot mismatch..
- There was indeed a bug, identified and flagged with mighty sandw1ch.
- He has narrowed it to a race condition and will be sending out a patch soon. I will then be able to merge it onto the EOS distribution collective Github as well.
- This was definitely a problem affected about 190 odd wallets at period 342 and caused the discrepancy on the snapshot.
- 29.05.2018 09:47:16, David P - EOS42 [@GunnisonCap]: That's interesting, be good to understand details later when you can expand
- 29.05.2018 09:49:32, Sharif Bouktila - EOS Dublin [@sharifbouktila]: From yesterday’s call we had agreed that the core teams will only launch one at a time
- 29.05.2018 09:50:33, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Obv others can do other things , but I think we need to position GhostBios as launching Mainnet
- 29.05.2018 09:51:07, Sharif Bouktila - EOS Dublin [@sharifbouktila]: I missed call today so I could be weeks behind current thinking
- 29.05.2018 09:54:40, Kevin Rose - EOS New York [@eosnewyork]: Yet the BIOS channel is still chugging along. Did they agree to join Ghostbusters? Or did Ghostbusters agree to join BIOS?
- 29.05.2018 09:55:01, Sharif Bouktila - EOS Dublin [@sharifbouktila]: It’s a merger
- 29.05.2018 09:55:21, Sharif Bouktila - EOS Dublin [@sharifbouktila]: BIOS is needed for launch
- 29.05.2018 09:55:44, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Validation will be independent
- 29.05.2018 09:56:34, Sharif Bouktila - EOS Dublin [@sharifbouktila]: [[Webpage]]https://medium.com/@eosdublin/ghostbios-de69a674d82e
- 29.05.2018 09:56:46, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Posted this this morning
- 29.05.2018 09:57:15, Kevin Rose - EOS New York [@eosnewyork]: I'm not sure what that tells me.
- 29.05.2018 09:57:32, Kevin Rose - EOS New York [@eosnewyork]: Did both groups agree to combine efforts?
- 29.05.2018 09:57:38, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Yes
- 29.05.2018 09:57:53, Kevin Rose - EOS New York [@eosnewyork]: Oh, well that's wonderful!
- 29.05.2018 09:58:18, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Unless something has changed since ...
- 29.05.2018 09:58:27, Kevin Rose - EOS New York [@eosnewyork]: Please excuse me as I clearly missed that.
- 29.05.2018 10:01:49, Sharif Bouktila - EOS Dublin [@sharifbouktila]: If that’s still the position, it should also be in the BP statements
- 29.05.2018 10:02:18, Sharif Bouktila - EOS Dublin [@sharifbouktila]: 1 approach to launch 1 mainnet
- 29.05.2018 10:02:40, Kevin Rose - EOS New York [@eosnewyork]: Absolutely
- 29.05.2018 10:04:45, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Anyone in here have a different understanding of where we are currently??
- 29.05.2018 10:06:23, Charles H - EOS 42 [@ankh2054]: I am on a eos-bios boot right now, but nobody here is running any VPN.
- .....29.05.2018 10:37:51, Jiri Jetmar | EOS Germany | [email protected] [@jirijetmar]: >>invited Peter Sveda EOS Germany [@PeterSvedaEOSGermany]<<
- 29.05.2018 10:40:07, Sharif Bouktila - EOS Dublin [@sharifbouktila]: We have 4 days to launch.
- How many dry runs can the potential ABPs commit to in those 4 days?
- Is one per day achievable?
- 29.05.2018 10:42:13, Charles H - EOS 42 [@ankh2054]: at least one per day I would hope.
- 29.05.2018 10:59:58, Sharif Bouktila - EOS Dublin [@sharifbouktila]: I'd suggest we have a schedule, Publish a list of participants who join each and capture any issues/comments/suggestions somewhere. - Do we have something like this currently??
- 29.05.2018 11:01:59, Roshan - EOS Authority [@telepound]: 1 per day atleast with the bios boot group. Just cautious about the 1st June code dump that might cause any breaking changes!
- .....29.05.2018 11:04:09, Matthew Darwin - EOS Nation [@mdarwin]: >>unsupported service message type: messageActionChatJoinedByLink<<
- 29.05.2018 11:08:08, Sharif Bouktila - EOS Dublin [@sharifbouktila]: That's a risk we are carrying. No way around it.
- 29.05.2018 11:11:14, Marc-Antoine Ross - EOS Canada [@marcantoineross]: Just caught up on the channel here. Great leadership folks! This morning we had 60 people on the call to launch stage 16 with eos-bios. We are really looking forward to working with EVERYONE on rehearsing. Maybe the schedule we proposed isn't right. We are looking for another launch at 1300UTC tomorrow but happy to change that. Winlin is also making progress on validation. The community still needs more validation!
- 29.05.2018 11:11:41, Alexandre Bourget - EOS Canada [@abourget]: Hey guys, I had a suggestion for the boot sequence.. I thought we could add an eosio.burned as a system account, and send any tokens that we don't want in circulation anymore, there..
- 29.05.2018 11:12:09, Alexandre Bourget - EOS Canada [@abourget]: Right now, we need to issue more token to pay for RAM and delegate stuff.. but we have no recall to destroy the extra..
- 29.05.2018 11:12:20, Alexandre Bourget - EOS Canada [@abourget]: we could, post-launch, destroy any overflow remaining in eosio to eosio.burned
- 29.05.2018 11:12:21, Alexandre Bourget - EOS Canada [@abourget]: what do you think ?
- 29.05.2018 11:15:46, Alexandre Bourget - EOS Canada [@abourget]: FOLKS: If we don't plan for that.. it's going to backfire with a very bad signal to the holders.. We said there would be 1B tokens.. if we issued 2B.. that dilutes the whole thing.. we need a plan to destroy the excess.
- 29.05.2018 11:16:49, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Agreed - We need to burn excess.
- 29.05.2018 11:25:01, Alexandre Bourget - EOS Canada [@abourget]: I think it should be the first operation after mainnet goes live.. that someone proposed an msig `transaction to be signed by active BPs, that transfers excess EOS from `eosio to eosio.burned ... I'll add that account creation in the boot seq.
- 29.05.2018 11:25:02, Charles H - EOS 42 [@ankh2054]: yes definitly a good plan.
- 29.05.2018 11:27:31, Jae Chung - HKEOS [@jaechung]: I think this is a good idea
- 29.05.2018 11:27:48, Jae Chung - HKEOS [@jaechung]: @abourget does this mean we should create an extra billion (11 billion instead of 10 billion)?
- 29.05.2018 11:30:01, Alexandre Bourget - EOS Canada [@abourget]: hmm.. Todd is saying this will change.. (in Jungle Testnet chan).. we'll need to sync.. darn :) thought changes were over
- 29.05.2018 11:40:29, Alexandre Bourget - EOS Canada [@abourget]: @xebb82 and friends, might want to read the conversation in the Jungle Testnet channel.. news and decisions
- 29.05.2018 11:42:33, Alexandre Bourget - EOS Canada [@abourget]: It *would* have been awesome that we cap to 1B, not a coin more at start.. but the solution to this is: no producers have any account, and users buy their own RAM..
- 29.05.2018 16:09:33, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Guys can we agree and publish a schedule for Launch Dry Runs for the next 4 days. (Wed, Thur, Fri, Sat)
- Should we have a signup sheet to see which teams are taking part each day?
- Do we have a list of BPs who are willing to take part in Launch on Sunday 3rd June?
- What time should we start on Sunday 3rd June? Suggesting 11:00 UTC How would that work for teams EAST and WEST ?
- 29.05.2018 16:28:09, David P - EOS42 [@GunnisonCap]: That’s a good idea, do you want to get a central spreadsheet up and we can build up the list. We need to ensure BPs from across Asia are fully engaged with these too and involved in the dry runs
- 29.05.2018 16:29:17, David P - EOS42 [@GunnisonCap]: 11am UTC works globally but frankly no time will be right for long, it seems as good as any.
- 29.05.2018 16:32:34, Shaheen Counts - EOS BlockSmith [@SeanPaul19]: Yes. Great initiative guys. A spreadsheet sounds ideal.
- 29.05.2018 16:38:35, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Ok will we use the same spreadsheet to collect contact details (email/phone numbers) of all potential BPs? or is there another solution to that yet?
- 29.05.2018 16:39:55, Sharif Bouktila - EOS Dublin [@sharifbouktila]: I suggest for these dry runs we keep it as close to production as possible - Keybase for comms etc..
- 29.05.2018 16:40:06, Marc-Antoine Ross - EOS Canada [@marcantoineross]: I think 1100UTC will be rough for west coast in North America. We've been running sessions at 1300UTC and it's still too early for them. Maybe 1400UTC would be a bit less worse?
- 29.05.2018 16:40:38, deck - EOS New York [@deckw]: anything stopping us from running two a day?
- 29.05.2018 16:40:55, Jiri Jetmar | EOS Germany | [email protected] [@jirijetmar]: Agree, 1400 UTC is fine..
- 29.05.2018 16:40:57, Marc-Antoine Ross - EOS Canada [@marcantoineross]: Not on our side !
- 29.05.2018 16:41:10, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Would we get production like participation twice a day?
- 29.05.2018 16:41:32, Marc-Antoine Ross - EOS Canada [@marcantoineross]: Yes on our side 🙂
- 29.05.2018 16:41:40, Jiri Jetmar | EOS Germany | [email protected] [@jirijetmar]: yes, we should. Training is now required..
- 29.05.2018 16:42:25, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Can I suggest that we keep BIOS launch group running for the ones who are training still.
- 29.05.2018 16:42:47, Sharif Bouktila - EOS Dublin [@sharifbouktila]: And have Production Dry Runs for the ones what have already participated ?
- 29.05.2018 16:44:12, Sharif Bouktila - EOS Dublin [@sharifbouktila]: i.e if you haven't done it before, take part in normal BIOS Boot , and once comfortable , "graduate" to our Production Dry Runs.
- 29.05.2018 16:45:48, Sharif Bouktila - EOS Dublin [@sharifbouktila]: I'd like the production dry runs to be the same as we do on Sunday. Same Time, Same Tools, Same Phases etc..
- 29.05.2018 17:05:44, Sharif Bouktila - EOS Dublin [@sharifbouktila]: It's getting late here. Can someone stateside get a spreadsheet together for Production Dry Runs please. Maybe @SeanPaul19 ??
- .....29.05.2018 17:06:13, Matthew Bryce 马修·布赖斯 - BitSpace [@matthewbryce]: >>invited Lars Ihler - BitSpace [@LarsIhler]<<
- 29.05.2018 18:03:03, Shaheen Counts - EOS BlockSmith [@SeanPaul19]: Yeah sure.
- 29.05.2018 20:35:02, Kyle - EOSVibes [@TheAwakenment]: 💯%
- 29.05.2018 21:49:23, Nathan Rempel | GenerEOS Sydney [@nsrempel]: {{FWD: Nathan Rempel | GenerEOS Sydney [@nsrempel], 29.05.2018 21:45:47}}
- @daniellarimer curious if you have any thoughts about bps granting themselves short names during the boot process
- 29.05.2018 21:49:23, Nathan Rempel | GenerEOS Sydney [@nsrempel]: {{FWD: Daniel Larimer [@daniellarimer], 29.05.2018 21:48:54}}
- they would have to get community approval for it
- 29.05.2018 21:49:24, Nathan Rempel | GenerEOS Sydney [@nsrempel]: {{FWD: Daniel Larimer [@daniellarimer], 29.05.2018 21:48:59}}
- I would advise against it
- 29.05.2018 21:50:06, Kevin Rose - EOS New York [@eosnewyork]: That solves that mystery.
- 29.05.2018 21:53:05, deck - EOS New York [@deckw]: so we are creating b1b1b1b1b1b1?
- 29.05.2018 21:53:12, deck - EOS New York [@deckw]: :)
- 29.05.2018 21:55:31, Igor Lins e Silva - EOS Rio [@igorls]: +1
- 29.05.2018 21:56:43, Nathan Rempel | GenerEOS Sydney [@nsrempel]: {{FWD: Nathan Rempel | GenerEOS Sydney [@nsrempel], 29.05.2018 21:54:21}}
- at the moment most boot processes seem to be making the assumption that the boot node will use eosio to create accounts for the 20 other participating producers prior to the 15% stake event
- 29.05.2018 21:56:43, Nathan Rempel | GenerEOS Sydney [@nsrempel]: {{FWD: Daniel Larimer [@daniellarimer], 29.05.2018 21:55:48}}
- if they do that then they should make it 12 char names
- 29.05.2018 21:57:13, Nathan Rempel | GenerEOS Sydney [@nsrempel]: {{FWD: Daniel Larimer [@daniellarimer], 29.05.2018 21:50:21}}
- it could be viewed as the initial producers taking advantage. This is why I wanted 10 liquid EOS so the initial producer canidates could buy "nice" 12 char names.
- 29.05.2018 21:57:14, Nathan Rempel | GenerEOS Sydney [@nsrempel]: {{FWD: Nathan Rempel | GenerEOS Sydney [@nsrempel], 29.05.2018 21:52:37}}
- Is that 10 liquid EOS in every snapshot account?
- 29.05.2018 21:57:14, Nathan Rempel | GenerEOS Sydney [@nsrempel]: {{FWD: Daniel Larimer [@daniellarimer], 29.05.2018 21:53:11}}
- assuming the account had over 11 EOS
- 29.05.2018 21:57:45, Nathan Rempel | GenerEOS Sydney [@nsrempel]: sent those backwards heh
- 29.05.2018 22:06:26, Nathan Rempel | GenerEOS Sydney [@nsrempel]: {{FWD: Daniel Larimer [@daniellarimer], 29.05.2018 22:04:55}}
- I think the initial "boostrap" producers should have "throw away accounts" that people don't vote for, everyone, including the initial producers, should have to register an account and a producer by the same method.
- In my boot process there would only be 1 initial producer to supervise the election, but the community seems to want 21 initial producers. Either way these initial producers should not even be registered to vote (and don't need to be registered in order to become initial producers).
- 29.05.2018 22:06:26, Nathan Rempel | GenerEOS Sydney [@nsrempel]: {{FWD: Daniel Larimer [@daniellarimer], 29.05.2018 22:05:30}}
- the bios contract can call "setprods" to the initial 21
- 29.05.2018 22:06:27, Nathan Rempel | GenerEOS Sydney [@nsrempel]: {{FWD: Daniel Larimer [@daniellarimer], 29.05.2018 22:05:46}}
- the eosio.system contract wouldn't even have producer records for them
- 29.05.2018 22:06:27, Nathan Rempel | GenerEOS Sydney [@nsrempel]: {{FWD: Daniel Larimer [@daniellarimer], 29.05.2018 22:05:51}}
- thus the initial prods could not be voted for
- 29.05.2018 22:16:44, Nathan Rempel | GenerEOS Sydney [@nsrempel]: {{FWD: Alexandre Bourget - EOS Canada [@abourget], 29.05.2018 22:14:47}}
- isolating the whole boot seq on a single node, makes it a *lot* faster, protects from injections, deadlines, and then swarms the blocks out..
- 29.05.2018 22:16:45, Nathan Rempel | GenerEOS Sydney [@nsrempel]: {{FWD: Daniel Larimer [@daniellarimer], 29.05.2018 22:14:58}}
- I agree with this
- 29.05.2018 22:16:45, Nathan Rempel | GenerEOS Sydney [@nsrempel]: {{FWD: Daniel Larimer [@daniellarimer], 29.05.2018 22:15:18}}
- this is the approach I advocate
- 29.05.2018 22:21:58, Craig Murray - EOS BlockSmith [@craigmurray100]: oh that would be interesting... so no 21 ABPs at all
- 29.05.2018 22:22:14, Craig Murray - EOS BlockSmith [@craigmurray100]: just 1 guy creates a node and says "ok everyone come here and regproducer and vote" from his 1 machine
- 29.05.2018 22:22:16, deck - EOS New York [@deckw]: that is interesting. Seems dicey if we think there will be an attack
- 29.05.2018 22:22:31, Craig Murray - EOS BlockSmith [@craigmurray100]: as long as that 1 node is really durable :)
- 29.05.2018 22:22:38, deck - EOS New York [@deckw]: could be done with nodes in front
- 29.05.2018 22:22:53, Craig Murray - EOS BlockSmith [@craigmurray100]: yeah a full topology is durable
- 29.05.2018 22:23:02, deck - EOS New York [@deckw]: hmm...
- 29.05.2018 22:23:07, deck - EOS New York [@deckw]: I don't hate it
- 29.05.2018 22:24:02, Nathan Rempel | GenerEOS Sydney [@nsrempel]: 1 node is easy to pause, provide validation proof, allow other connections to read the block and do simulataneous validation, then allow producers to connect and register
- 29.05.2018 22:24:30, deck - EOS New York [@deckw]: Would be really quick too.
- 29.05.2018 22:24:32, Craig Murray - EOS BlockSmith [@craigmurray100]: it makes things a lot easier
- 29.05.2018 22:25:24, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: have a cloudfront business plan ahead of it
- 29.05.2018 22:25:36, deck - EOS New York [@deckw]: './eos-bios boot --single' and put nodes in front and have at it
- 29.05.2018 22:25:45, Craig Murray - EOS BlockSmith [@craigmurray100]: ok, so let's have each BP launch a full topology and then submit it to everyone, and we can have 179 nodes, each one that every other BP has to then register on and we can tell all the wallet portals about the 179 nodes that are candidates
- 29.05.2018 22:26:01, Craig Murray - EOS BlockSmith [@craigmurray100]: <sarcasm>
- 29.05.2018 22:26:14, deck - EOS New York [@deckw]: Still waiting on that sarcasm font
- 29.05.2018 22:26:18, Craig Murray - EOS BlockSmith [@craigmurray100]: it does seem like a likely way to have multiple mainnets
- 29.05.2018 22:27:10, Nathan Rempel | GenerEOS Sydney [@nsrempel]: well, depends if everyone wants to be a boot or if we can agree on "one" or a small selection of boot nodes
- 29.05.2018 22:27:15, Nathan Rempel | GenerEOS Sydney [@nsrempel]: validate them all and choose one
- 29.05.2018 22:27:21, Craig Murray - EOS BlockSmith [@craigmurray100]: i guess we could all use eos-bios to create weights and the boot node just never reconnects
- 29.05.2018 22:27:42, Craig Murray - EOS BlockSmith [@craigmurray100]: whoever has the highest total weight we have already agreed to "support 1 mainnet"
- 29.05.2018 22:27:54, Craig Murray - EOS BlockSmith [@craigmurray100]: we use eos itself to form consensus on a single BP group to run the bios node
- 29.05.2018 22:28:07, Craig Murray - EOS BlockSmith [@craigmurray100]: we all regproducer and wait for 15%
- 29.05.2018 22:32:50, Shaheen Counts - EOS BlockSmith [@SeanPaul19]: 😂
- 29.05.2018 22:39:06, Craig Murray - EOS BlockSmith [@craigmurray100]: I'd be looking at @igorls, @xebb82 , and @abourget as possible candidates to run the 1 boot node that gets us to 15%.
- 29.05.2018 22:39:56, Nathan Rempel | GenerEOS Sydney [@nsrempel]: Yeah honestly whoever has the most experience making sure the boot and validation is accurate is what's important
- 29.05.2018 22:40:07, Nathan Rempel | GenerEOS Sydney [@nsrempel]: and that it wont go down once all the producers connect
- 29.05.2018 22:46:24, Shaheen Counts - EOS BlockSmith [@SeanPaul19]: 1 initial producer seems to solve a lot of problems.
- 29.05.2018 22:51:34, Alexandre Bourget - EOS Canada [@abourget]: that's in the contract though, b1 is hardcoded in there...
- 29.05.2018 23:00:07, Igor Lins e Silva - EOS Rio [@igorls]: That is one single change that doesn't hurt much ;) but I'm okay with b1 too
- 30.05.2018 02:35:04, Shaheen Counts - EOS BlockSmith [@SeanPaul19]: [[Webpage]]https://docs.google.com/spreadsheets/d/1wvk5eOtb8hscIWXdaVP8eED0-Ec_SKiENSX_RVgTIiA/edit?usp=sharing
- 30.05.2018 02:39:50, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Thanks @SeanPaul19
- 30.05.2018 03:29:28, Sharif Bouktila - EOS Dublin [@sharifbouktila]: I second either one of @igorls @xebb82 @abourget to be our boot node.
- I would suggest that one of them should be in each Production Dry Run.
- For go-live, we need a way to pick one. Personally I'd suggest a dance off live streamed on EOS Go , but Rock Paper Scissors could work also. (or another way)
- Questions:
- 1. Should all Launch BPs sign something - BP agreement or similar ??
- 2. Should only the ABPs be in the secure comms channel? Do we have one for everyone and then move to a smaller one with just the 21?
- 3. Can we set the criteria for go-no go on Sunday. ( Validated snapshot , BPs online and checked in etc.. )
- .....30.05.2018 03:40:22, Boseok Kang [@boseokkang]: >>unsupported service message type: messageActionChatJoinedByLink<<
- 30.05.2018 03:54:17, WinLin-EOSBIXIN [@hiwinlin]: Validation Tool Progress:
- the functions have been realized mostly, I have test in stage16.testnets.eoscanada.com and waiting to next stage check.
- Feature:
- Contracts validation
- Account privilege check
- Snapshot balance validation
- Support eos-bios BIOS and easily to DIY
- 30.05.2018 03:54:22, WinLin-EOSBIXIN [@hiwinlin]: [[Webpage]]https://github.com/EOSBIXIN/eostoolkit/tree/dev/eos_validator
- 30.05.2018 04:08:02, Eric - sw/eden [@xebb82]: I'll PM u a bit.
- 30.05.2018 04:44:06, Charles H - EOS 42 [@ankh2054]: +1 on @xebb82 @igorls @abourget as the possible boot nodes. Can I suggest we have a call today to go through this.
- .....30.05.2018 05:56:14, Tim Weston - GenerEOS (BP Sydney) [@timweston]: >>invited Ralf Weinand | GenerEOS (BP Sydney) [@RalfWei]<<
- 30.05.2018 07:13:10, WinLin-EOSBIXIN [@hiwinlin]: [[Webpage]]Validation Tool Progress:
- add theories description in Readme
- find a way to make sure NO extra system account can be created
- https://github.com/EOSBIXIN/eostoolkit/tree/master/eos_validator
- 30.05.2018 07:13:14, WinLin-EOSBIXIN [@hiwinlin]: [[Photo]]
- .....30.05.2018 07:14:52, Sharif Bouktila - EOS Dublin [@sharifbouktila]: >>unsupported service message type: messageActionPinMessage<<
- 30.05.2018 07:17:55, Sharif Bouktila - EOS Dublin [@sharifbouktila]: I have pinned the sign up sheet for the Launch Dry Runs
- I suggest that on Saturday's one we aim to launch a testnet with EOS 1.0 and leave it running for an extended period. Maybe keep the SYS token name?
- We need a GitHub repo not just for the code base, but also for everyone to log issues.
- Any suggestions?
- 30.05.2018 07:58:21, Alexandre Bourget - EOS Canada [@abourget]: There's a few folks that practiced being a boot node in the other channel. They did it successfully.. eosnation, eoslaomao and a few others.
- 30.05.2018 08:00:41, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Is 3 not enough??
- 30.05.2018 08:01:37, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Personally I don't mind.
- 30.05.2018 08:18:25, Alexandre Bourget - EOS Canada [@abourget]: Well the eos-bios process was to let the community decide (through votes on chain). I do think they will make a fair choice.
- Not sure I want to go back and say it's going to be X and Y
- 30.05.2018 08:20:38, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Ok so BPs opt in for Boot
- Any risk some will do it incorrectly? And get picked?
- 30.05.2018 08:22:39, Alexandre Bourget - EOS Canada [@abourget]: always a risk, but if they do.. we just outvote them and do it 5 minutes after..
- 30.05.2018 08:22:46, Alexandre Bourget - EOS Canada [@abourget]: so not a big risk they screw up long-term anything
- 30.05.2018 08:32:53, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Apart from optics, any concern is asking teams in advance if they are putting themselves in to be bios node?
- Can we ask for the same disco files for Friday and Sat dry runs ?
- 30.05.2018 10:51:59, Sharif Bouktila - EOS Dublin [@sharifbouktila]: I'm assuming John Doe wasn't able to participate in a Dry Run launch today alone.
- Do we think we can get a combined dry run done tomorrow ??
- @abourget @igorls @xebb82 Do we work on making the BIOS launches that we have production dry runs?? How many are taking part in BIOS there today??
- 30.05.2018 10:55:59, Marc-Antoine Ross - EOS Canada [@marcantoineross]: Quick update: over 70 people on the call, roughly 35 active BP accounts, @igorls meshed in as an ABP and producing. We are testing pausing the chain now just to see if it works.
- 30.05.2018 10:56:14, Marc-Antoine Ross - EOS Canada [@marcantoineross]: Some of us are using production style environment
- 30.05.2018 10:58:01, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Ok super, so it is easier to "upgrade" what you guys are doing and formalise them into Production Dry Runs than to try start anything else.
- 30.05.2018 11:00:05, Marc-Antoine Ross - EOS Canada [@marcantoineross]: That's a good question... There is still a lot of training for some... Maybe saying we do a production run where everyone will be aggressively DDoS'ed as we go would be the barrier to entry
- 30.05.2018 11:01:01, Marc-Antoine Ross - EOS Canada [@marcantoineross]: I'd like everybody to be able to join, try, and prove they can handle it... We will be surprised I am sure
- 30.05.2018 11:03:35, Sharif Bouktila - EOS Dublin [@sharifbouktila]: I'm sure we are all conscious of the time closing in here.
- We need as many production dry runs as possible ( using Keybase , back up comms channels etc.. )
- How many are possible??
- 30.05.2018 11:05:48, Eric - sw/eden [@xebb82]: We have had a couple of meetings with Alex and Marc and we have presented a possible solution for a compromise.
- Marc and Alex seemed positive but they have asked for more time to think about it and we are still waiting for confirmation. We have a few details we need to dig into and discuss further.
- They are currently focused on launching / testing and we have been promised a follow-up meeting before Jae gets to bed today. In the mean-time we are trying to start detailing the plan for acceptance testing. Would love help there.
- We are hoping that we asap can announce that we have found a solution to move forward asap.
- 30.05.2018 11:07:01, Marc-Antoine Ross - EOS Canada [@marcantoineross]: I have not seen the chart yet Eric. Have you shared that with me?
- 30.05.2018 11:07:35, Eric - sw/eden [@xebb82]: We had Igor send it over, right?
- 30.05.2018 11:08:18, Marc-Antoine Ross - EOS Canada [@marcantoineross]: I don't see it in: Email, Telegram, Slack, WeChat, LinkedIn, Facebook... am I missing something ? 🙂
- 30.05.2018 11:08:45, Eric - sw/eden [@xebb82]: Shipping over what I have over telegram.
- 30.05.2018 11:09:24, Marc-Antoine Ross - EOS Canada [@marcantoineross]: thanks
- 30.05.2018 11:10:20, Eric - sw/eden [@xebb82]: Anytime.
- 30.05.2018 12:02:48, aza eosar [@p17er0bl3do]: mostly agree with dan comments, shouldnt be that hard to create the initial accounts with: 1st EOS staked, next 10 EOS liquid, and the rest staked
- 30.05.2018 12:05:45, Eric - sw/eden [@xebb82]: We are creating accounts for every eth add with >=1 EOS
- You mean buy ram for 1 EOS, right?
- 30.05.2018 12:09:56, aza eosar [@p17er0bl3do]: yes right
- 30.05.2018 17:59:48, Sharif Bouktila - EOS Dublin [@sharifbouktila]: How about this suggestion for Production Launch Dry Runs
- @abourget - Thursday Dry Run
- @igorls - Friday Dry Run
- @xebb82 - Sat 13:00 UTC Dry Run
- Random among BIOS boots - Sat 16:00 Dry Run
- Sunday - 13:00 UTC - Launching ...
- 30.05.2018 18:00:31, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Named people would be BOOT node and provide feedback / guidance for the group?
- 30.05.2018 18:00:55, David P - EOS42 [@GunnisonCap]: I like it a lot
- 30.05.2018 18:02:41, Alexandre Bourget - EOS Canada [@abourget]: We're already organizing staged launches each day.. but they are always random
- 30.05.2018 18:03:01, Alexandre Bourget - EOS Canada [@abourget]: The problem with Sunday 13:00 UTC is it's not 24 hours after freeze. Monday 1300 UTC is (over 24h actually)
- 30.05.2018 18:03:18, Alexandre Bourget - EOS Canada [@abourget]: It won't hurt to have everyone check the snapshot for issues.. and concentrate on that for a day (especially during the weekend)
- 30.05.2018 18:05:41, Sharif Bouktila - EOS Dublin [@sharifbouktila]: I know that but we need to get from what you are currently running to something that is production ready.
- E.g are all the BPs you are working with on Keybase ?? Is that what we are using ? We need a way to get this into a run book.
- If you can think of a better way than planned dry runs which teams sign up to , then shout.
- 30.05.2018 18:08:00, Sharif Bouktila - EOS Dublin [@sharifbouktila]: @telepound When is the Freeze ? And when do you think we can start?
- 30.05.2018 18:08:37, Alexandre Bourget - EOS Canada [@abourget]: Well the announcements we've pinned in the EOSIO BIOS Boot channel have always gathered huge crowd.. It works fine I think.. And also, the whole community is on Telegram.. I might have preferred Keybase myself,.. but moving this late in the game isn't worth it I think.
- 30.05.2018 18:08:48, Alexandre Bourget - EOS Canada [@abourget]: I'm going to be offline for ~4h
- 30.05.2018 18:15:23, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Personally I don't care what tool we use, as long as everyone agrees to be on it and we have a back up plan (or two) in case it goes off line.
- Anyone else got some ideas on what we should do between now and Sunday?
- .....30.05.2018 21:19:11, Kyunghwan Kim - EOSeoul.io [@resvera]: >>unsupported service message type: messageActionChatJoinedByLink<<
- .....30.05.2018 21:23:04, - eoseoul.io DS Lim [@dslim]: >>unsupported service message type: messageActionChatJoinedByLink<<
- .....30.05.2018 21:50:21, Sungmin Ma - EOSeoul.io [@moomooland]: >>unsupported service message type: messageActionChatJoinedByLink<<
- 30.05.2018 23:49:48, Kevin Rose - EOS New York [@eosnewyork]: Hey everyone, the most recent joint statement is ready to be published in BP launch group.
- .....31.05.2018 00:24:49, Hubert Chang-EOSeoul.io [@hjhc00]: >>unsupported service message type: messageActionChatJoinedByLink<<
- 31.05.2018 04:14:00, Roshan - EOS Authority [@telepound]: When is the freeze: ERC20 EOS tokens are frozen at June 2, 2018 at 21:59:59 UTC. [Each period is always 23 hours, so not sure why how some got 24 hours for the freeze?]
- Snapshot start: We will probably run the final snapshot on 2nd June at 22:15 UTC. Will also put a high spec server at it. Snapshot will be published on github and we will compare it with other snapshots.
- We previously had a bug with snapshot where two runs were not the same. These bugs are fixed and should work as expected.
- cc @RohanWillAnswer
- 31.05.2018 04:17:22, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Thanks.
- So Target launch Sunday 3rd June 13:00 UTC looks right to me.
- 31.05.2018 04:22:03, Roshan - EOS Authority [@telepound]: yes, that looks reasonable. however, i have never ran a full snapshot import with user balance & stakes etc. I hope to do that with @abourget & bios BPs group members at some point in the next few days. we are also running a full snapshot right now (#348) with the new snapshot code.
- 31.05.2018 04:38:56, David P - EOS42 [@GunnisonCap]: Be interested in an update on how the full snapshot goes Roshan, thanks
- 31.05.2018 04:40:27, Roshan - EOS Authority [@telepound]: sure will get back to this group
- 31.05.2018 04:40:49, David P - EOS42 [@GunnisonCap]: Great work on the ongoing snapshot analysis by the way gents
- 31.05.2018 04:42:22, Roshan - EOS Authority [@telepound]: thanks, we will also be testing yesterday's patch to confirm it works.
- 31.05.2018 05:08:05, Eric - sw/eden [@xebb82]: I would love to do that but saturday will be "heavy feature testing day" for me.
- We will have to spin up a lot of chains per day until then, and really start focusing on trying to test as much as possible as quickly as possible :D
- 31.05.2018 08:17:43, Alexandre Bourget - EOS Canada [@abourget]: Guys I wrote a small doc for those who don't want to use the eos-bios software, yet want to participate in ABPs and launch with the community around eos-bios:
- https://github.com/eoscanada/eos-bios/blob/master/README.nobios.md
- 31.05.2018 10:45:42, Rohan - EOSauthority [@RohanWillAnswer]: Hello, did a snapshot to period 348 earlier today. It took about 2 hours with —resume flag (faster and recommended now). That’s probably the fastest. Waiting for the others to run 348 to compare and raise any issues.
- 31.05.2018 10:46:12, deck - EOS New York [@deckw]: [[Webpage]]348 is up here for us. https://github.com/eosnewyork/snapshots
- 31.05.2018 10:46:42, Rohan - EOSauthority [@RohanWillAnswer]: The genesis repository will stop being updated and will also be deleted today @sandw1ch .. The distribution collective repo is updated in case we need it..
- 31.05.2018 10:49:27, Rohan - EOSauthority [@RohanWillAnswer]: Did you run on the latest master branch of the genesis code?
- 31.05.2018 10:50:01, deck - EOS New York [@deckw]: should be yes
- 31.05.2018 10:52:37, Rohan - EOSauthority [@RohanWillAnswer]: Does not look like it, you still seem to be having the “$10m hack issue” unfixed EOS6MRyAjQq8ud7hVNYcfnVPJqcVpscN5So8BhtHuGYqET5GDW5CV that key was in the burn list.. Your snapshot.csv seems to contain it..
- 31.05.2018 10:53:02, Rohan - EOSauthority [@RohanWillAnswer]: The good news is we have very little difference
- 31.05.2018 10:53:02, deck - EOS New York [@deckw]: is the distribution collective not up-to-date?
- 31.05.2018 10:53:18, Rohan - EOSauthority [@RohanWillAnswer]: When did you run this?
- 31.05.2018 10:53:22, deck - EOS New York [@deckw]: this morning
- 31.05.2018 10:53:41, Rohan - EOSauthority [@RohanWillAnswer]: I suspect you pulled before the final commits from @sandw1ch..
- 31.05.2018 10:53:42, deck - EOS New York [@deckw]: 2ish hours ago
- 31.05.2018 10:53:55, Rohan - EOSauthority [@RohanWillAnswer]: Maybe you just missed it? Can you try pulling again?
- 31.05.2018 10:53:58, deck - EOS New York [@deckw]: hmm ok will pull and run again
- 31.05.2018 10:54:25, Rohan - EOSauthority [@RohanWillAnswer]: The good news is the amounts match perfect! Just the keys disagree..
- 31.05.2018 10:58:55, Rohan - EOSauthority [@RohanWillAnswer]: Did you use —resume?
- 31.05.2018 11:01:30, deck - EOS New York [@deckw]: node snapshot —load_config —poll —period=350 —resume —verbose_mt —recalculate-wallets
- 31.05.2018 11:04:55, deck - EOS New York [@deckw]: I am now running with the version from the eosio repo instead of the collective.
- .....31.05.2018 11:50:31, Rohan - EOSauthority [@RohanWillAnswer]: >>invited Sandwich [@sandw1ch]<<
- 31.05.2018 11:52:10, Sandwich [@sandw1ch]: They are the same except a change in a readme file.
- 31.05.2018 11:54:27, deck - EOS New York [@deckw]: Yup as you explained in the other channel I was running with the wrong parameters.
- 31.05.2018 11:54:44, Sandwich [@sandw1ch]: Gotchya 👍
- 31.05.2018 11:54:51, Sandwich [@sandw1ch]: Just placing it here for later reference.
- 31.05.2018 11:55:00, Sandwich [@sandw1ch]: If anyone has questions, today is the day for them.
- 31.05.2018 11:55:15, deck - EOS New York [@deckw]: ok appreciate all your hard work
- 31.05.2018 11:55:15, Sandwich [@sandw1ch]: We can arrange a video call if you find that valuable.
- 31.05.2018 11:57:38, deck - EOS New York [@deckw]: I think that call would consist of RTFM honestly. I should have read the README first :)
- 31.05.2018 12:14:21, Rohan - EOSauthority [@RohanWillAnswer]: [[Webpage]]This is looking all good right? https://github.com/EOS-distribution-collective/genesis
- 31.05.2018 12:28:25, Sandwich [@sandw1ch]: Yes
- 31.05.2018 12:33:44, Sandwich [@sandw1ch]: I found something that could be adjusted in docs
- 31.05.2018 17:58:58, Eric - sw/eden [@xebb82]: Hello everyone,
- We have been working on the steps for launching and validating the main net.
- This is the flow we believe is necessary to test critical functionality and allow token holders and opportunity to validate the chain before opening it up for voting.
- Please have a look at it.
- I'll be going to bed really soon - but let's pick it up tomorrow and see if we can get some agreement.
- 31.05.2018 17:59:11, Eric - sw/eden [@xebb82]: [[Document, size 188'738 bytes]]
- 31.05.2018 18:36:42, Sharif Bouktila - EOS Dublin [@sharifbouktila]: [[Photo]]
- 31.05.2018 18:37:10, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Just on Jungle chat , what happens when we get a bug??
- 31.05.2018 19:10:47, Roman CryptoLions [@romanskaskiw]: Also this advice:
- 31.05.2018 19:10:50, Roman CryptoLions [@romanskaskiw]: [[Photo]]
- 01.06.2018 04:01:46, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Everyone - Can you please take a walk through the validation process flow posted by @xebb82
- Each block producer should be able to state to the EOS community what is the minimum scope of tests that they need to see successfully passed to open the chain up for voting.
- We have issued statements that the Mainnet Launch Team will announce when the EOS community can begin voting.
- We need to agree how we do this.
- Massive amounts of work and professionalism has got us to today. I believe one more solid push will carry us over the line.
- 01.06.2018 04:06:02, Adrian - SaltBlock [@adrian_saltblock]: +1
- 01.06.2018 04:56:20, David P - EOS42 [@GunnisonCap]: Fully agree with this, we cannot have for example a chain being fast created by one group that completes no proper validation testing being presented as the mainnet. So we need to ensure there is collective understanding and consensus on this key point. Thanks Sharif how do you propose collating this?
- 01.06.2018 05:00:50, Eric - sw/eden [@xebb82]: We need to make sure that for us to be able to declare a chain a possible main net candidate we need to run validation. +1 on this.,
- 01.06.2018 05:02:34, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: Our snapshot tools cant even reach consensus...
- 01.06.2018 05:02:40, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: Up to 80 differences
- 01.06.2018 05:02:58, Eric - sw/eden [@xebb82]: Nice :)
- 01.06.2018 05:03:44, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: Not nice at all 😞
- 01.06.2018 05:04:26, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: Hard to reach consensus on even a snapshot
- 01.06.2018 05:04:50, Nathan Rempel | GenerEOS Sydney [@nsrempel]: Why are the tools generating different results? Any ideas?
- 01.06.2018 05:05:33, David P - EOS42 [@GunnisonCap]: The point is that nobody gets to launch a fast chain, skip validation testing and attempt to call that the mainnet. Not least that will not have global BP consensus required, so would cause a big problem. @abourget do you fully agree with this point?
- 01.06.2018 05:07:35, Nathan Rempel | GenerEOS Sydney [@nsrempel]: Even if the tools don't reach concensus ... Can they atleast say 'yes it's valid'
- 01.06.2018 05:07:52, David P - EOS42 [@GunnisonCap]: Presenting a chain as the mainnet for voting and activation on the global exchanges that had fundamental functionality issues would be a huge mistake, not least us not fulfilling our duties as BPC’s at this crucial time. It would therefore be useful to ensure everybody here confirms they agree we need to complete robust validation testing - and probably what that entails.
- 01.06.2018 05:08:08, David P - EOS42 [@GunnisonCap]: A good question
- 01.06.2018 05:08:30, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: No there is no way to say something is valid
- 01.06.2018 05:08:37, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: Validation is literally a diff of all snapshots
- 01.06.2018 05:08:59, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: Ultimate validation would be all snapshot containing 0 differences
- 01.06.2018 05:10:04, Nathan Rempel | GenerEOS Sydney [@nsrempel]: Agreed but why is that so hard right now? Is it cause everyone is copying the snapshot at different times? Cause it's not locked yet? Or are you saying even if everyone uses the same snapshot and same chain.. they get different results
- 01.06.2018 05:11:15, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: Parity and snapshots work in mysterious ways
- 01.06.2018 05:11:36, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: The chain we are talking about is ethereum
- 01.06.2018 05:11:47, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: People running snapshots are getting different results
- 01.06.2018 05:13:23, Nathan Rempel | GenerEOS Sydney [@nsrempel]: If one person takes a single snapshot off ethereum and compares it to an EOS chain they load.. how does that diff compare?
- 01.06.2018 05:13:43, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: we are not at EOS loading step
- 01.06.2018 05:13:52, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: you need consensus on which snapshot to use in the first place
- 01.06.2018 05:14:10, Nathan Rempel | GenerEOS Sydney [@nsrempel]: I see
- 01.06.2018 05:15:51, Nathan Rempel | GenerEOS Sydney [@nsrempel]: Once the erc-20 contract is frozen perhaps those differences will stop? Or does the parity mechanism extract the snapshot differently every time based on blockhash and other random variables. Forgive me if these are dumb questions.. j haven't had a chance to work with the parity/Genesis code yet :(
- 01.06.2018 05:17:08, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: Not my place to answer, perhaps @RohanWillAnswer or @igorls ?
- 01.06.2018 05:19:20, Rohan - EOSauthority [@RohanWillAnswer]: We had some code updates yesterday which will hopefully make it all match correctly.. @igorls has already run the latest snapshot for period #349, I will be running it as well and we will let you know.. Fingers crossed.
- 01.06.2018 05:24:53, Sharif Bouktila - EOS Dublin [@sharifbouktila]: How about this for some targets.
- Today - 14:00 UTC - Ghostbusters Test Launch ( This should be done as close as possible to production launch as possible)
- Today - 18:00 UTC - Validation Call to agree minimum testing which needs to be completed for MAINNET Launch Group to announce chain is open for voting
- Saturday 13:00 UTC - Launch EOS 1.0 Testnet with a commitment to keep it alive for testing for 30 days.
- My view is that all system / technical testing should happen on the 1.0 Testnet (Pre-Production) and it gives us a test bed to implement and test bug fixes before trying them in Live.
- We need an agreed way of working. The chain will get launched , I've no doubt but without solid processes for testing and validation every bug after will be a bloody nightmare to get implemented.
- 01.06.2018 05:32:27, Nathan Rempel | GenerEOS Sydney [@nsrempel]: Sounds great
- 01.06.2018 05:38:29, David P - EOS42 [@GunnisonCap]: We are pushing it 14:00 today for the testnet launch, can you find out if that needs to be pushed back an hour or two. Otherwise yes we absolutely need a call to agree minimum acceptable testing to be completed for the mainnet. Anything less is not a valid (or validated) chain.
- 01.06.2018 05:39:57, Sharif Bouktila - EOS Dublin [@sharifbouktila]: The Run today is already planned by the Ghostbusters team as a practice launch. I think as many BPs as possible should take part.
- 01.06.2018 05:40:04, Pete BlockMatrix [@blockmatrixpete]: Sounds like a good plan, more than happy to commit to everything mentioned here
- 01.06.2018 05:41:54, David P - EOS42 [@GunnisonCap]: I know, but the call I was on at 2am I got the impression that time may need to adjust slightly so worth checking is all I’m saying.
- 01.06.2018 05:42:22, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Ah ok - got it
- 01.06.2018 05:42:32, David P - EOS42 [@GunnisonCap]: Great, to achieve our collective consensus we need more of this
- 01.06.2018 05:42:56, Pete BlockMatrix [@blockmatrixpete]: 👍
- 01.06.2018 05:44:13, Sharif Bouktila - EOS Dublin [@sharifbouktila]: + 1 now is not the time to stay quiet people.
- There are millions of peoples money and reputations on the line here.
- It is a Block Producers duty to protect the interests of Token Holders before, during and after launch
- 01.06.2018 07:23:18, Alexandre Bourget - EOS Canada [@abourget]: Of course.
- 01.06.2018 07:24:11, David P - EOS42 [@GunnisonCap]: Thank you, then we are all in agreement and should have no risk of an unvalidated chain being proposed as a mainnet candidate.
- 01.06.2018 07:24:49, David P - EOS42 [@GunnisonCap]: Next then is confirming exactly what the validation tests required are for all chains booted to pass.
- 01.06.2018 07:58:33, deck - EOS New York [@deckw]: [[Webpage]]Hey everyone I wrote up a python validation script using a 'semi-complete' python implementation of cleos. I am going to attempt to validate against both the ghostbuster testnet and eos-bios launches today. Anyone who has time please take a look or try it out for yourself. Cheers.
- https://github.com/eosnewyork/eospy
- 01.06.2018 08:26:10, Alexandre Bourget - EOS Canada [@abourget]: Folks about unregd:
- * The unregistered_snapshot.csv now contains an account name.
- * The previous proposition was to put all funds in eosio.unregd and then later create accounts and transfer tokens out of that account.
- * If there is an account name now, so:
- * We could create that account, just like the registered, transfer their funds on that account, and delegate its authority to eosio.unregd. In the future, that contract could do an inline transaction to set a new updateauth public key on that account, for the benefit of the person proving they own keys.
- What do you think?
- 01.06.2018 08:34:46, Adrian - SaltBlock [@adrian_saltblock]: Sounds like an extra step to validate, at a time where there's enough on the plate. Given the end result is the same, my question would be: is this really necessary right now?
- 01.06.2018 08:45:42, aza eosar [@p17er0bl3do]: +1
- 01.06.2018 08:50:47, WinLin-EOSBIXIN [@hiwinlin]: +1
- 01.06.2018 08:53:01, WinLin-EOSBIXIN [@hiwinlin]: even we record the unregister account in the block chain, the BPs may not want to support the recovery. In other side, if the BPs want do the recovery, they can get the unregistered account in the ETH chian.
- So let keep it simple.
- 01.06.2018 08:53:55, WinLin-EOSBIXIN [@hiwinlin]: and make sure the mainnet is clean with the necessities
- 01.06.2018 09:00:47, Ralf Weinand | GenerEOS (BP Sydney) [@RalfWei]: However if we do it at genesis, then we have a clean start and avoid inconsistencies / disputes later. We should validate this at the beginning IMO.
- 01.06.2018 10:12:08, mini - EOS Argentina [@moonxor]: [[Webpage]]hey, we just push our validator with some of the scripts we used for our testnet aurora.
- https://github.com/EOSArgentina/firestarter
- The boot and validation (85k accounts) take 15 minutes aprox.
- 01.06.2018 10:12:18, Marc-Antoine Ross - EOS Canada [@marcantoineross]: 👏
- .....01.06.2018 10:39:55, Jiri Jetmar | EOS Germany | [email protected] [@jirijetmar]: >>invited Niclas Hedhman - EOS Germany [@niclash]<<
- .....01.06.2018 19:18:11, alex k - EOS Brazil [@bakudd]: >>unsupported service message type: messageActionChatJoinedByLink<<
- 01.06.2018 19:32:41, deck - EOS New York [@deckw]: Hey everyone I appear to be missing somethign simple when I'm trying to parse the cleos get code <account> does anyone know how that hash is generated?
- 01.06.2018 19:48:44, Alexandre Bourget - EOS Canada [@abourget]: Sha256 of content
- 01.06.2018 19:50:24, deck - EOS New York [@deckw]: what content? wasm? wast?
- 01.06.2018 20:07:50, Alexandre Bourget - EOS Canada [@abourget]: Wasm always
- 01.06.2018 20:08:31, deck - EOS New York [@deckw]: ok that is what I thought.
- 02.06.2018 02:43:15, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Distribution is closed.
- EOS 1.0 is released.
- Take a moment to think about how best we should spend our time between now and when the millions of people who are trusting us to get this done, start using our platform.
- Focus on what is important and impactful.
- Check your egos at the desk, you can pick them up on the other side.
- 02.06.2018 02:44:24, jem [@jemxpat]: >Check your egos at the desk, you can pick them up on the other side.
- egggsactly!
- 02.06.2018 02:54:17, Eric - sw/eden [@xebb82]: 👍
- 02.06.2018 03:46:10, David P - EOS42 [@GunnisonCap]: Somebody pin this 👆
- 02.06.2018 03:47:43, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: 100%
- 02.06.2018 03:47:49, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: The first thing we need
- 02.06.2018 03:47:55, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: Is consensus on validation
- 02.06.2018 03:48:03, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: Lets test the different external validation tools
- 02.06.2018 03:48:25, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: To make sure there is some sort of consensus, otherwise there could be 3 chains, and we wouldnt even agree on which one is valid
- 02.06.2018 03:49:05, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: List all the external validation tools we have so far, I know about EOSBIXIN, what else is there?
- 02.06.2018 03:49:22, David P - EOS42 [@GunnisonCap]: That is indeed the priority, because consensus on validation is the first step to enabling section
- 02.06.2018 03:49:35, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: And its a way to stop differences
- 02.06.2018 03:49:40, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: If we have consensus on validation
- 02.06.2018 03:49:43, Jae Chung - HKEOS [@jaechung]: Igor’s tool
- 02.06.2018 03:50:04, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: Then the first chain to launch with 15% of votes and fully verified by that consensus can be mainnet
- 02.06.2018 03:50:07, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: There is no need to argue
- 02.06.2018 03:50:09, Michael Yeates [@michaelyeates]: ram pricing makes validation difficult
- 02.06.2018 03:50:57, Michael Yeates [@michaelyeates]: we need to work out a way to record how much was paid for ram at the time the account was created
- 02.06.2018 03:51:50, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: Okay, any technical solutions to this?
- 02.06.2018 03:52:16, Michael Yeates [@michaelyeates]: im thinking of a couple
- 02.06.2018 03:53:50, jem [@jemxpat]: I have privately asked Dan for an immediate update of the account creation RAM purchase.
- At bios, we have a bunch of account to create. As each account requires some RAM to hold the account data, some RAM purchase must be done to allocate that RAM. With the current RAM pricing mechanism, the first user account created cost very little to allocate 8k. As we get more and more accounts, the price continuously increases, so that accounts created later cost massively more than those created earlier.
- The variable fee makes third party account validation nearly impossible. External client queries for the account balance will show the balance minus the RAM fees, which is very, very hard to reconcile with the snapshot data, since the external probe has no ability to know the order of registration of the accounts and how the Bancor price has changed. Fixed or zero fee is the only way, it seems.
- We suggest to B1 a code change which will make initial account creation RAM fees at bios free.
- 02.06.2018 04:30:14, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: This would be a much needed fix
- 02.06.2018 04:30:28, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: Then we can finally do proper third party account validation
- 02.06.2018 04:32:07, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Is it possible for someone in the community to make this change?
- 02.06.2018 04:34:32, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: If we all have valid third party tools that multiple people can run just like the snapshot tool, then it does not matter how the chain is launched. The chain must reach the 15% threshold and have multiple BPs run the external validation tools and publically announce their results for it to be a valid chain.
- 02.06.2018 04:36:12, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: So far we have validation tools from Igor, EOSBIXIN, and EOS Argentina, that can hopefully verify deterministically after the RAM costs changes occur
- .....02.06.2018 04:38:44, Kyunghwan Kim - EOSeoul.io [@resvera]: >>invited Hyunwoong JI - EOSeoul.io [@EOSeoul_hyunwoong]<<
- 02.06.2018 04:46:56, Sungmin Ma - EOSeoul.io [@moomooland]: I personally think Jem's opinion is valid.
- Eosio thinks that Ram costs to create account should be borne by eosio, and the expense should be covered by EOSio at 4% which is left as work proposal.
- And it will be a hurdle to charge new users to join the EOS ecosystem in the future, and it will not help EOS in the long run.
- For a better direction, the 8kb used to create an account at EOS suggests to process the charge for free.
- 02.06.2018 04:49:11, Sharif Bouktila - EOS Dublin [@sharifbouktila]: That’s right.
- It is up to each BP to be satisfied that the chain is validated before declaring it open for staking.
- 100% agree we need to come at this from the validating view point.
- 02.06.2018 04:49:50, jem [@jemxpat]: Yes, but at this point I hope we can minimize the deviation from EOSIO. Any early changes will make ongoing synchronization slightly more difficult. Best if B1 will simply do the changes. There are multiple architectural options to accomplish it, so I would hate to have us implement it, then B1 does it later using a different approach.
- 02.06.2018 04:53:46, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Agreed.
- Could we have a pull request ready ? B.1 might accept it into their repo?
- We are open source now.
- 02.06.2018 04:53:55, jem [@jemxpat]: 나성민씨, I agree. I will ask for the change initally in the bios, then later, when we know how to create new accounts, we can use the same cost basis method.
- 02.06.2018 06:01:56, Alexandre Bourget - EOS Canada [@abourget]: The way I handle this right now is that eosio buys RAM for each ERC-20 account.. at market price. For that we need to issue a bit more EOS, and we can all agree to destroy the excess by sending it to eosio.burned (an account created for that purpose) after launch.
- 02.06.2018 06:02:20, Alexandre Bourget - EOS Canada [@abourget]: This way we provide 8kb RAM to everyone.
- 02.06.2018 06:03:32, Michael Yeates [@michaelyeates]: i dont think we need to issue or burn, the unregistered tokens which were sent to contract addresses are highly unlikely to be reliably claimed and if they are then we have the option to issue the extra tokens needed at the time
- 02.06.2018 06:04:25, Alexandre Bourget - EOS Canada [@abourget]: We have to put unreg'd money aside right now, and have that contract hold it. One or the other.. it doesn't change much in the end.
- 02.06.2018 06:04:45, Alexandre Bourget - EOS Canada [@abourget]: We can inspect this morning's chain and tweak that before mainnet, still a few hours.
- 02.06.2018 06:04:55, Michael Yeates [@michaelyeates]: yes, the point is that it is not good for public relations
- 02.06.2018 06:05:09, Michael Yeates [@michaelyeates]: people will appear to lose tokens on the swap
- 02.06.2018 06:06:34, Charles H - EOS 42 [@ankh2054]: People won’t agree with their current balances if we allow public validation.
- 02.06.2018 06:07:33, Alexandre Bourget - EOS Canada [@abourget]: If eosio buys RAM, then the hodler's accounts are untouched.
- 02.06.2018 06:07:39, Alexandre Bourget - EOS Canada [@abourget]: If we have hodlers pay for RAM, then what you say is true: their account won't match the new chain.
- 02.06.2018 06:08:04, Charles H - EOS 42 [@ankh2054]: Aha I see :)
- 02.06.2018 06:08:52, Alexandre Bourget - EOS Canada [@abourget]: I would have hoped for that collaboration weeks ago..
- 02.06.2018 06:09:02, Charles H - EOS 42 [@ankh2054]: I think it’s best we avoid having users pay as those created last will also pay more.
- 02.06.2018 06:09:21, Alexandre Bourget - EOS Canada [@abourget]: yes, also true, unless we buy RAM with a fixed amount of EOS.. but then not everyone gets the same amount of RAM. Those inserted last get less RAM
- 02.06.2018 06:10:33, Alexandre Bourget - EOS Canada [@abourget]: A full injection of the chain, with the algorithm stated above, "invests" 19727.0815 EOS into buying RAM for its new users.
- .....02.06.2018 06:10:51, Alexandre Bourget - EOS Canada [@abourget]: >>invited Daniel Larimer [@daniellarimer]<<
- 02.06.2018 06:11:03, Alexandre Bourget - EOS Canada [@abourget]: @daniellarimer ^^ that risks being on chain.
- 02.06.2018 06:12:49, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Let’s work from now forward.
- 02.06.2018 06:13:19, Charles H - EOS 42 [@ankh2054]: Can the EOS come from all the EOS sitting in the wrong contracts?
- 02.06.2018 06:14:08, Sharif Bouktila - EOS Dublin [@sharifbouktila]: No - this needs to be audit-able
- 02.06.2018 06:15:40, Charles H - EOS 42 [@ankh2054]: Deduct from BP’s first month payouts? Worker proposal?
- 02.06.2018 06:18:29, jem [@jemxpat]: Cleanest solution is to just set new user account RAM cost to 0. But, I really like Alex's method!
- 02.06.2018 06:19:46, jem [@jemxpat]: With a 0 cost for new accounts, my only concern is that the allocated RAM might cause some unexpected interaction with the Bancor algorithm.
- 02.06.2018 06:26:26, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Any reason we can't used Alex's solution for this?
- On the call last week when it was explained, nobody saw a reason it wouldn't work.
- 02.06.2018 06:27:55, Alexandre Bourget - EOS Canada [@abourget]: We have a rehearsal this morning 1300 UTC. Let's all try to validate the output.
- We can do the exact same thing tomorrow.
- I will whitelist contracts and freeze the chain as requested by the community.
- 02.06.2018 06:29:07, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Are we going to keep today's chain alive as a testnet for an extended period ??
- 02.06.2018 06:29:41, Michael Yeates [@michaelyeates]: the ghost account is the one that freezes the chain
- 02.06.2018 06:29:49, Alexandre Bourget - EOS Canada [@abourget]: I don't think so.. unless people ask for it...
- 02.06.2018 06:30:38, Sharif Bouktila - EOS Dublin [@sharifbouktila]: I'm asking, do you know think its needed? A solid test net where we can test any bug fixes etc..
- 02.06.2018 06:31:00, Alexandre Bourget - EOS Canada [@abourget]: Freezing can be done with -x and selectively with --contract-whitelist.
- This morning, I'll start with -x on dissemination (start paused, so not producing).
- Then slowly open up some contracts through contract-whitelist..
- 02.06.2018 06:31:14, Alexandre Bourget - EOS Canada [@abourget]: and I suggest we open to all contract only when the community is ready and says "go mainnet"
- 02.06.2018 06:31:44, Alexandre Bourget - EOS Canada [@abourget]: well you can always replicate that chain on your machine with a single command.. so I'm not sure we need a testnet, do we ?
- 02.06.2018 06:32:15, Sharif Bouktila - EOS Dublin [@sharifbouktila]: I thinking of a process where agreement is easier to implement when we have issues.
- 02.06.2018 06:32:31, Eric - sw/eden [@xebb82]: I'm fine with that solution Alex.
- We could also pay for it from the unregistered funds, and then repay that part as a worker proposal from the 4% inflation.
- 02.06.2018 06:33:26, Alexandre Bourget - EOS Canada [@abourget]: Yeah.. any manipulations of the issuance can be agreed by the 21 elected.. they can propose transactions that move things around.. so I'm not so concerned about a few balances not 100% right .. provided its just a "transfer" after
- 02.06.2018 06:34:19, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Agreement of 21 elected BPs is going to be a challenge for a while.
- 02.06.2018 06:34:26, Alexandre Bourget - EOS Canada [@abourget]: We want snapshot absolutely right, but eosio.burned and eosio balances can fluctuate a bit
- 02.06.2018 06:34:51, Alexandre Bourget - EOS Canada [@abourget]: But those funds can't move anyway.. it's just a question of perception of the maximum supply and perceived dilution.
- 02.06.2018 06:35:14, Sharif Bouktila - EOS Dublin [@sharifbouktila]: I've no issue with the approach re: funding
- 02.06.2018 06:40:30, Kyunghwan Kim - EOSeoul.io [@resvera]: RAM balance of bancor is initialized once by int64_t(_gstate.free_ram()) and changed via setram action when RAM capacity is increased. If with a 0 cost for new accounts and used bytes of RAM for new accounts are removed from RAM balance(base.balance.amount) correctly, there is no side effect. Price of next 'normal' RAM buy or sell will just increase as expected(Bancor way).
- 02.06.2018 06:43:02, Alexandre Bourget - EOS Canada [@abourget]: do you mean from the eosio.ram account ?
- 02.06.2018 06:43:37, Alexandre Bourget - EOS Canada [@abourget]: so moving the 19k EOS from eosio.ram to eosio.burned for example, would make price start again ?
- 02.06.2018 06:43:39, Alexandre Bourget - EOS Canada [@abourget]: Let me try that.
- 02.06.2018 06:48:26, Kyunghwan Kim - EOSeoul.io [@resvera]: I mean rammarket initialization on system_contract constructor in eosio.system.cpp. It’s fundamental bancor system, not balance of eosio.ram account.
- 02.06.2018 06:48:48, Alexandre Bourget - EOS Canada [@abourget]: Right, so no real way to "reset" that, or is there ?
- 02.06.2018 06:50:39, Alexandre Bourget - EOS Canada [@abourget]: no, from what I can read.. that's set in stone.. so we'll handle the excess post-launch
- 02.06.2018 06:54:37, Alexandre Bourget - EOS Canada [@abourget]: SOMETHING WE NEED:
- we need someone to take the ricardian contracts and stuff them into the .abi files from the eos-mainnet and publish that.
- Who's going to jump on that ?
- 02.06.2018 07:08:56, Kyunghwan Kim - EOSeoul.io [@resvera]: Price of RAM is only determined by bancor balance of RAM. So to implement “with 0 cost and 0 fee” RAM for newaccount, we need another buyram and sellram action just for newaccount. Maybe ‘allocNewAccountRAM’? Its argument should be with not EOS but with required RAM amount. If RAM used by newaccount creation is regarded like ‘total_ram_bytes_reserved’ and is removed from bancor RAM balance, it seems okay without paying EOS.
- 02.06.2018 07:20:32, Kevin Rose - EOS New York [@eosnewyork]: [[Webpage]]I’ve missed a lot of messages but here’s one Buddy made. https://github.com/eosnewyork/eospy
- 02.06.2018 07:22:35, deck - EOS New York [@deckw]: This makes an assumption about rambuying that has changed since made. I'm updating now.
- 02.06.2018 07:23:07, David P - EOS42 [@GunnisonCap]: There is no scope for 0 cost 0 fee allocNewAccountRAM being abused in the future is there? Presumably this is a restricted function proposal just for the mainnet boot?
- 02.06.2018 07:27:04, jem [@jemxpat]: I think it could be used to onboard unregistered token holders post launch, if we want.
- 02.06.2018 07:28:16, Kevin Rose - EOS New York [@eosnewyork]: Thanks Buddy.
- 02.06.2018 07:28:16, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: Okay we have @igorls , @deckw , @moonxor and EOSBIXIN’s tools
- 02.06.2018 07:28:25, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: thats four external validators
- 02.06.2018 07:28:29, Kevin Rose - EOS New York [@eosnewyork]: @syed_jafri
- 02.06.2018 07:28:59, David P - EOS42 [@GunnisonCap]: That’s fair, just want to ensure we do not open a back door to potential RAM abuse
- 02.06.2018 07:34:30, Kyunghwan Kim - EOSeoul.io [@resvera]: I don’t have specific position on that func. If consensus made, eosio.system can be updated as intended.
- 02.06.2018 07:40:10, Sharif Bouktila - EOS Dublin [@sharifbouktila]: [[Webpage]]We've started a Document with the features which have been released in 1.0 which we believe need validation testing before we can propose a chain for mainnet
- https://docs.google.com/document/d/1Bsi1__u7U4Jvl9JAomdtWu7U0DaWGPcmD9FIAzL_pQM/edit#heading=h.qtvfems4zceg
- 02.06.2018 07:41:21, Sharif Bouktila - EOS Dublin [@sharifbouktila]: The Main one I see is White/Black Listing of smart contracts. We think this is essential to stop a contract if it had issues.
- This was a principle promise which EOS made to the community.
- 02.06.2018 07:54:42, Alexandre Bourget - EOS Canada [@abourget]: I'm going to try that today in the launch, start with production paused. Seed the other nodes. THen restart with contract-whitelist one by one..
- 02.06.2018 07:55:23, David P - EOS42 [@GunnisonCap]: Excellent work Sharif, we’ll take a look asap
- 02.06.2018 07:56:43, Michael Yeates [@michaelyeates]: this method would need 21/21 approval. in practice it would be easier to null the code
- 02.06.2018 07:57:10, Alexandre Bourget - EOS Canada [@abourget]: During ABP period, it's only the ABPs... so potentially a single person.
- 02.06.2018 07:57:43, Sharif Bouktila - EOS Dublin [@sharifbouktila]: How many BPs would need to agree to that ? After go-live I'm talking. Say in 60 days when a dAPP goes rouge.
- 02.06.2018 07:58:32, Michael Yeates [@michaelyeates]: 15/21
- 02.06.2018 07:59:07, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Should be nice and easy then, considering how well we are all working together. 🙂
- 02.06.2018 07:59:54, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Also , we don't believe there is any way for an Arbitrator to be called , so we need an offline interface established.
- 02.06.2018 08:01:52, Alexandre Bourget - EOS Canada [@abourget]: The whitelisting doesn't send a signal to the network.. it merely ignores transactions from the blocks produced locally.. so 15/21 doesn't cut it.. You would need all producers to sync on that
- 02.06.2018 08:33:43, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Does anyone else think this is a problem?
- After Mainnet we will have BPs (which probably haven't been involved in the launch at all) and we haven't yet agreed how they should contact each other.
- Let alone make a decision on when and how to stop a contract.
- Regardless of how this weekend goes, we need to think about how things get done after launch.
- 02.06.2018 08:45:38, Michael Yeates [@michaelyeates]: has anyone fun a full validation on the token balances from yesterdays ghostbusters launch?
- 02.06.2018 08:54:49, deck - EOS New York [@deckw]: I have not. Doing local tests I went through all the accounts in ~10 minutes. Over the network it slows dramatically.
- 02.06.2018 08:54:57, deck - EOS New York [@deckw]: Working on speed ups now
- 02.06.2018 09:18:05, Eric - sw/eden [@xebb82]: If you sync a node you should do it quickly, right?
- 02.06.2018 09:18:24, deck - EOS New York [@deckw]: Yes sir.
- 02.06.2018 09:18:34, deck - EOS New York [@deckw]: give me the endpoints!!!
- 02.06.2018 09:20:12, Eric - sw/eden [@xebb82]: @michaelyeates u have one open, right?
- 02.06.2018 09:24:23, deck - EOS New York [@deckw]: Tried connecting to this address that I was given: 54.229.241.156:19876 and tried the snapshot @michaelyeates gave me but had issues with the database:
- `Dynamic exception type: boost::exception_detail::clone_impl<boost::exception_detail::error_info_injector<std::runtime_error> >
- std::exception::what: database created by a different compiler, build, boost version, or operating system`
- I didn't want to pester so I started working on my end
- 02.06.2018 09:25:18, Eric - sw/eden [@xebb82]: oh, what was in that snapshot?
- 02.06.2018 09:25:36, Eric - sw/eden [@xebb82]: Looks like u got the state as well?
- 02.06.2018 09:25:56, Eric - sw/eden [@xebb82]: rm -r data-dir/state
- Then restart nodeos with —replay-blockchain
- 02.06.2018 09:28:36, deck - EOS New York [@deckw]: same error with state dir removed
- 02.06.2018 09:30:19, Eric - sw/eden [@xebb82]: ok, then remove everything in data-dir and just sync all blocks, shouldn't take any time anyways..
- 02.06.2018 09:31:39, deck - EOS New York [@deckw]: I think @michaelyeates has me blocked.
- 1867630ms thread-0 net_plugin.cpp:1954 operator() ] connection failed to 54.229.241.156:19876: Connection refused
- 02.06.2018 09:31:48, deck - EOS New York [@deckw]: or whoever that is
- 02.06.2018 09:52:10, deck - EOS New York [@deckw]: @xebb82 and whoever want to know. Through first 10000 accounts there are 188 errors ( this includes b1)
- 02.06.2018 09:54:54, deck - EOS New York [@deckw]: https://d3ck.keybase.pub/gb_bad_accounts.txt
- 02.06.2018 09:55:08, deck - EOS New York [@deckw]: some formatting issues there
- 02.06.2018 10:20:57, deck - EOS New York [@deckw]: Hmm this may be my issue.
- 02.06.2018 10:21:48, deck - EOS New York [@deckw]: It looks like if there is less than 10 EOS in the account you guys stake all of it?
- 02.06.2018 10:24:12, Pete BlockMatrix [@blockmatrixpete]: [[Webpage]]im pretty sure that this was the logic applied: https://t.me/jungletestnet/25344 - and the code interpretation is here: https://github.com/CryptoLions/EOS-ERC20-Distribution/blob/master/drop.sh
- 02.06.2018 10:25:37, deck - EOS New York [@deckw]: great that makes sense to what I just worked through.
- 02.06.2018 10:26:00, Pete BlockMatrix [@blockmatrixpete]: sweet 👍
- 02.06.2018 10:46:11, Eric - sw/eden [@xebb82]: We used Bohdans script.. It treats it differently if there's > or < 10
- 02.06.2018 10:47:03, Bohdan CryptoLions 🦁 [@BohdanCryptoLions]: [[Webpage]]it works according to that answer: https://t.me/jungletestnet/25344
- 02.06.2018 10:47:52, deck - EOS New York [@deckw]: Yup thanks guys @BohdanCryptoLions @xebb82 . I have one fix and then I should have some good things for you
- 02.06.2018 10:48:38, Eric - sw/eden [@xebb82]: Great work deck! :D
- 02.06.2018 10:48:55, Hyunwoong JI - EOSeoul.io [@EOSeoul_hyunwoong]: When it comes to initial 8 kb ram allocation for every account, there could be two possible ways to solve it.
- 1. Suggestion by Alex : Issue more EOS to buy all the RAM for every account and destroy the excess.
- 2. Calculate all the RAM which is needed to create all the account and remove the amount from initial balance of the Bancor contract.
- The two methods are similar but as a result, the following differences exist.
- Method 1 will increase the EOS balance of the Bancor contract and Method 2 will not. This is a matter of choice. We will implement method 2 just in case.
- 02.06.2018 10:55:00, jem [@jemxpat]: 👍
- .....02.06.2018 11:04:28, Roman CryptoLions [@romanskaskiw]: >>invited Sudip 🦁 [@JTF420]<<
- .....02.06.2018 11:08:42, Sharif Bouktila - EOS Dublin [@sharifbouktila]: >>unsupported service message type: messageActionPinMessage<<
- 02.06.2018 11:35:39, Hyunwoong JI - EOSeoul.io [@EOSeoul_hyunwoong]: or combined one. after processing method of Alex, remove added EOS from balance of Bancor contract.
- 02.06.2018 11:52:06, mini - EOS Argentina [@moonxor]: Anyone tried firestarter validation?
- 02.06.2018 12:05:33, deck - EOS New York [@deckw]: @BohdanCryptoLions @xebb82 I haven't run a full run yet but it looks like there is an issue with accounts that hold 2 EOS.
- 02.06.2018 12:06:08, deck - EOS New York [@deckw]: It looks like they have staked the .90 and bought ram for .1 but they do not have a balance of EOS. Maybe I'm missing something
- 02.06.2018 12:40:09, Bohdan CryptoLions 🦁 [@BohdanCryptoLions]: For all RAM 0.1 EOS
- 2 EOS <= 11 EOS :
- STAKE_NET "0.4500 EOS"
- STAKE_CPU "0.4500 EOS"
- TRANSFER=$(($EOS-1)) = 1 EOS
- 02.06.2018 12:50:14, deck - EOS New York [@deckw]: yea but that doesn't look like what happened on chain. For example account : ha4timzqgage
- cleos get account ha4timzqgage
- shows that they have the .45 cpu and .45 net staked.
- now cleos get currency balance eosio.token ha4timzqgage EOS should return 1.0000 EOS however it simply returns nothing.
- 02.06.2018 12:56:35, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Has anyone got a page ready where we will be able to allow Users (outside world) to query their balances on EOS Mainnet candidate?
- 02.06.2018 13:03:38, Bohdan CryptoLions 🦁 [@BohdanCryptoLions]: In this case probable went something wrong.. @igorls please check logs.
- 02.06.2018 13:06:17, deck - EOS New York [@deckw]: It is weird that most of the failures I saw happened with accounts with 2 EOS. Just worth noting. Full logs to come
- 02.06.2018 13:07:20, Bohdan CryptoLions 🦁 [@BohdanCryptoLions]: can you give some account examples? I'll check in Jungles
- 02.06.2018 13:07:29, Bohdan CryptoLions 🦁 [@BohdanCryptoLions]: after I will do a drop
- 02.06.2018 13:08:59, deck - EOS New York [@deckw]: yes sir
- 02.06.2018 13:10:50, deck - EOS New York [@deckw]: gi4tcnqgenes
- gm2danygenes
- gqydqoagenes
- gq3tonqgenes
- gy3tonygenes
- ha3dsmqgenes
- he4tqnqgenes
- gezdonrsgene
- geztcnzugene
- ge2tknbvgene
- ge4dmmjzgene
- gi3domrxgene
- gi3tcmzygene
- gmytqmjrgene
- gm2tgmrvgene
- gm3tonzugene
- gm4tembqgene
- gm4tqnrtgene
- gm4tqnrtgene
- gqytaobrgene
- gqytimrqgene
- gq2tgnbqgene
- gq3tcnrrgene
- guydamrtgene
- 02.06.2018 13:11:59, Bohdan CryptoLions 🦁 [@BohdanCryptoLions]: thanks.. I am prepering to drop right now, so in 1-2 h we can check this
- 02.06.2018 13:15:58, Daniel Larimer [@daniellarimer]: What key are you associated with B1
- 02.06.2018 13:16:04, Bohdan CryptoLions 🦁 [@BohdanCryptoLions]: checking distribution:
- gi4tcnqgenes 1.0000
- gm2danygenes 1.0000
- gqydqoagenes 1.0000
- 02.06.2018 13:16:27, Bohdan CryptoLions 🦁 [@BohdanCryptoLions]: "0x00000000000000000000000000000000000000b1","b1","EOS5cujNHGMYZZ2tgByyNEUaoPLFhZVmGXbZc9BLJeQkKZFqGYEiQ","100000000.0100"
- 02.06.2018 13:17:19, Daniel Larimer [@daniellarimer]: I’ll double triple check this in a bit.
- 02.06.2018 13:20:48, Sharif Bouktila - EOS Dublin [@sharifbouktila]: @daniellarimer any thoughts on the checks we should do to validate the chain before releasing it?
- 02.06.2018 13:22:19, Sandwich [@sandw1ch]: That's been checked after every release.
- 02.06.2018 14:19:45, mini - EOS Argentina [@moonxor]: hey I was checking the the mainnet repo, it would be nice to manage it like the bitcoin repo, where you rebase periodically and your commits are always on top of the latest block one release
- 02.06.2018 14:24:32, deck - EOS New York [@deckw]: Can I get the snapshot you are using so I can confirm I am using the right one
- 02.06.2018 14:29:12, mini - EOS Argentina [@moonxor]: [[Webpage]]https://github.com/EOS-Mainnet/eos/commit/a416db018c043958671b80f0b70b9d0b85c8075b due to cache , this wouldnt work if you already build with core symbole name sys. it would be better to just set the coresymbole name without the if.
- 02.06.2018 14:38:52, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Guys,
- We only have 2 things to be concerned about now.
- 1. Validation Plan - Which as a BP we want executed to declare MainNet
- 2. Time of Mainnet Launch - eg. 13:00 UTC tomorrow.
- I have pinned a Google Document to this and some people have added some suggestions.
- We should work to have develop it into a test plan and also agree a way for people to report issues to a core group to check/escalate.
- I know lots of people are feeling the pressure at the moment.
- If you are tired take some rest.
- If you have questions or concerns please share them.
- I've no doubt that we will get this chain launched and validated soon.
- We need to self organise and focus on only what is critical.
- 02.06.2018 14:44:05, Kyunghwan Kim - EOSeoul.io [@resvera]: [[Webpage]]Implemented 'mitigatebiosramcost' action just in case, but not tested. This is combined approach of @EOSeoul_hyunwoong . It only adjusts EOS balance of Bancor when using 'mitigatebiosramcost' action with EOS used for new account. And it can be called only once.
- https://gist.github.com/redjade/c22b2d7fcab068af67c0de31724eb710
- 02.06.2018 14:45:47, Hyunwoong JI - EOSeoul.io [@EOSeoul_hyunwoong]: 👍
- 02.06.2018 15:20:27, Marc-Antoine Ross - EOS Canada [@marcantoineross]: 👏
- 02.06.2018 15:32:29, mini - EOS Argentina [@moonxor]: We are adding more validations, but we don't know if there Is consensus about the following.
- 1- Token distribución: how much Is liquid, how much for CPU and how much for net,
- 2- RAM, Is going to be 64? This should be validated
- 3- max supply, 50billons?
- 4- what would happend with the remaining EOS that eosio would have, should we send them to the workers fund account or should eosio keep them?
- .....02.06.2018 15:33:02, Ross - Cypherglass [@Ross_Cypherglass]: >>invited Leo Ribeiro [@leordev]<<
- 02.06.2018 15:35:43, Patryk @perduta from Tokenika [@perduta]: i think we should burn them, but that's only a opinion
- 02.06.2018 15:46:03, mini - EOS Argentina [@moonxor]: The eosio.token contract can't burn tokens :(
- 02.06.2018 15:47:06, Bohdan CryptoLions 🦁 [@BohdanCryptoLions]: you can send it to EOS1111111111111111111111111111111114T1Anm
- 02.06.2018 15:48:32, mini - EOS Argentina [@moonxor]: You can't Transfer to an address
- 02.06.2018 15:49:09, mini - EOS Argentina [@moonxor]: You csn Transfer to an account with no authorities
- 02.06.2018 15:49:33, Bohdan CryptoLions 🦁 [@BohdanCryptoLions]: 👍
- 02.06.2018 15:49:53, mini - EOS Argentina [@moonxor]: You were saying that?
- 02.06.2018 15:50:36, Bohdan CryptoLions 🦁 [@BohdanCryptoLions]: you are right.. but I didnt try that... or create account with that address
- 02.06.2018 15:59:30, mini - EOS Argentina [@moonxor]: I think what would work Is: 1-50/50 for distribution,
- 2 ram 64gb,
- 3 50B of max supply,
- 4- we send Any remainings to the workers fund.
- But this should be put to consensus
- 02.06.2018 16:01:00, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: I am fine with this, lets get everyone else writing a validation tool to agree to this
- 02.06.2018 16:01:10, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: @deckw @igorls
- 02.06.2018 16:01:19, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: anyone have the telegram for EOSBIXIN? they should be in here
- 02.06.2018 16:03:58, deck - EOS New York [@deckw]: sounds good. @abourget to ensure 50B is the max supply. I think he set it to 10B
- 02.06.2018 17:39:18, Alexandre Bourget - EOS Canada [@abourget]: According to my calculation, with 5% inflation yearly, 10B is largely enough..
- 02.06.2018 17:39:24, Alexandre Bourget - EOS Canada [@abourget]: Do you have a reason to put 50B there ?
- 02.06.2018 17:39:41, Alexandre Bourget - EOS Canada [@abourget]: in 10 years.. we can probably tweak the contract to tweak that number a bit..
- 02.06.2018 17:42:04, Alexandre Bourget - EOS Canada [@abourget]: @xebb82 @jemxpat tested the "contract-whitelist".. and it works well. you can stop any action from going through.
- 02.06.2018 17:42:07, mini - EOS Argentina [@moonxor]: 10B sounds fair, 50B was a more of "just in case"
- 02.06.2018 17:42:24, Alexandre Bourget - EOS Canada [@abourget]: we can tweak that later
- 02.06.2018 18:40:37, Alexandre Bourget - EOS Canada [@abourget]: EVERYONE:
- Please run any validation on chain: stage20.eb.infra.eosamsterdam.net:9876
- It uses snapshot 350/0.. see pinned message on the EOSIO BIOS Boot channel
- 02.06.2018 19:11:32, mini - EOS Argentina [@moonxor]: [[Webpage]]hey, more info about our validation tool
- https://steemit.com/@eosargentina/firestarter-boot-and-validation-tool
- 02.06.2018 19:32:09, mini - EOS Argentina [@moonxor]: hey would love to, do you have a block where you know for sure that eosio has resigned?
- 02.06.2018 19:47:48, Marc-Antoine Ross - EOS Canada [@marcantoineross]: From Alex:
- TO ALL:
- * Validation succeeded 696158/696158, took 28 minutes to inject full snapshot. 350/0
- * I updated https://abourget.keybase.pub/stage20/genesis.json .. RESET YOUR CHAIN with that genesis.json and execute meshing like you did this morning..
- * Genesis public key: EOS6QbvV1pE9UdxbozZVPPiNyFTrvUZCnohpEYpib8jKUKyy7r1sA
- * Resignation block: 3189
- * Chain ID: 51b4f3b0800c7af8c4f2ce5d5038517b7090cecc0b7e63b8d6cab079038a8568
- 02.06.2018 19:49:26, Nathan Rempel | GenerEOS Sydney [@nsrempel]: Was this a dry run validation? Just to clarify
- 02.06.2018 19:51:04, Marc-Antoine Ross - EOS Canada [@marcantoineross]: full snapshot. 350/0 so before the contract freeze
- .....02.06.2018 19:52:00, Marc-Antoine Ross - EOS Canada [@marcantoineross]: >>invited Denis | EOS Nation [@DenisCarriere]<<
- 02.06.2018 19:52:19, Zhao Yu@ EOSLaoMao.com [@zhaoyu]: not the final version
- 02.06.2018 19:52:43, Rohan - EOSauthority [@RohanWillAnswer]: I missed this group..
- 02.06.2018 19:52:47, Marc-Antoine Ross - EOS Canada [@marcantoineross]: @RohanWillAnswer @DenisCarriere How's the snapshot consensus going?
- 02.06.2018 19:52:51, Rohan - EOSauthority [@RohanWillAnswer]: [[Webpage]]Here is the final snapshot v1 on the EOS Authority repo
- https://github.com/eosauthority/genesis/tree/master/snapshot-files/final/1
- 02.06.2018 19:53:13, Rohan - EOSauthority [@RohanWillAnswer]: I didn’t get a chance to compare with anyone else. I’ve had another machine and it was a perfect match!
- 02.06.2018 19:53:13, Marc-Antoine Ross - EOS Canada [@marcantoineross]: can you please PR eos-bios?
- 02.06.2018 19:54:26, Rohan - EOSauthority [@RohanWillAnswer]: I just know of @igorls and @deckw who are nearly there.. Also I suspect others are close..
- 02.06.2018 19:55:00, mini - EOS Argentina [@moonxor]: Awesome!
- 02.06.2018 19:55:31, Marc-Antoine Ross - EOS Canada [@marcantoineross]: let us know how the validation goes please
- 02.06.2018 19:58:58, Alexandre Bourget - EOS Canada [@abourget]: Around 3479.. there's no transactions in blocks after that for the most part.
- 02.06.2018 19:59:17, Alexandre Bourget - EOS Canada [@abourget]: Check the output.log on keybase (same dir as the genesis.json)
- 02.06.2018 20:12:06, Denis | EOS Nation [@DenisCarriere]: @abourget I'm syncing to @monetashi 's endpoint and for some reason when I query eosio there's a very large balance left over:
- $ cleos get currency balance eosio.token eosio
- 1560977134.1908 EOS
- Also when I try to query my own accounts, none of them are present in the chain.
- I'm using chain_id => 51b4f3b0800c7af8c4f2ce5d5038517b7090cecc0b7e63b8d6cab079038a8568
- 02.06.2018 20:14:28, deck - EOS New York [@deckw]: can someone confirm their nodeos --version output for 1.0.1?
- 02.06.2018 20:14:46, Nathan Rempel | GenerEOS Sydney [@nsrempel]: nodeos —version
- 157394272
- 02.06.2018 20:14:50, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: 157394272
- 02.06.2018 20:14:56, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: thats right off of eosio branch
- 02.06.2018 20:14:57, deck - EOS New York [@deckw]: ok cool thank you guys
- 02.06.2018 20:39:08, mini - EOS Argentina [@moonxor]: just finish a team meating, sorry for the delay.
- 02.06.2018 20:43:01, mini - EOS Argentina [@moonxor]: [[Webpage]]so, I just added validation criteria inside the google doc that @sharifbouktila posted.
- I think we should divide that in 2 areas, one is software testing(unit testing), wich i think most of them are already in block one unit test, and then the validation criteria for the chain.
- https://docs.google.com/document/d/1Bsi1__u7U4Jvl9JAomdtWu7U0DaWGPcmD9FIAzL_pQM/edit#
- We already added most of this validation to firestarter, and we are working on adding the rest
- 02.06.2018 20:46:54, mini - EOS Argentina [@moonxor]: it would be awesome if we can reach consensus and made a joint statement about chain validation params.
- 02.06.2018 20:53:58, mini - EOS Argentina [@moonxor]: @marcantoineross @abourget
- http://stage20.eb.infra.eosamsterdam.net:9876/v1/chain/get_info
- 02.06.2018 20:54:48, Dafeng Guo - EOS Asia [@dftle]: 👍
- 02.06.2018 20:56:55, mini - EOS Argentina [@moonxor]: need a p2p endpoint to validate
- 02.06.2018 20:58:00, Zhao Yu@ EOSLaoMao.com [@zhaoyu]: try ours
- 02.06.2018 20:58:07, Zhao Yu@ EOSLaoMao.com [@zhaoyu]: fullnode.eoslaomao.com:443
- 02.06.2018 20:58:35, Shaheen Counts - EOS BlockSmith [@SeanPaul19]: +1
- 02.06.2018 20:59:45, mini - EOS Argentina [@moonxor]: same block?
- 02.06.2018 21:12:21, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: Update on snapshots guys?
- 02.06.2018 21:30:16, Tim Weston - GenerEOS (BP Sydney) [@timweston]: @unlockrs Do we have a generic announcment for when the cahin is enabled and can announce its ready to vote on
- 02.06.2018 21:35:27, mini - EOS Argentina [@moonxor]: [[Photo]]
- 02.06.2018 21:36:41, mini - EOS Argentina [@moonxor]: we skip the genesis validation, and we found this
- 02.06.2018 21:36:48, mini - EOS Argentina [@moonxor]: [[Photo]]
- 02.06.2018 21:36:53, Rick Schlesinger - EOS New York [@unlockrs]: @eosnewyork
- .....02.06.2018 21:36:55, Chitty 🅑 - EOS Argentina [@JChitty]: >>invited Sebastian Castro - EOS Argentina 🇦🇷 [@sebicas]<<
- 02.06.2018 21:39:56, Chitty 🅑 - EOS Argentina [@JChitty]: that means that the owner and the active authorities are controlled by a key instead of the eosio account
- 02.06.2018 21:41:13, Dafeng Guo - EOS Asia [@dftle]: @moonxor that's the report for @zhaoyu 's endpoint?
- 02.06.2018 21:41:22, mini - EOS Argentina [@moonxor]: yes, is the only one i have
- 02.06.2018 21:43:13, Dafeng Guo - EOS Asia [@dftle]: @abourget @zhaoyu check the screenshot above, is this a version issue?
- 02.06.2018 21:46:27, WinLin-EOSBIXIN [@hiwinlin]: inital_timestamp not equal? but I check the picture and they are the same
- 02.06.2018 21:47:05, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: hahah
- 02.06.2018 21:47:05, deck - EOS New York [@deckw]: missing .000
- 02.06.2018 21:47:11, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: this
- 02.06.2018 21:47:16, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: different significance
- 02.06.2018 21:47:47, mini - EOS Argentina [@moonxor]: The first one Is extracted, from the blockchain, the second one Is the one you use
- 02.06.2018 21:47:54, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: eosio.ramfree fail verification?
- 02.06.2018 21:49:11, Chitty 🅑 - EOS Argentina [@JChitty]: yes
- 02.06.2018 21:49:48, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: eosio.ramfree not resigned, correct?
- 02.06.2018 21:50:28, mini - EOS Argentina [@moonxor]: @syed_jafri
- .....02.06.2018 21:51:05, Chitty 🅑 - EOS Argentina [@JChitty]: >>invited Gustavo Mick - EOS Argentina [@gustavomick]<<
- 02.06.2018 21:52:10, mini - EOS Argentina [@moonxor]: Yes the screenshot says that
- 02.06.2018 21:56:56, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: @abourget update the genesis in your new chain
- 02.06.2018 22:10:57, Alexandre Bourget - EOS Canada [@abourget]: the latest genesis was from ~4h ago.. I haven't started another one after that..
- 02.06.2018 22:52:39, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: {{FWD: WinLin-EOSBIXIN [@hiwinlin], 02.06.2018 22:51:35}}
- [[Photo]]
- 02.06.2018 22:52:58, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: {{FWD: WinLin-EOSBIXIN [@hiwinlin], 02.06.2018 22:51:56}}
- I have successed to validation the eos-bios boot --signal 👏
- 02.06.2018 22:54:03, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: Looks like there is *some* post-chain validation consensus between 2 tools, we need more though
- 02.06.2018 22:56:46, mini - EOS Argentina [@moonxor]: so @abourget @zhaoyu did you use this snapshot?
- https://raw.githubusercontent.com/eoscanada/eos-bios/master/files/snapshot.csv
- 02.06.2018 22:58:06, Marc-Antoine Ross - EOS Canada [@marcantoineross]: [[Webpage]]the one listed here: https://github.com/eoscanada/eos-bios/blob/master/sample_config/my_discovery_file.yaml
- 02.06.2018 22:58:26, Marc-Antoine Ross - EOS Canada [@marcantoineross]: That is the one we are pushing now (shortly)
- 02.06.2018 22:58:50, Alexandre Bourget - EOS Canada [@abourget]: no sorry. that's old.. I'm using EOS Authority,s snapshots right now.
- I took them out
- 02.06.2018 22:59:57, Nathan Rempel | GenerEOS Sydney [@nsrempel]: stage20 doesn't have my erc-20 account :(
- 02.06.2018 23:00:04, Sebastian Castro - EOS Argentina 🇦🇷 [@sebicas]: [[Webpage]]https://github.com/eosauthority/genesis/tree/master/snapshot-files/final/1
- 02.06.2018 23:00:43, mini - EOS Argentina [@moonxor]: Ok great thanks
- 02.06.2018 23:02:01, Alexandre Bourget - EOS Canada [@abourget]: the chain was forked :) we were testing handoff of ABP, and the person doing the handoff forked the chain..
- 02.06.2018 23:02:07, Alexandre Bourget - EOS Canada [@abourget]: I'm about to push a new genesis.json :)
- 02.06.2018 23:02:22, Nathan Rempel | GenerEOS Sydney [@nsrempel]: hahaha
- 02.06.2018 23:03:09, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: Confirmation from eosDAC, EOS Authority and Cryptolions on snapshots matching up!
- 02.06.2018 23:04:43, Alexandre Bourget - EOS Canada [@abourget]: wwwwwooott
- 02.06.2018 23:05:02, Nathan Rempel | GenerEOS Sydney [@nsrempel]: awesome
- 02.06.2018 23:05:29, Marc-Antoine Ross - EOS Canada [@marcantoineross]: w00t! that is great
- 02.06.2018 23:06:19, Kevin Rose - EOS New York [@eosnewyork]: 👏 👏
- 02.06.2018 23:07:20, Dafeng Guo - EOS Asia [@dftle]: that's awesome!
- 02.06.2018 23:07:53, mini - EOS Argentina [@moonxor]: awesome!
- 02.06.2018 23:08:09, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: [[Webpage]]I have not manually confirmed the cryptolion snapshot, but here is:
- eosdac: https://github.com/eosdac/conversion/tree/master/eos%20final%20snapshot/v1
- eosauthority: https://github.com/eosauthority/genesis/tree/master/snapshot-files/final/2
- 02.06.2018 23:09:53, Zhao Yu@ EOSLaoMao.com [@zhaoyu]: 👍👍
- .....02.06.2018 23:13:29, - eoseoul.io DS Lim [@dslim]: >>invited Yongkil Kwon - EOSeoul.io<<
- 02.06.2018 23:17:29, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: Update on news, we are back to 2 confirmed snapshot, Bohdan and Rohan matched for period 350 but Bohdan’s server crashed for final snapshot!
- 02.06.2018 23:17:47, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: [[Webpage]]These are matching snapshots so far:
- eosdac: https://github.com/eosdac/conversion/tree/master/eos%20final%20snapshot/v1
- eosauthority: https://github.com/eosauthority/genesis/tree/master/snapshot-files/final/2
- 02.06.2018 23:21:27, mini - EOS Argentina [@moonxor]: [[Photo]]
- 02.06.2018 23:21:28, mini - EOS Argentina [@moonxor]: @zhaoyu @dftle @marcantoineross @abourget if the information is valid(
- peer fullnode.eoslaomao.com:443, the snapshot marc gave us, the block number, and the genesis)
- there are some accouns that are not created, and some that dosent have the right stake
- 02.06.2018 23:21:56, mini - EOS Argentina [@moonxor]: one of the accounts that dosent exist is g43tanryhege
- 02.06.2018 23:22:22, mini - EOS Argentina [@moonxor]: we check that by hand, we are looking the wrong stakes to see if its an error on our side
- 02.06.2018 23:22:37, Marc-Antoine Ross - EOS Canada [@marcantoineross]: The chain has not been updated yet, it's still the run from this afternoon which has been forked on ABP handoff
- 02.06.2018 23:22:53, Alexandre Bourget - EOS Canada [@abourget]: Do you want to peer with a new chain ?
- 02.06.2018 23:23:08, Alexandre Bourget - EOS Canada [@abourget]: I have a new genesis.json and have loaded the final/2 snapshot from EOS Authority..
- 02.06.2018 23:24:27, mini - EOS Argentina [@moonxor]: well, if the block that you guys gave me was before the fork, it dosent matter. the accounts should be created before eosio resigns.
- 02.06.2018 23:24:29, mini - EOS Argentina [@moonxor]: sure
- 02.06.2018 23:24:52, Alexandre Bourget - EOS Canada [@abourget]: genesis.json: {"initial_timestamp":"2018-06-03T02:47:08","initial_key":"EOS6P3bsRaNLLLmmqTCf8XwCRPSbPUFivkcDsjkh36WR8KGjqTrLh"}
- 02.06.2018 23:25:04, Alexandre Bourget - EOS Canada [@abourget]: give me a peer address, I'll sync you the blocks and you can quadruple-check :)
- 02.06.2018 23:26:26, Marc-Antoine Ross - EOS Canada [@marcantoineross]: The node was not synced so the fork happened before the resignation block
- 02.06.2018 23:27:24, mini - EOS Argentina [@moonxor]: [[Webpage]]I need a
- - p2p endpoint
- - block number after eosio resigns
- We are going to use
- genesis:
- {"initial_timestamp":"2018-06-03T02:47:08","initial_key":"EOS6P3bsRaNLLLmmqTCf8XwCRPSbPUFivkcDsjkh36WR8KGjqTrLh"}
- snapshot:
- https://github.com/eosauthority/genesis/tree/master/snapshot-files/final/2
- 02.06.2018 23:27:56, mini - EOS Argentina [@moonxor]: that make sense :D.
- 02.06.2018 23:29:48, mini - EOS Argentina [@moonxor]: there is the problem with the stakes also.
- 02.06.2018 23:30:22, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: were you validating against period 350 or latest
- 02.06.2018 23:30:45, mini - EOS Argentina [@moonxor]: to this block @syed_jafri
- 02.06.2018 23:31:03, Marc-Antoine Ross - EOS Canada [@marcantoineross]: do you have details ?
- 02.06.2018 23:31:21, mini - EOS Argentina [@moonxor]: yes, 1 sec
- 02.06.2018 23:35:20, mini - EOS Argentina [@moonxor]: [[Photo]]
- 02.06.2018 23:35:53, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: rounding errors
- 02.06.2018 23:37:52, Alexandre Bourget - EOS Canada [@abourget]: interesting :P
- 02.06.2018 23:38:11, mini - EOS Argentina [@moonxor]: [[Document, size 34'700 bytes]]
- 02.06.2018 23:38:12, Alexandre Bourget - EOS Canada [@abourget]: what am I seeing.. so L is what is in the account, C is CPU and N is Net ?
- 02.06.2018 23:38:28, mini - EOS Argentina [@moonxor]: exactly
- 02.06.2018 23:38:39, mini - EOS Argentina [@moonxor]: most likley the csv was wrong
- 02.06.2018 23:38:54, Alexandre Bourget - EOS Canada [@abourget]: ok, is it possible that the split algo does the rounding resolution on the other way
- 02.06.2018 23:39:33, Alexandre Bourget - EOS Canada [@abourget]: [[Webpage]]no no.. look: https://github.com/eoscanada/eos-bios/blob/master/bios/ops.go#L417-L419
- 02.06.2018 23:39:47, Alexandre Bourget - EOS Canada [@abourget]: if there is rounding on the first, the other will get the off-by-one..
- 02.06.2018 23:40:21, Alexandre Bourget - EOS Canada [@abourget]: if your validation tool uses a slightly different algo, it might be off-by-one..
- 02.06.2018 23:40:31, Alexandre Bourget - EOS Canada [@abourget]: so maybe we can tolerate a difference of 0.0001 EOS, what do you think ?
- 02.06.2018 23:41:47, mini - EOS Argentina [@moonxor]: most likely we are looking to a chain that wasnt boot with the same snapshot
- 02.06.2018 23:42:16, mini - EOS Argentina [@moonxor]: 162 != 164
- 02.06.2018 23:42:34, mini - EOS Argentina [@moonxor]: I would think that there shouldnt be no tolerance
- 02.06.2018 23:43:09, Alexandre Bourget - EOS Canada [@abourget]: well we need the same rounding algorithm
- 02.06.2018 23:43:18, Alexandre Bourget - EOS Canada [@abourget]: can you report the two unequal values ?
- 02.06.2018 23:44:35, WinLin-EOSBIXIN [@hiwinlin]: I think we need to use the Decimal type to calculate and compare
- 02.06.2018 23:45:03, Nathan Rempel | GenerEOS Sydney [@nsrempel]: setup the new genesis, waiting for blocks to start coming in
- 02.06.2018 23:45:05, Nathan Rempel | GenerEOS Sydney [@nsrempel]: chain id 67fc1919ce5f203b0ba1f315ad9059a04056b17b9ac056ccba7a5c8ebd98554a
- 02.06.2018 23:48:22, mini - EOS Argentina [@moonxor]: >>>(100000.0 + 772116.0 + 772116.0)/10000.0
- 164.4232
- vs 162.4232 in the chain
- 02.06.2018 23:49:33, Nathan Rempel | GenerEOS Sydney [@nsrempel]: no getting anything synced... if someone wants to peer with p2p.genereos.io:9876 would be sweet
- 02.06.2018 23:54:19, mini - EOS Argentina [@moonxor]: we need a synced peer :(
- 02.06.2018 23:55:08, Alexandre Bourget - EOS Canada [@abourget]: let's put peering into in BIOS Boot..
- 02.06.2018 23:55:08, Alexandre Bourget - EOS Canada [@abourget]: I'm about to diffuse a new genesis.json ..
- 03.06.2018 00:02:14, mini - EOS Argentina [@moonxor]: instead of leaving the initial configuration empty, you can run
- nodeos —print-genesis , to generate a valid genesis.
- 03.06.2018 00:03:47, mini - EOS Argentina [@moonxor]: the validator shows the first error, in the stage 20 case, the first error was the timestamp, but the initial params were also wrong, because they were empty.
- 03.06.2018 00:17:43, Zhao Yu@ EOSLaoMao.com [@zhaoyu]: agreed
- 03.06.2018 00:18:04, Alexandre Bourget - EOS Canada [@abourget]: it just means using the defaults.. should we have validation errors because of that ?
- 03.06.2018 00:18:43, Alexandre Bourget - EOS Canada [@abourget]: which validation errors were triggered on the timestamp ?
- 03.06.2018 00:19:07, mini - EOS Argentina [@moonxor]: the genesis used to launch is not the same as the one in the chain, we are validating that
- 03.06.2018 00:19:27, mini - EOS Argentina [@moonxor]: [[Webpage]]https://docs.google.com/document/d/1Bsi1__u7U4Jvl9JAomdtWu7U0DaWGPcmD9FIAzL_pQM/edit#heading=h.qtvfems4zceg
- .....03.06.2018 00:24:52, - eoseoul.io DS Lim [@dslim]: >>invited Sangwook Lee_EOSeoul.io [@NEOPLY]<<
- 03.06.2018 00:53:03, Alexandre Bourget - EOS Canada [@abourget]: you only need to check the chain_id.. it's a hash of timestamp + initialpubkey
- 03.06.2018 00:53:27, Alexandre Bourget - EOS Canada [@abourget]: you want to check parameters too ? eos-bios doesn't write parameters.. it relies on the default params of the boot node
- 03.06.2018 00:54:46, Alexandre Bourget - EOS Canada [@abourget]: is that still an issue ? this doesn't seem like a rounding error, does it ?
- 03.06.2018 00:59:47, mini - EOS Argentina [@moonxor]: I need the p2p endpoint and the block number to validate, if you could give me one would be awesome. I'm going home right now to get some sleep :( , but i can do it in a couple of hours
- 03.06.2018 01:00:11, Alexandre Bourget - EOS Canada [@abourget]: ok, get some rest.. it's important.. I'll have an endpoint and block height to validate..
- 03.06.2018 01:00:28, Alexandre Bourget - EOS Canada [@abourget]: the chain is currently frozen, in the sense that no tx can get in (locked by contract-whitelist)
- 03.06.2018 01:00:56, mini - EOS Argentina [@moonxor]: :) i'm in a car right now otherwise would be happy to do it now
- .....03.06.2018 01:08:45, Hahn - EOS NodeOne [@hahnryu]: >>invited Leo Seo - EOS NodeOne [@sspleo]<<
- 03.06.2018 02:10:31, Rohan - EOSauthority [@RohanWillAnswer]: Hello, Good morning.. I’m back up.. Did anyone else get the final snapshots?
- 03.06.2018 02:10:48, Rohan - EOSauthority [@RohanWillAnswer]: Trying to catch up with all the other messages
- .....03.06.2018 02:40:38, Tim Weston - GenerEOS (BP Sydney) [@timweston]: >>invited Tom Fu GenerEOS (BP Sydney) [@Fuble]<<
- 03.06.2018 03:05:42, Rohan - EOSauthority [@RohanWillAnswer]: Ok, I have some good news..
- 03.06.2018 03:06:04, Rohan - EOSauthority [@RohanWillAnswer]: [[Webpage]]This is kinda amazing!! I can’t believe it..
- EOS Authority - Folder 2 (our last version) is a perfect match with EOS DAC i compared this morning!
- Do a diff on snapshot.csv and snapshot_unregistered.csv
- Links:
- https://github.com/eosauthority/genesis/tree/master/snapshot-files/final/2 (EOS Authority - Final 2 - 6 hours ago)
- https://github.com/eosdac/conversion/tree/master/eos%20final%20snapshot/v1 (EOS Dac v1 - 4 hours ago)
- Will wait for other BPs snapshots as well.
- @syed_jafri @sandw1ch @Lifex10 @igorls @abourget @BohdanCryptoLions
- 03.06.2018 03:07:04, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Awesome !!!
- 03.06.2018 03:08:02, Hahn - EOS NodeOne [@hahnryu]: Great! Much earlier than I thought!
- 03.06.2018 03:08:09, Shaheen Counts - EOS BlockSmith [@SeanPaul19]: That’s Fantastic!!!
- 03.06.2018 03:09:51, Shaheen Counts - EOS BlockSmith [@SeanPaul19]: [[GIF, size 132'456 bytes]]
- 03.06.2018 03:10:31, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: We still want atleast 3 independant confirmations, ideally 5
- 03.06.2018 03:10:40, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: I believe @deckw’s node kept failing too
- 03.06.2018 03:11:47, Shaheen Counts - EOS BlockSmith [@SeanPaul19]: Agreed. But this is a great start.
- 03.06.2018 03:12:28, Rohan - EOSauthority [@RohanWillAnswer]: I had two nodes running. One crashed and had to be restarted..
- 03.06.2018 03:12:36, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: Yes, we’ve known EOS Authority and eosDAC have been matching for over 4 hours
- 03.06.2018 03:12:56, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: He said
- 03.06.2018 03:12:59, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: he restarted twice
- 03.06.2018 03:13:03, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: doing it 3rd time..
- 03.06.2018 03:56:23, Sam - EOS Dublin [@samnoble]: That's not cool. What are you crashing on?
- 03.06.2018 03:57:06, Rohan - EOSauthority [@RohanWillAnswer]: That was yesterday.. A restart worked fine..
- 03.06.2018 03:57:18, Sam - EOS Dublin [@samnoble]: Awesome
- .....03.06.2018 03:58:50, Sharif Bouktila - EOS Dublin [@sharifbouktila]: >>changed group name to "EOS Validation (EMLG)"<<
- 03.06.2018 04:01:19, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Can we agree to use this channel for EOS MAINNET LAUNCH validation communications.
- I believe all key people are here
- Please note we cannot launch until we have Governance Documents , looks like they will be a couple of hours.
- Cool heads guys.
- Let’s do this!!
- 03.06.2018 04:02:23, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: We had @maoyifeng, @DenisCarriere and @dftle work hard today in including the Ricardian contracts
- 03.06.2018 04:02:37, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: We also included the Consitution into the abi
- 03.06.2018 04:04:55, Sharif Bouktila - EOS Dublin [@sharifbouktila]: I see @dftle in the other discussions , it’s coming together!
- 03.06.2018 04:10:29, Igor Lins e Silva - EOS Rio [@igorls]: {{FWD: Igor Lins e Silva - EOS Rio [@igorls], 03.06.2018 04:10:17}}
- [[Document, size 19'949'615 bytes]]
- .....03.06.2018 04:13:55, Lachezar Iordanov [@LachezarIordanov]: >>unsupported service message type: messageActionChatJoinedByLink<<
- .....03.06.2018 04:14:00, Alex - EOS SOCAL [@gmory]: >>unsupported service message type: messageActionChatJoinedByLink<<
- .....03.06.2018 04:14:24, Lachezar Iordanov [@LachezarIordanov]: >>invited J.T. - EOS the World [@HartParr]<<
- .....03.06.2018 04:14:25, Libai - Eosio.SG [@Libai_EosioSG]: >>unsupported service message type: messageActionChatJoinedByLink<<
- .....03.06.2018 04:15:08, Guillaume @eostitan.com [@gyom_eos_titan]: >>unsupported service message type: messageActionChatJoinedByLink<<
- .....03.06.2018 04:15:17, Ben Sigman [@Bensig]: >>unsupported service message type: messageActionChatJoinedByLink<<
- .....03.06.2018 04:15:42, Leon - EOSIO.SG [@leonsun]: >>unsupported service message type: messageActionChatJoinedByLink<<
- .....03.06.2018 04:16:17, Andrey Marchuk [@madz7z]: >>unsupported service message type: messageActionChatJoinedByLink<<
- .....03.06.2018 04:16:31, [email protected] [@qianxiaofeng]: >>unsupported service message type: messageActionChatJoinedByLink<<
- .....03.06.2018 04:16:54, Jack - AcroEOS [@gjacklee]: >>unsupported service message type: messageActionChatJoinedByLink<<
- .....03.06.2018 04:17:31, Jack - AcroEOS [@gjacklee]: >>unsupported service message type: messageActionChatAddUsers<<
- 03.06.2018 04:17:38, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Welcome everyone, let’s keep this focused on Validation and Mainnet Prep.
- Eric and I are the only admins.
- Any shit or unwanted peeps let me know.
- 03.06.2018 04:27:08, Rohan - EOSauthority [@RohanWillAnswer]: Sounds good..
- 03.06.2018 04:27:27, Rohan - EOSauthority [@RohanWillAnswer]: [[Webpage]]We’ve also loaded our latest snapshot on https://eosauthority.com
- 03.06.2018 04:27:32, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: [[Webpage]]VALIDATION REPORT:
- I have personally validated the output files from EOS Authority, EOS Rio and eosDAC and they all have the same output files and output checksums!
- Here is an ongoing repository with all snapshots: https://github.com/eoscafe/eos-snapshot-validation
- 03.06.2018 04:27:44, Rohan - EOSauthority [@RohanWillAnswer]: No reports of any issues on balances..
- 03.06.2018 04:30:07, Rohan - EOSauthority [@RohanWillAnswer]: Excellent news! 👍
- 03.06.2018 04:30:29, Rohan - EOSauthority [@RohanWillAnswer]: Awesome!!
- 03.06.2018 04:31:01, Ralf Weinand | GenerEOS (BP Sydney) [@RalfWei]: Excellent work guys! 3 independent matching validations. We are getting close...
- 03.06.2018 04:33:13, Roshan - EOS Authority [@telepound]: excellent.
- 03.06.2018 05:07:14, Hahn - EOS NodeOne [@hahnryu]: I just shared the news to the community that EOSAuthority&eosDAC snapshots were found identical. And there was a person who claimed that there's difference in what is seen between MEW contract key string/EOSCountdown and EOS Authority&eosDAC snapshot. He said he registered the same eth wallet twice to make sure it is properly registered, once in last feb, and once in 28th May. While MEW Contract Key String & eoscountdown.com shows his newly resigered public key, The eosdac/eosauthority snapshot seems to show his old one.
- His ETH Address : 0xF828CCEc5c3744D0F56a67cfab736066fc3Bd6A6
- [Mew Contract Key String / eoscountdown.com] :
- EOS8QunSrNAfdtyx2MT5gnYkisJVNYFwUJ8ttMDKVuCTLgfkxEko3
- [eosauthority / eos dac Snapshot] :
- EOS66VYof6zt8yRXQYjw7zKMWQgWZpoBgXHUU5cp8iTUCFVa4uAUP
- He's now confused which is right one.
- Guess there might be more of these cases where people registered the same ETH address for more than twice.
- 03.06.2018 05:17:15, David P - EOS42 [@GunnisonCap]: Logically the registration did not work on the second attempt properly due to lack of gas etc. Otherwise it would have overwritten the previous.
- 03.06.2018 05:18:40, Hahn - EOS NodeOne [@hahnryu]: He thought it's overwritten successfuly when he checked from eoscountdown.com, but what happen to appear in the snapshot is still the old one.
- 03.06.2018 05:18:49, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: He has 5000 EOS
- 03.06.2018 05:18:51, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: registered
- 03.06.2018 05:19:05, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: with ETH key 0xF828CCEc5c3744D0F56a67cfab736066fc3Bd6A6
- 03.06.2018 05:19:09, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: tell him to vote for me 😉
- 03.06.2018 05:22:14, Dafeng Guo - EOS Asia [@dftle]: great news
- .....03.06.2018 06:40:38, Dan Levine [@dannymaximo]: >>unsupported service message type: messageActionChatJoinedByLink<<
- .....03.06.2018 06:41:49, CRS [@sss_CRS_sss]: >>unsupported service message type: messageActionChatJoinedByLink<<
- .....03.06.2018 06:52:12, Eric_EOSARMY.io [@Eric_EOSARMY]: >>unsupported service message type: messageActionChatJoinedByLink<<
- 03.06.2018 07:29:07, Sharif Bouktila - EOS Dublin [@sharifbouktila]: [[Webpage]]https://zoom.us/j/581561479
- 03.06.2018 07:29:22, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Opening a Zoom Room for Validation
- 03.06.2018 07:35:07, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: [[Webpage]]🚀 VALIDATION UPDATE 🚀
- I have verified snapshots from the following 5 BPs + 1 Independant to match:
- 1. EOS Authority
- 2. EOS Rio
- 3. eosDAC
- 4. EOSphere
- 5. EOS New York
- 6. Max Infeld
- The snapshot from EOS Store had 2 discrepancies (they were only team using another tool than genesis, they used EOSBIXIN/eostoolkit). The discrepancies were in the accounts:
- 1. 0x94a717c2a6f8e1b2b217fe9b49d04b8de0e430ce
- 2. 0x69d645ad13ca271ddee97da896dae0822516ecfe
- EOS Asia is currently running a snapshot, message me with any other snapshots you find.
- The updated repo with all snapshots is here: https://github.com/eoscafe/eos-snapshot-validation
- .....03.06.2018 07:36:11, Ken (EOSphere) [@ken_eosphere]: >>unsupported service message type: messageActionChatJoinedByLink<<
- 03.06.2018 07:36:12, Roman CryptoLions [@romanskaskiw]: 👍
- 03.06.2018 07:43:26, David P - EOS42 [@GunnisonCap]: Excellent work Syed
- 03.06.2018 07:44:28, Leonard Dorlöchter - EOS network by Advanced Blockchain AG [@EOSnetwork]: 👍
- .....03.06.2018 08:02:28, Jose Toriello [@josetoriello]: >>unsupported service message type: messageActionChatJoinedByLink<<
- .....03.06.2018 08:03:07, Stuardo Rodríguez 🇬🇹 [@stuardo_str]: >>unsupported service message type: messageActionChatJoinedByLink<<
- 03.06.2018 08:03:36, Zhao Yu@ EOSLaoMao.com [@zhaoyu]: there are two accounts missing in EosStore snapshot
- 03.06.2018 08:04:46, Sharif Bouktila - EOS Dublin [@sharifbouktila]: I would suggest if we have 5 matching snapshots we are good. And others need to validate against that base if they still see the need.
- 03.06.2018 08:06:23, deck - EOS New York [@deckw]: Agreed. I believe the next step is to come to an agreement on the final code to use correct?
- .....03.06.2018 08:06:36, Luis Fernando G : EOSMESOdotio 🇬🇹 [@Maski_LF]: >>unsupported service message type: messageActionChatJoinedByLink<<
- 03.06.2018 08:06:47, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Yes @jemxpat what's the current situation on this?
- 03.06.2018 08:07:06, Zhao Yu@ EOSLaoMao.com [@zhaoyu]: we checked these two accounts with both the etherscan.io and etherchain.org, pretty sure EosScore snapshot missed them out
- 03.06.2018 08:07:20, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Cool
- .....03.06.2018 08:07:40, Sharif Bouktila - EOS Dublin [@sharifbouktila]: >>unsupported service message type: messageActionPinMessage<<
- 03.06.2018 08:07:47, Zhao Yu@ EOSLaoMao.com [@zhaoyu]: with balance ~12 EOS and 10 EOS each
- 03.06.2018 08:08:36, Sharif Bouktila - EOS Dublin [@sharifbouktila]: I've pinned @syed_jafri last update so teams can check against that.
- I will continue to pin the most recent validation update as we go through today.
- 03.06.2018 08:09:19, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Have a zoom room ready if we need it. Don't see any need at the moment.
- 03.06.2018 08:11:04, Dafeng Guo - EOS Asia [@dftle]: I like how Syed did it for snapshot validation which we try to set specific goals that once reached we are confident to move to next stage.
- 03.06.2018 08:29:27, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Agreed , awesome job
- 03.06.2018 08:39:29, Devin Lee KOREOS.IO [@indend007]: How about these accounts with "0" balance? Is this removed snapshot? Even if is already registered?
- 03.06.2018 08:42:46, Marc-Antoine Ross - EOS Canada [@marcantoineross]: Hi everyone! How are we doing on validation of the last run? Can I help centralize all issues found somewhere?
- 03.06.2018 08:45:52, Sharif Bouktila - EOS Dublin [@sharifbouktila]: [[Webpage]]Can teams join us here https://zoom.us/j/581561479
- 03.06.2018 08:46:07, Sharif Bouktila - EOS Dublin [@sharifbouktila]: to discuss next steps
- 03.06.2018 08:57:01, Sharif Bouktila - EOS Dublin [@sharifbouktila]: [[Webpage]]https://docs.google.com/document/d/1Bsi1__u7U4Jvl9JAomdtWu7U0DaWGPcmD9FIAzL_pQM/edit#heading=h.qtvfems4zceg
- .....03.06.2018 08:57:07, Sharif Bouktila - EOS Dublin [@sharifbouktila]: >>unsupported service message type: messageActionPinMessage<<
- 03.06.2018 08:58:17, Sharif Bouktila - EOS Dublin [@sharifbouktila]: I've pinned his document. We need to agree what tests are needed for Successful Validation
- .....03.06.2018 09:10:19, Marc-Antoine Ross - EOS Canada [@marcantoineross]: >>invited Josh Kauffman - EOS Canada [@joshkauff]<<
- 03.06.2018 09:19:28, Sharif Bouktila - EOS Dublin [@sharifbouktila]: [[Webpage]]We will then take the information from the Pinned Document and work into this Test Plan
- https://docs.google.com/spreadsheets/d/1t4p1tWs0fXf28nF2w4m5wwzO1cAknYs5XQGJLAIf1pU/edit?usp=sharing
- 03.06.2018 09:19:49, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Please fire ahead if you can. It should be very clear and simple to follow
- 03.06.2018 09:28:26, Sharif Bouktila - EOS Dublin [@sharifbouktila]: [[Document, size 188'738 bytes]]
- 03.06.2018 09:28:26, Sharif Bouktila - EOS Dublin [@sharifbouktila]: As a starting point I've included the tests as listed in this Flow Diagram.
- 03.06.2018 09:59:34, Josh Kauffman - EOS Canada [@joshkauff]: Hey all. Just wanted to pass along the info that all changes that were needed for Governance documents, as well as updating the ABIs have all been done and merged. All good to go.
- 03.06.2018 10:00:29, Jose Toriello [@josetoriello]: I can confirm this.
- 03.06.2018 10:00:33, Nathan Rempel | GenerEOS Sydney [@nsrempel]: woo
- 03.06.2018 10:00:59, Jose Toriello [@josetoriello]: Meaning it is confirmed. Done
- 03.06.2018 10:01:26, mao EOSREAL [@maoyifeng]: GREAT!
- 03.06.2018 10:03:26, Daniel Larimer [@daniellarimer]: 😎
- 03.06.2018 10:03:53, Marc-Antoine Ross - EOS Canada [@marcantoineross]: Great work team! Let's make sure this is in the validation spreadsheet
- 03.06.2018 10:05:51, Rick Schlesinger - EOS New York [@unlockrs]: 👍
- 03.06.2018 10:06:55, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Can I suggest a cut-off of 14:30 UTC for suggesting tests (24 mins from now)
- 03.06.2018 10:07:52, Tim Weston - GenerEOS (BP Sydney) [@timweston]: 👍🏻
- 03.06.2018 10:08:39, Shaheen Counts - EOS BlockSmith [@SeanPaul19]: Awesome!
- 03.06.2018 10:32:54, Alexandre Bourget - EOS Canada [@abourget]: @xebb82 I didn't yet do the block split of resignation actions.. do you think I should ?
- 03.06.2018 10:33:30, Alexandre Bourget - EOS Canada [@abourget]: I'll push this chain out for people to test.. and I'll try to fit it in.. so the next run could be forkable like you wanted.
- 03.06.2018 10:39:17, Sharif Bouktila - EOS Dublin [@sharifbouktila]: [[Document, size 437'889 bytes]]
- 03.06.2018 10:40:22, Sharif Bouktila - EOS Dublin [@sharifbouktila]: This is what we have in the spreadsheet now.
- If we can all move and work there. And make sure we have tests listed and the approach to test each one.
- We will review them on the Zoom call in 1 hour from now 15:40
- Agreed ???????
- .....03.06.2018 10:47:06, Sharif Bouktila - EOS Dublin [@sharifbouktila]: >>unsupported service message type: messageActionPinMessage<<
- 03.06.2018 10:47:20, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Test Plan is now pinned
- 03.06.2018 11:10:35, Sharif Bouktila - EOS Dublin [@sharifbouktila]: We have 28 people in our Zoom Call
- We will be reviewing the Test Plan to try reach consensus at 15:45 UTC - 35 Minutes from now.
- .....03.06.2018 11:15:55, Joe EOS Asia [@gaijinJOE]: >>invited 彭 皓 [@hericpeng]<<
- 03.06.2018 11:47:27, Sharif Bouktila - EOS Dublin [@sharifbouktila]: [[Webpage]]If you aren't already in can you join our Zoom Call
- https://zoom.us/j/581561479
- 03.06.2018 11:47:40, Sharif Bouktila - EOS Dublin [@sharifbouktila]: We can start to look at the Test Plan in 5 minutes
- 03.06.2018 11:48:12, mini - EOS Argentina [@moonxor]: +1
- 03.06.2018 12:59:38, David P - EOS42 [@GunnisonCap]: Can’t make it unfortunately but about to review the plan and will add comments @sharifbouktila
- 03.06.2018 13:18:46, David P - EOS42 [@GunnisonCap]: Few changes made, one I added in relates to whether we want to carry out a coordinated transaction as part of the pause/restart of the chain to simulate implementation of an arbitration ruling.
- 03.06.2018 13:19:07, David P - EOS42 [@GunnisonCap]: Arguably that's one we can do on a testnet and doesn't need to be for the mainnet validation actually
- 03.06.2018 13:19:27, David P - EOS42 [@GunnisonCap]: I'll scrap it, but we do need to practice for such scenarios post launch
- 03.06.2018 14:32:20, Sam - EOS Dublin [@samnoble]: [[Webpage]]All, thanks again for the meeting just now, I think we made some awesome progress! A quick summary of the plan between now and the next meeting which will take place at 1300 UTC on Monday 4th June 2018.
- - Firestarter will be used to perform chain validation with the following testers:
- - Buddy @ EOS New York
- - Trenton @ CSX.IO
- - Mini @ EOS Argentina
- - Everyone else will use the master spreadsheet to perform testing on the net that EOS Canada (EOSIO BIOS BOOT) have running to test as many of the test cases as possible. Sheet is here: https://docs.google.com/spreadsheets/d/1t4p1tWs0fXf28nF2w4m5wwzO1cAknYs5XQGJLAIf1pU/edit#gid=0
- Please remember to document what you do and your results as we can use this for auditing later.
- During the call tomorrow, we'll aim to:
- - Compare notes on our testing
- - Detail any concerns/issues/bugs and how to address them
- - Agree on the next steps
- We're getting very close to getting EOS in to the hands of the community, we just have to continue working together towards this common goal and take it step by step. Remember, it's a marathon and not a race!
- See you all tomorrow at 1300 UTC!
- 03.06.2018 14:38:36, Jose Toriello [@josetoriello]: 👍 Thanks for the leadership
- 03.06.2018 14:38:58, Jose Toriello [@josetoriello]: And also @moonxor even though he is reluctant
- 03.06.2018 14:39:11, Jose Toriello [@josetoriello]: 😜
- 03.06.2018 14:55:47, Denis | EOS Nation [@DenisCarriere]: ⭐️Bang on!
- 03.06.2018 14:57:16, Marc-Antoine Ross - EOS Canada [@marcantoineross]: Great work everyone! Let's leave no stone unturned!
- 03.06.2018 15:21:46, mini - EOS Argentina [@moonxor]: lol, I can help to coordinate the validation if you guys want to, no problem
- 03.06.2018 15:21:50, mini - EOS Argentina [@moonxor]: +1
- 03.06.2018 15:23:04, deck - EOS New York [@deckw]: taking a look now.
- 03.06.2018 15:25:24, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: We need atleast 5 testers
- 03.06.2018 15:25:42, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: With 3 testers, if 1 person is a little off, you’re not sure if the chain is screwed or they screwed up
- 03.06.2018 15:26:16, Rick Schlesinger - EOS New York [@unlockrs]: Agreed. Need more testers please.
- 03.06.2018 15:26:45, mini - EOS Argentina [@moonxor]: we can make a call in about 3/4 hours
- 03.06.2018 15:26:59, mini - EOS Argentina [@moonxor]: so I can show you how to run firestarter
- 03.06.2018 15:27:05, mini - EOS Argentina [@moonxor]: maybe i can work in more documentation
- 03.06.2018 15:27:16, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: A chain is not validated in my opinion without 5 independant testers atleast
- 03.06.2018 15:27:24, mini - EOS Argentina [@moonxor]: +1
- 03.06.2018 15:27:30, mini - EOS Argentina [@moonxor]: I think every abp should validate the chain
- 03.06.2018 15:27:33, mini - EOS Argentina [@moonxor]: but thats my opinion
- 03.06.2018 15:27:36, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: agreed
- 03.06.2018 15:27:50, Rick Schlesinger - EOS New York [@unlockrs]: +1
- 03.06.2018 15:29:47, deck - EOS New York [@deckw]: he said signing himself up :)
- 03.06.2018 15:30:21, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: I could, but I would rather stay independant coordinating the tests
- 03.06.2018 15:30:44, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: I believe doing the testing sort of adds bias to which other validations you confirm
- 03.06.2018 15:34:33, deck - EOS New York [@deckw]: I was just teasing. What you have been doing adds a ton of value for sure. Might be worthwhile to have you takeover @moonxor's job of Test Manager
- 03.06.2018 15:35:37, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: Yes we need @moonxor to focus on teaching people how to use the tool
- 03.06.2018 15:36:10, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: The tool README doesnt look too complicated we could have a dozen or so BPs run through it
- 03.06.2018 15:40:37, deck - EOS New York [@deckw]: not at all complicated. Running now
- 03.06.2018 15:45:13, mini - EOS Argentina [@moonxor]: ❤️
- 03.06.2018 15:48:21, mini - EOS Argentina [@moonxor]: I think the tool is quite simple and dosent need a lot of teaching. Also my english skills are not top notch
- 03.06.2018 15:52:17, mini - EOS Argentina [@moonxor]: I wouldnt be comfortable with anyone been a manager right now, I think we need p2p consensus not a manager. I like what @dftle said about dividing ourselves in tasks.
- we need:
- 1- reach consensus on what a valid blockchain is (BOOT CONSENSUS)
- 2- Before opening the main chain we need to test if the SOFTWARE is working fine. we need to reach a consensus on what means "software working fine".
- after we reach consensus on 1 and 2 we then can think about who would test that and how.
- 03.06.2018 15:54:56, Basit Raza - EOSVibes [@boltxyz]: [[Webpage]]https://github.com/EOSVibes/eos-validation-results
- here is result of check contracts, check accounts, intitial EOS balance using eospy validation script
- 03.06.2018 16:02:32, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Agreed not a manager , but appreciate the work to put some structure on the Test Approach and Plan
- 03.06.2018 16:03:01, mini - EOS Argentina [@moonxor]: yes, i can do that %100.
- 03.06.2018 16:04:12, deck - EOS New York [@deckw]: Hey someone used my tool!!
- 03.06.2018 16:04:23, deck - EOS New York [@deckw]: Even better it worked haha
- 03.06.2018 16:04:57, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: I think manager is a bad term, what we all mean is coordinator
- 03.06.2018 16:05:00, deck - EOS New York [@deckw]: Err I mean of course it would
- 03.06.2018 16:05:27, mini - EOS Argentina [@moonxor]: i like temporary sudoer ❤️
- 03.06.2018 16:05:35, mini - EOS Argentina [@moonxor]: anyone running the mainnet repo?
- 03.06.2018 16:05:51, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: jungle is I believe
- 03.06.2018 16:06:04, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: theyre not using the exact final snapshot though
- 03.06.2018 16:06:09, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: theyre using eosauthority/v1
- 03.06.2018 16:06:14, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: which is missing 2 accounts
- 03.06.2018 16:06:20, mini - EOS Argentina [@moonxor]: i think we should start filling the ricardiang contracts fields in the eosiosystem.abi otherwise the chain would not have a constituttion
- 03.06.2018 16:06:29, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: we have already been doing that
- 03.06.2018 16:06:33, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: 🙂
- 03.06.2018 16:06:36, mini - EOS Argentina [@moonxor]: awesome :D
- 03.06.2018 16:07:09, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: I found out they were missing 48 hours ago, and now @maoyifeng @dftle and @DenisCarriere have worked hard to put most of them in
- 03.06.2018 16:07:30, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: We also put the constitution in eosio.system
- 03.06.2018 16:07:45, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: [[Webpage]]https://github.com/EOS-Mainnet/eos/tree/launch-rc
- .....03.06.2018 16:18:50, Lars Ihler - BitSpace [@LarsIhler]: >>invited Manuel Lains (spectral) [@spectralvision]<<
- 03.06.2018 16:19:45, Alexandre Bourget - EOS Canada [@abourget]: Check the chain that is out there, "get code eosio -a dump.abi"
- 03.06.2018 16:19:56, Alexandre Bourget - EOS Canada [@abourget]: Someone needs to cross check that too :-)
- 03.06.2018 16:20:04, Alexandre Bourget - EOS Canada [@abourget]: Same for eosio.token
- 03.06.2018 16:22:26, mini - EOS Argentina [@moonxor]: yes we are working on validating the chain with an endpoint that we took from eos-bios channel
- 03.06.2018 16:22:41, mini - EOS Argentina [@moonxor]: and updating some stuff to firestarter
- 03.06.2018 16:27:24, Alexandre Bourget - EOS Canada [@abourget]: Super cool :-) you bring a lot of value here mini :-)
- 03.06.2018 16:28:08, mini - EOS Argentina [@moonxor]: would have love to finish firestarter a week or 2 ago :(. But thanks :D
- 03.06.2018 18:15:04, Sharif Bouktila - EOS Dublin [@sharifbouktila]: https://reddit.app.link/UMoFK3PtsN
- Anyone else see this yet?
- 03.06.2018 18:17:09, mini - EOS Argentina [@moonxor]: yes, we are talking about that
- 03.06.2018 18:18:41, Sandwich [@sandw1ch]: 🔥🔥🔥
- 03.06.2018 18:19:14, Stuardo Rodríguez 🇬🇹 [@stuardo_str]: He makes it sound so relaxing, like launching a main net is like a day at the spa!
- 03.06.2018 18:24:57, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: Its actually the way it should be conveyed to voters
- 03.06.2018 18:25:33, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: Its good to keep them calm for us to focus on launch
- 03.06.2018 18:25:39, Stuardo Rodríguez 🇬🇹 [@stuardo_str]: I agree, we are shareing that video. Thank you.
- 03.06.2018 18:28:12, Jose Toriello [@josetoriello]: Great video
- 03.06.2018 18:29:15, Marcin @eosgreen.io [@marcin_eosgreen]: ❤️ , he should make more videos like that, great presentation
- 03.06.2018 19:16:15, Stuardo Rodríguez 🇬🇹 [@stuardo_str]: A question.. when you buy ram (like when you create a new account with the param: --buy-ram-kbytes 8) that money/EOS spent... who gets them?
- 03.06.2018 19:16:19, Stuardo Rodríguez 🇬🇹 [@stuardo_str]: Are they burned?
- 03.06.2018 19:16:40, Daniel Larimer [@daniellarimer]: Eosio.ram.
- 03.06.2018 19:17:00, Daniel Larimer [@daniellarimer]: When you sell you get it back. Most of it
- 03.06.2018 19:18:33, Stuardo Rodríguez 🇬🇹 [@stuardo_str]: but, how can I sell the ram I used to create the new account? Would I need to destroy the account?
- 03.06.2018 19:18:47, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: You cannot
- 03.06.2018 19:18:49, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: Its spent
- 03.06.2018 19:19:05, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: Unless they pay for it themselves
- 03.06.2018 20:14:24, mini - EOS Argentina [@moonxor]: @abourget @marcantoineross we fix some stuff that were wrong in our side, and the validation looks fine, the only thing thats not working is the genesis, and we cant validate global params due to the problem with the genesis, but the rest is looking goooooood
- 03.06.2018 20:14:57, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: is it still the decimal point problems in genesis?
- 03.06.2018 20:15:10, mini - EOS Argentina [@moonxor]: is not just the problem with the decimal points
- 03.06.2018 20:15:49, mini - EOS Argentina [@moonxor]: what I would recommend to eos-bios is to generate a genesis.json with nodeos and use that to boot
- 03.06.2018 20:16:24, mini - EOS Argentina [@moonxor]: we are validating that the parameters that you use to boot the network are the same as the ones when eosio resigns.
- 03.06.2018 20:16:54, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: ah, needs all params as well
- 03.06.2018 20:16:59, mini - EOS Argentina [@moonxor]: yes
- 03.06.2018 20:17:00, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: using nodeos sets them to empty
- 03.06.2018 20:17:04, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: but eos-bios is missing it
- 03.06.2018 20:17:08, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: yes, we should do that
- .....03.06.2018 20:17:24, J.T. - EOS the World [@HartParr]: >>invited Brett Jeffery @ EOS the World [@EOStheWorld]<<
- 03.06.2018 20:17:25, mini - EOS Argentina [@moonxor]: ïs a small detail but i think is important otherwise you can boot with x parameters and change them
- 03.06.2018 20:18:41, Alexandre Bourget - EOS Canada [@abourget]: can you validate that we're simply using the default values ? I'd like to keep the genesis.json dissemination super easy
- 03.06.2018 20:18:46, Alexandre Bourget - EOS Canada [@abourget]: like being able to tweet it :P
- 03.06.2018 20:19:53, Alexandre Bourget - EOS Canada [@abourget]: so you'd be validating that no one fiddled with the defaults
- 03.06.2018 20:20:25, mini - EOS Argentina [@moonxor]: that no one fiddled with the params you use at boot
- 03.06.2018 20:20:41, mini - EOS Argentina [@moonxor]: the idea is to have a vainilla validator
- 03.06.2018 20:20:45, mini - EOS Argentina [@moonxor]: so it can work with any tool :(
- 03.06.2018 20:54:57, Ricardo Marlière - EOS Brazil [@rbmarliere]: using https://github.com/rbmarliere/validate_eos/blob/master/validate_eos.sh to validate balances here
- .....03.06.2018 20:57:01, Jack - AcroEOS [@gjacklee]: >>invited 광영 정<<
- 03.06.2018 20:59:12, Chitty 🅑 - EOS Argentina [@JChitty]: [[Webpage]]During this morning call, some of you were asking about how to quickly inject the snapshot into blockchain, we created this post that goes trough the techniques we used in FireStarter... this code is free to use and you can integrated into your tools https://steemit.com/eosio/@eosargentina/injecting-the-full-erc20-snapshot-in-under-6-minutes-firestarter
- 03.06.2018 21:36:25, David P - EOS42 [@GunnisonCap]: Nice article guys, and great work on Firestarter
- .....03.06.2018 22:07:02, Benjamin Wang | EOShenzhen [@iBenjaminWang]: >>unsupported service message type: messageActionChatJoinedByLink<<
- .....03.06.2018 22:09:38, mao EOSREAL [@maoyifeng]: >>invited Sherman EOSREAL [@ShermanXie]<<
- 03.06.2018 22:16:12, mini - EOS Argentina [@moonxor]: <3
- 03.06.2018 22:20:04, Alexandre Bourget - EOS Canada [@abourget]: @moonxor ohh! now I understand why you want to check the params! because you tweak them on-chain !
- 03.06.2018 22:20:39, Alexandre Bourget - EOS Canada [@abourget]: so you want to make sure they're back to their original values!
- 03.06.2018 22:21:17, mini - EOS Argentina [@moonxor]: yes :) but anyone can do it not only for injection
- 03.06.2018 22:21:40, Alexandre Bourget - EOS Canada [@abourget]: yes.. and it affects the whole network
- 03.06.2018 22:21:46, mini - EOS Argentina [@moonxor]: exactly!
- 03.06.2018 22:22:06, Alexandre Bourget - EOS Canada [@abourget]: and there's no issue syncing for nodes connecting ?
- 03.06.2018 22:22:42, Alexandre Bourget - EOS Canada [@abourget]: they'll process as fast as possible and keep the huge transactions?
- 03.06.2018 22:22:52, Alexandre Bourget - EOS Canada [@abourget]: but mini, can your validation re-check for the default values ?
- 03.06.2018 22:23:07, Alexandre Bourget - EOS Canada [@abourget]: I'd prefer not distributing a big fat genesis.json...
- 03.06.2018 22:23:36, mini - EOS Argentina [@moonxor]: we are already doing that
- 03.06.2018 22:23:55, Alexandre Bourget - EOS Canada [@abourget]: if I have time, I'd like to try your injection strategy.. and you'll make sure all params are reset to defaults.. so we'll be safe
- 03.06.2018 22:24:09, Alexandre Bourget - EOS Canada [@abourget]: very good.. thanks a lot :)
- 03.06.2018 22:24:33, mini - EOS Argentina [@moonxor]: sure! no problem :)
- 03.06.2018 22:24:48, mini - EOS Argentina [@moonxor]: if anyone has other chain to validate please let us know
- 03.06.2018 22:25:04, Alexandre Bourget - EOS Canada [@abourget]: (let me know too :)
- 03.06.2018 22:25:05, Alexandre Bourget - EOS Canada [@abourget]: g'night
- 03.06.2018 22:25:13, mini - EOS Argentina [@moonxor]: haha sure
- 03.06.2018 22:25:15, mini - EOS Argentina [@moonxor]: o/
- 03.06.2018 22:25:58, Lucas @EOS Beijing [@LucasEosBeijing]: hello,eveyone
- 03.06.2018 22:29:58, mini - EOS Argentina [@moonxor]: hi o/
- 03.06.2018 22:42:52, Lucas @EOS Beijing [@LucasEosBeijing]: @abourget Hello, does the contract change?
- 03.06.2018 22:43:21, Lucas @EOS Beijing [@LucasEosBeijing]: Where are you now?
- 03.06.2018 23:58:49, Nathan Rempel | GenerEOS Sydney [@nsrempel]: ./validate_chain_at_block.sh 000017dee519f1785e571480fe72989ac2dc7e4d00ecd0aa925aba6e7af414e4
- Waiting for snapshot to be taken ........ (please wait)
- 03.06.2018 23:58:55, Nathan Rempel | GenerEOS Sydney [@nsrempel]: waiting and waiting 🤣
- 03.06.2018 23:59:30, mini - EOS Argentina [@moonxor]: in the data folder
- 03.06.2018 23:59:37, mini - EOS Argentina [@moonxor]: you can do tail -f name.log
- 03.06.2018 23:59:49, mini - EOS Argentina [@moonxor]: we are working in a progress bar :( sorry guys
- 03.06.2018 23:59:58, Nathan Rempel | GenerEOS Sydney [@nsrempel]: not complaining ;)
- 04.06.2018 00:00:00, Nathan Rempel | GenerEOS Sydney [@nsrempel]: i know its huge
- 04.06.2018 00:00:04, mini - EOS Argentina [@moonxor]: im complaining 😆
- 04.06.2018 00:00:07, Nathan Rempel | GenerEOS Sydney [@nsrempel]: hahaha
- 04.06.2018 00:00:07, Lucas @EOS Beijing [@LucasEosBeijing]: waiting......
- 04.06.2018 00:00:32, Lucas @EOS Beijing [@LucasEosBeijing]: It's better to go to bed.
- 04.06.2018 00:00:48, Nathan Rempel | GenerEOS Sydney [@nsrempel]: difficult, its 2pm in sunny sydney
- 04.06.2018 00:01:12, mini - EOS Argentina [@moonxor]: we try it with our boot that injects everything in under 100 blocks, so when we have to snapshot the block 6000 takes longer and you dont have any warning message
- 04.06.2018 00:01:23, mini - EOS Argentina [@moonxor]: but its the replay
- 04.06.2018 00:03:31, Nathan Rempel | GenerEOS Sydney [@nsrempel]: firestarter/validator/this/data ?
- 04.06.2018 00:04:16, Sandwich [@sandw1ch]: 😍
- 04.06.2018 00:07:23, Nathan Rempel | GenerEOS Sydney [@nsrempel]: can't find name.log
- 04.06.2018 00:07:28, Nathan Rempel | GenerEOS Sydney [@nsrempel]: only eos.log
- 04.06.2018 00:07:36, Gustavo Mick - EOS Argentina [@gustavomick]: pull again
- 04.06.2018 00:07:51, Nathan Rempel | GenerEOS Sydney [@nsrempel]: kk
- 04.06.2018 00:07:55, Gustavo Mick - EOS Argentina [@gustavomick]: 🙂
- 04.06.2018 00:10:03, mini - EOS Argentina [@moonxor]: eos.log :) not name.log, didnt remember the name, si should have put <name>.log
- 04.06.2018 00:10:25, mini - EOS Argentina [@moonxor]: tail -f eos.log
- 04.06.2018 00:10:26, Nathan Rempel | GenerEOS Sydney [@nsrempel]: I assume i need to update the genesis in "this" to match? or atleast the timestamp and start key
- 04.06.2018 00:11:37, Gustavo Mick - EOS Argentina [@gustavomick]: just to confirm, are you trying to validate, or are you trying to validate an specific chain
- 04.06.2018 00:11:57, Nathan Rempel | GenerEOS Sydney [@nsrempel]: trying to validate against the latest bios boot chain
- 04.06.2018 00:12:02, Nathan Rempel | GenerEOS Sydney [@nsrempel]: that im synced to
- 04.06.2018 00:13:25, Gustavo Mick - EOS Argentina [@gustavomick]: yes you need to updated that file or you can use parameters
- 04.06.2018 00:15:00, Nathan Rempel | GenerEOS Sydney [@nsrempel]: ok i can see its starting to sync blocks now
- 04.06.2018 00:15:18, Gustavo Mick - EOS Argentina [@gustavomick]: nice
- 04.06.2018 00:15:56, Gustavo Mick - EOS Argentina [@gustavomick]: sorry if there is any glitch on those params, please push a pr if you think there is something to help with
- 04.06.2018 00:16:07, Nathan Rempel | GenerEOS Sydney [@nsrempel]: 👍
- 04.06.2018 00:16:18, Nathan Rempel | GenerEOS Sydney [@nsrempel]: see how it works out once fully synced
- 04.06.2018 00:18:23, Gustavo Mick - EOS Argentina [@gustavomick]: ok, depending on your hw, snapshot validation could take between 5-15 min. let us know if you have any issue.
- 04.06.2018 00:22:39, Nathan Rempel | GenerEOS Sydney [@nsrempel]: Waiting for snapshot to be taken ........ (please wait)
- 1329899ms thread-0 snapshot_plugin.cpp:89 operator() ] Taking snapshot at block 6110 (000017dee519f1785e571480fe72989ac2dc7e4d00ecd0aa925aba6e7af414e4)...
- Snapshot saved in /home/nathan/firestarter/validator/this/files/snapshot.bin
- Converting snapshot to json ...
- 04.06.2018 00:22:41, Nathan Rempel | GenerEOS Sydney [@nsrempel]: going well so far
- 04.06.2018 00:24:11, Nathan Rempel | GenerEOS Sydney [@nsrempel]: got an error
- 04.06.2018 00:24:12, Nathan Rempel | GenerEOS Sydney [@nsrempel]: Converting snapshot to json ...
- 1362909ms thread-0 snap2json.cpp:182 main ] 8 out_of_range_exception: Out of Range
- read datastream of length 0 over by 1
- {"method":"read","len":0,"over":1}
- thread-0 datastream.cpp:6 throw_datastream_range_error
- error unpacking uint64_t
- {"type":"uint64_t"}
- thread-0 raw.hpp:616 unpack
- 1362909ms thread-0 snap2json.cpp:182 main ] :
- Validating EOS blockchain snapshot against "EOS Vanilla validator (v0.1)"
- Loading EOS snapshot FAIL
- ##########################################################
- No JSON object could be decoded
- Traceback (most recent call last):
- File "validator.py", line 27, in <module>
- getattr(validator,'validate')(args)
- File "/home/nathan/firestarter/validator/vanilla_validator.py", line 316, in validate
- eos_snapshot = load_eos_snapshot(args)
- File "/home/nathan/firestarter/validator/vanilla_validator.py", line 70, in load_eos_snapshot
- snapshot = json.load(file)
- File "/usr/lib/python2.7/json/__init__.py", line 291, in load
- **kw)
- File "/usr/lib/python2.7/json/__init__.py", line 339, in loads
- return _default_decoder.decode(s)
- File "/usr/lib/python2.7/json/decoder.py", line 364, in decode
- obj, end = self.raw_decode(s, idx=_w(s, 0).end())
- File "/usr/lib/python2.7/json/decoder.py", line 382, in raw_decode
- raise ValueError("No JSON object could be decoded")
- ValueError: No JSON object could be decoded
- 04.06.2018 00:24:56, mini - EOS Argentina [@moonxor]: you need to update the eosio plugin
- 04.06.2018 00:25:25, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: Hey mini why do we have to merge snaoshot-plugin
- 04.06.2018 00:25:32, mini - EOS Argentina [@moonxor]: there was a problem with the plugin so we contacted the creator of the plugin, rebuild the nodeos with the latest changes
- 04.06.2018 00:25:59, mini - EOS Argentina [@moonxor]: we use that plugin to take the snapshot :)
- 04.06.2018 00:26:08, Nathan Rempel | GenerEOS Sydney [@nsrempel]: ok, so i was on eos-mainnet/eos tag v1.0.1 and applied git merge origin/snapshot-plugin
- 04.06.2018 00:26:16, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: does the chain need that on mainnet?
- 04.06.2018 00:26:38, Nathan Rempel | GenerEOS Sydney [@nsrempel]: although i hadn't pulled, perhaps missing some update
- 04.06.2018 00:26:40, mini - EOS Argentina [@moonxor]: no you just need it for the validation, the endpoint dosent need that
- 04.06.2018 00:27:09, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: So I am trying to understand this
- 04.06.2018 00:27:11, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: I need a local chain
- 04.06.2018 00:27:15, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: to validate external chain?
- 04.06.2018 00:27:19, mini - EOS Argentina [@moonxor]: for example, if we want to validate a chain booted by EOS Cafe, you dont need that plugin, we just run a nodeos with that plugin against your endpoint
- 04.06.2018 00:27:28, mini - EOS Argentina [@moonxor]: no, a local nodeos
- 04.06.2018 00:27:46, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: [[🐻 Sticker, size 39'322 bytes]]
- 04.06.2018 00:27:50, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: ofcourse, thank you!
- 04.06.2018 00:28:08, mini - EOS Argentina [@moonxor]: no man , no problem! we didnt document anything, sorry
- 04.06.2018 00:28:09, Nathan Rempel | GenerEOS Sydney [@nsrempel]: was hoping it would be a little faster on the machine itself haha
- 04.06.2018 00:28:37, mini - EOS Argentina [@moonxor]: the replay is the problem
- 04.06.2018 00:28:41, mini - EOS Argentina [@moonxor]: :(
- 04.06.2018 00:28:54, Nathan Rempel | GenerEOS Sydney [@nsrempel]: ok i'll try running it local
- 04.06.2018 00:29:03, Nathan Rempel | GenerEOS Sydney [@nsrempel]: with the latest snapshot merge
- 04.06.2018 00:29:10, mini - EOS Argentina [@moonxor]: we cant do anything with that, but if alex or anyone launching use our firestarter injection technique, we could all enjoy some speed :D
- 04.06.2018 01:00:40, mini - EOS Argentina [@moonxor]: ok, i added some stuff to the validator, you now can run everything with arguments, and it shows you the eos.log file instead of the sad waitin message
- 04.06.2018 01:03:31, mini - EOS Argentina [@moonxor]: so, if you want to validate a chain the steps are
- 1- inside /firestarter/validator, create a folder with a name, ex ghost
- 2- inside ghost put your genesis and snapshot files
- 3- run
- ./validate_chain_at_block.sh \
- --network=ghost \
- --block= < block_id> \
- --p2p=<endpoint_ip>:<endpoint_port>
- 04.06.2018 02:08:33, Nathan Rempel | GenerEOS Sydney [@nsrempel]: i get it going off with
- 468008ms thread-0 producer_plugin.cpp:751 start_block ] Not producing block because I don't have the private key for EOS7m4uBgRXBM5DbXFX7FTrNiZyRzxfktmto6qQTzc2z3BhgAGHKM
- 04.06.2018 02:08:58, Nathan Rempel | GenerEOS Sydney [@nsrempel]: thats the initial key in the genesis
- 04.06.2018 02:11:54, mini - EOS Argentina [@moonxor]: You run the./boot.sh with the default config?
- 04.06.2018 02:12:16, Nathan Rempel | GenerEOS Sydney [@nsrempel]: im not trying to boot
- 04.06.2018 02:12:36, Nathan Rempel | GenerEOS Sydney [@nsrempel]: ./validate_chain_at_block.sh —network=eos —block=000017dee519f1785e571480fe72989ac2dc7e4d00ecd0aa925aba6e7af414e4 —p2p=p2p.genereos.io:9876
- 04.06.2018 02:13:00, Nathan Rempel | GenerEOS Sydney [@nsrempel]: just a remote validate
- 04.06.2018 02:13:46, mini - EOS Argentina [@moonxor]: The validator shouldnt produce blocks, im in a cab right now let me get home and i take a look at it
- 04.06.2018 02:13:55, Nathan Rempel | GenerEOS Sydney [@nsrempel]: cool man
- 04.06.2018 02:18:53, Nathan Rempel | GenerEOS Sydney [@nsrempel]: i might have figured it
- 04.06.2018 02:19:09, Nathan Rempel | GenerEOS Sydney [@nsrempel]: i had a config file at the default macOS location that might have been loading a config.ini
- 04.06.2018 02:19:16, Nathan Rempel | GenerEOS Sydney [@nsrempel]: removed it... seems to be ok now
- 04.06.2018 02:26:34, mini - EOS Argentina [@moonxor]: great! would you like to report an issue to our repo? I think validating full macOS compatibility would be awesome.
- 04.06.2018 02:27:15, Nathan Rempel | GenerEOS Sydney [@nsrempel]: sure can do
- 04.06.2018 02:27:58, mini - EOS Argentina [@moonxor]: thanks!
- 04.06.2018 02:30:43, Nathan Rempel | GenerEOS Sydney [@nsrempel]: getting block now 🤞
- 04.06.2018 02:32:04, mini - EOS Argentina [@moonxor]: :D
- 04.06.2018 02:32:31, mini - EOS Argentina [@moonxor]: depending on the chain, its going to take a short or long amount of time, but at least you can see the blocks comming :)
- 04.06.2018 02:32:47, Nathan Rempel | GenerEOS Sydney [@nsrempel]: 3000/6000 haha
- 04.06.2018 02:34:27, mini - EOS Argentina [@moonxor]: \o/
- .....04.06.2018 02:37:49, jem [@jemxpat]: >>invited Timothy Lewis [@trustlayer]<<
- 04.06.2018 02:41:42, Nathan Rempel | GenerEOS Sydney [@nsrempel]: Snapshot saved in /Users/nathan.rempel/firestarter/validator/this/files/snapshot.bin
- Converting snapshot to json ...
- Validating EOS blockchain snapshot against "EOS Vanilla validator (v0.1)"
- Loading EOS snapshot OK
- Loading ETH snapshot OK
- Load EOS genesis OK
- Validating genesis FAIL
- > please check /var/folders/xg/vmswb7md08l3rcf0kjkq973j_h0nqb/T/tmpWLYS9v for details
- Verifying global params vs genesis OK
- Verifying user account creation OK
- Verifying user account stake OK
- Verifying user account code OK
- Verifying user account permission OK
- Verifying privileged accounts OK
- Verifying extra accounts FAIL
- > Extra account found guztqobqgqge
- > Extra account found eosio.burned
- > Extra account found eosio.unregd
- > Extra account found gqytcmrrgige
- 04.06.2018 02:42:19, Nathan Rempel | GenerEOS Sydney [@nsrempel]: very cool
- 04.06.2018 02:46:12, Nathan Rempel | GenerEOS Sydney [@nsrempel]: i copied snapshot.csv from eosbios-stage20 and got Validating EOS blockchain snapshot against "EOS Vanilla validator (v0.1)"
- Loading EOS snapshot OK
- Loading ETH snapshot OK
- Load EOS genesis OK
- Validating genesis FAIL
- > please check /var/folders/xg/vmswb7md08l3rcf0kjkq973j_h0nqb/T/tmpqZbpWb for details
- Verifying global params vs genesis OK
- Verifying user account creation OK
- Verifying user account stake OK
- Verifying user account code OK
- Verifying user account permission OK
- Verifying privileged accounts OK
- Verifying extra accounts FAIL
- > Extra account found eosio.burned
- > Extra account found eosio.unregd
- 04.06.2018 02:46:41, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: eosbios-stage20 snapshot is not the final snapshot
- 04.06.2018 02:47:08, Nathan Rempel | GenerEOS Sydney [@nsrempel]: I understand, just cool to see the difference
- .....04.06.2018 03:10:32, lo chaiching [@lochaiching]: >>unsupported service message type: messageActionChatJoinedByLink<<
- 04.06.2018 03:12:48, Sukesh Tedla (EOSGreen) [@Sukeshtedla]: [[Webpage]]Hi Guys, I have been actively monitoring the news and this is not to mock anyone. I just read this article, I know everyone is busy but it's worth a while reading and maybe reacting on it: https://www.trustnodes.com/2018/06/03/just-10-addresses-hold-nearly-50-eos-tokens
- 04.06.2018 03:17:45, mini - EOS Argentina [@moonxor]: ❤️❤️❤️, it was easy right?
- 04.06.2018 03:28:32, Timothy Lewis [@trustlayer]: [[Webpage]]I would like to gather approval for VULN testing by BPC administrators.
- Please fill the BP tab on the spreadsheet to acknoledge testing acceptance.
- https://docs.google.com/spreadsheets/d/1dgmt_Hk5lqhxkvC5yv9lApzogPlUzQE9QItB15UYprk/edit?usp=sharing
- Please join the EOSIO Security discussion:
- https://t.me/joinchat/FTzsFlKJbwM5wXHeLKP9gA
- EOSIO SECURITY CALL FOR DISCOVERY AT 12:00 UTC
- .....04.06.2018 04:40:27, Jack - AcroEOS [@gjacklee]: >>invited Moon@Acroeos [@moonlightbp]<<
- 04.06.2018 04:41:00, Sharif Bouktila - EOS Dublin [@sharifbouktila]: I read it, it’s a BS article - we are too busy to react to this type of stuff.
- 04.06.2018 04:43:02, Sukesh Tedla (EOSGreen) [@Sukeshtedla]: I agree. Just wanted to share it. I appreciate all the efforts from each and everyone. Thanks for the great work you guys are doing.
- 04.06.2018 04:44:37, Sharif Bouktila - EOS Dublin [@sharifbouktila]: We have a test plan pinned to this group , can I suggest that you add your suggestion for testing to it, and try make a case for it on our 13:00 UTC call
- I could see why each BP may or may not want to take part, but I’m not sure it is relevant to chain validation.
- 04.06.2018 04:47:05, Timothy Lewis [@trustlayer]: It has absolute relevance. We are coordinating with a large group of BPCs now in the EOSIO security channel. We have multiple indipendant security groups testing and will share our results during the 13:00 call.
- 04.06.2018 04:47:27, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: isnt the call at 12:00?
- 04.06.2018 04:47:33, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: the security one atleast
- 04.06.2018 04:47:41, Timothy Lewis [@trustlayer]: Correct, the security call will be 12:00
- 04.06.2018 04:51:57, Timothy Lewis [@trustlayer]: EOSIO Security channel: https://t.me/joinchat/FTzsFlKJbwM5wXHeLKP9gA
- 04.06.2018 08:27:02, deck - EOS New York [@deckw]: My output differed slightly:
- Converting snapshot to json ...
- Validating EOS blockchain snapshot against "EOS Vanilla validator (v0.1)"
- Loading EOS snapshot OK
- Loading ETH snapshot OK
- Load EOS genesis OK
- Validating genesis FAIL
- > please check /tmp/tmpeGTk3G for details
- Verifying global params vs genesis FAIL
- > Unable to compare global params, initial_configuration not found in genesis
- Verifying user account creation OK
- Verifying user account stake OK
- Verifying user account code OK
- Verifying user account permission OK
- Verifying privileged accounts OK
- Verifying extra accounts FAIL
- > Extra account found guztqobqgqge
- > Extra account found eosio.burned
- > Extra account found eosio.unregd
- > Extra account found gqytcmrrgige
- 04.06.2018 08:35:00, Alexandre Bourget - EOS Canada [@abourget]: which chain_id ? and what is your snapshot reference ?
- 04.06.2018 08:35:20, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: The one provided in stage20 snapshot folder
- 04.06.2018 08:35:24, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: In firestarter
- 04.06.2018 08:35:58, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: Does not match with the consensus snapshot, can check again
- 04.06.2018 08:36:07, Alexandre Bourget - EOS Canada [@abourget]: @moonxor so snapshot validation of chain 579a649aae8f660aa1abbab262437596d1f388f709b0b94a9fd6bba479889ea5 failed ?!
- 04.06.2018 08:36:59, Alexandre Bourget - EOS Canada [@abourget]: The snapshot used was https://github.com/eoscanada/eos-bios/blob/master/sample_config/my_discovery_file.yaml#L56
- 04.06.2018 08:37:30, Alexandre Bourget - EOS Canada [@abourget]: Can you confirm it's a different one ? I haven't seen any newer one
- 04.06.2018 08:39:09, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: I checked against
- firestarter/validator/eosbios-stage20/snapshot.csv
- 04.06.2018 08:39:26, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: Not on chain
- 04.06.2018 08:41:26, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: That one is correct
- 04.06.2018 08:41:33, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: I dont know what they have in their folder
- 04.06.2018 08:43:25, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Excellent work so far on Validation. We will be starting our call in 20 mins - 13:00 UTC
- I will post Zoom here in 10 mins.
- 04.06.2018 08:56:13, Sharif Bouktila - EOS Dublin [@sharifbouktila]: [[Webpage]]https://zoom.us/j/551514141
- 04.06.2018 09:11:52, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: The sha256 hash of the snapshot.csv used is 6df61f12f96f89c907fac14a021d788c9e77098952a6c5494c7999d2e79d0a35
- 04.06.2018 09:16:15, Alexandre Bourget - EOS Canada [@abourget]: CHAIN ID: 579a649aae8f660aa1abbab262437596d1f388f709b0b94a9fd6bba479889ea5
- 04.06.2018 09:16:23, Denis | EOS Nation [@DenisCarriere]: {{FWD: Alexandre Bourget - EOS Canada [@abourget], 03.06.2018 11:54:00}}
- To those nearby, can you provide with a p2p endpoints, setup with the following genesis.json:
- {"initial_timestamp":"2018-06-03T13:04:31","initial_key":"EOS7m4uBgRXBM5DbXFX7FTrNiZyRzxfktmto6qQTzc2z3BhgAGHKM"}
- * Chain ID: 579a649aae8f660aa1abbab262437596d1f388f709b0b94a9fd6bba479889ea5
- * Genesis JSON: {"initial_timestamp":"2018-06-03T13:04:31","initial_key":"EOS7m4uBgRXBM5DbXFX7FTrNiZyRzxfktmto6qQTzc2z3BhgAGHKM"}
- * Resignation block: 2683
- * Frozen head block: 6110
- Another RC chain is ready, I'll peer it out for validation. I'll do a better announcement in a short while.
- v1.0.1
- .....04.06.2018 10:08:14, cupide EOS Pacific: >>unsupported service message type: messageActionChatJoinedByLink<<
- .....04.06.2018 10:08:40, Jacky - JEDA [@newbdez33]: >>unsupported service message type: messageActionChatJoinedByLink<<
- .....04.06.2018 10:09:48, Mark-eosantpool [@mark_cike]: >>unsupported service message type: messageActionChatJoinedByLink<<
- .....04.06.2018 10:13:11, Mathieu Boulianne - EOS Canada [@mathieubo]: >>unsupported service message type: messageActionChatJoinedByLink<<
- .....04.06.2018 10:13:37, zhangzc | Onechain [@zhangzc]: >>unsupported service message type: messageActionChatJoinedByLink<<
- .....04.06.2018 10:14:39, [email protected] [@crazybits]: >>unsupported service message type: messageActionChatJoinedByLink<<
- .....04.06.2018 10:16:17, Ren Kaidi - EOSWTZ [@FoxMr]: >>unsupported service message type: messageActionChatJoinedByLink<<
- .....04.06.2018 10:21:54, kenneth: >>unsupported service message type: messageActionChatJoinedByLink<<
- .....04.06.2018 10:21:55, Jason Wang - EOS Pacific [@jason_wang_4595]: >>unsupported service message type: messageActionChatJoinedByLink<<
- .....04.06.2018 10:22:06, Ville Sundell [@SolariusFI]: >>unsupported service message type: messageActionChatJoinedByLink<<
- .....04.06.2018 10:22:21, Elliot@EOS Pacific [@Elliot521]: >>unsupported service message type: messageActionChatJoinedByLink<<
- .....04.06.2018 10:24:14, Discha W [@afeidischa]: >>unsupported service message type: messageActionChatJoinedByLink<<
- .....04.06.2018 10:28:34, Che Hao Yang [@hankych]: >>unsupported service message type: messageActionChatJoinedByLink<<
- .....04.06.2018 10:30:47, Joshua Qiu @EOSeco [@Joshuaqiu]: >>unsupported service message type: messageActionChatJoinedByLink<<
- .....04.06.2018 10:31:11, Angelo Wang: >>unsupported service message type: messageActionChatJoinedByLink<<
- 04.06.2018 10:45:29, Sam - EOS Dublin [@samnoble]: Dear all, thanks again for another productive call. I think we can all agree we're another step closer to a successful and secure EOS launch. Whilst it might seem slow and frustrating, just take a second to think about the scale of what we're doing here. As a quick re-cap:
- - Alex (@abourget) from EOS Canada will now work to push out a new test net and release the private key via the validation channel. Everyone is then free (and encouraged) to use this to create their own test nets
- - Alex from EOS Canada will modify the duration for which tokens are staked to reduce the lock in time to expedite testing. This will NOT be loaded in to the new test net and will be optional for people to load in to their own net
- - Jem (@jemxpat) will run point on this and will facilitate communication with the Ghostbusters team
- - The Ghostbusters team will mesh in to the new network to ensure crossover
- - Everyone will now conduct their own testing either on one of the existing test nets, or one they stand up themselves with the view to using the validation plan as a reference in order to perform their own validation testing of the platform.
- - We should be working towards a position where each BP is satisfied that the current version of the platform is ready for prime time, to launch the main net (pending feedback from the security testing). This will be in the form of some kind of joint statement, TBD.
- - We'll reconvene at 1300 UTC Tuesday June 5th
- - Ben (@Bensig) will work on a public statement that we’ll release to keep the community abreast of the situation.
- We’ll use this channel for coordination.
- 04.06.2018 11:05:37, Josh Kauffman - EOS Canada [@joshkauff]: Just wanted to check and try to grab a more global consensus from BPs. The contents of the chain that was spun up yesterday (chain id: 579a649aae8f660aa1abbab262437596d1f388f709b0b94a9fd6bba479889ea5), do we all agree upon it containing the correct information (constitution, ricardians, snapshot, source code...)? Want to ensure we're all on the same page.
- 04.06.2018 11:06:58, Nathan Rempel | GenerEOS Sydney [@nsrempel]: 👍
- 04.06.2018 11:07:11, Nathan Rempel | GenerEOS Sydney [@nsrempel]: We confirmed
- 04.06.2018 11:07:40, Nathan Rempel | GenerEOS Sydney [@nsrempel]: Although, does it have the very latest ricardian updates?
- 04.06.2018 11:10:05, mini - EOS Argentina [@moonxor]: we didnt test for everything in the spread sheet yet. Was looking good for sure.
- 04.06.2018 11:12:23, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: Have we agreed on a specific commit to launch with?
- 04.06.2018 11:12:29, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: Should we release a launch tag
- 04.06.2018 11:12:55, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: People seem to be pulling v1.0.1 and adding consitution/etc from governance repo
- 04.06.2018 11:13:06, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: That may not line up like a release tag
- .....04.06.2018 11:18:29, Bonnie Wang [@eostarkBP]: >>unsupported service message type: messageActionChatJoinedByLink<<
- 04.06.2018 11:22:53, Sharif Bouktila - EOS Dublin [@sharifbouktila]: I think we need a Launch Tag
- 04.06.2018 11:24:30, James Sutherland - Cypherglass.com [@JamesSutherland]: +1
- 04.06.2018 11:25:19, Guillaume @eostitan.com [@gyom_eos_titan]: agreed
- .....04.06.2018 11:25:52, [email protected] [@PennyYip]: >>unsupported service message type: messageActionChatJoinedByLink<<
- .....04.06.2018 11:27:18, skep: >>unsupported service message type: messageActionChatJoinedByLink<<
- 04.06.2018 11:29:02, Nathan Rempel | GenerEOS Sydney [@nsrempel]: Also agree
- 04.06.2018 11:31:16, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: Have asked in the right channel for that to happen
- 04.06.2018 11:35:00, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: A launch tag is being worked on by Cesar!
- 04.06.2018 11:45:37, Denis | EOS Nation [@DenisCarriere]: By any chance, could we start this next call earlier (proposing 1200 UTC). We could divide the meeting into two parts, first hour can be discussions of validation/testing and 2nd part would be taking/making the decisions. That way if BP's cannot be there for the validation discussion, they can easily join in on the "action" part.
- 1200 UTC (validation/security phase)
- First hour to discussing our validation/testing/security findings
- 1300 UTC (action phase)
- Decide if we have a mainnet candidate:
- - IF TRUE => Launch mainnet candidate and open a limited amount of API endpoints
- - IF FALSE => Define which validation steps failed & propose another meeting in +24 hours
- 04.06.2018 11:46:23, Craig Murray - EOS BlockSmith [@craigmurray100]: just coordinate with the security channel where the earlier of the two meetings is organized
- 04.06.2018 11:46:35, Craig Murray - EOS BlockSmith [@craigmurray100]: im all for it being 1 call 👍
- 04.06.2018 11:47:23, Denis | EOS Nation [@DenisCarriere]: Yes the security discussion could happen on the first hour (1200 UTC), that way we have a good "flow" during the meeting and not have people drop off to join other Zoom video calls
- 04.06.2018 11:48:44, mini - EOS Argentina [@moonxor]: i rather continue with 1300 utc, but if everyone is on board I can make it to 1200 , but is one hour less of sleep for me :P
- 04.06.2018 12:07:31, Alexandre Bourget - EOS Canada [@abourget]: ALL:
- I'm currently injecting the test run, without resignation.
- Genesis JSON: {"initial_timestamp":"2018-06-04T15:31:02","initial_key":"EOS8VtXjprEnsyzyGbo7xXCjcE4omxKhEv5bkJkjUkefHQ84b2j5A"}
- I'll peer it out for people to observe. I would appreciate at least 3 peers to swarm the blocks out.
- 04.06.2018 12:12:00, Niclas Hedhman - EOS Germany [@niclash]: You mean regular full nodes to connect?
- 04.06.2018 12:12:29, Alexandre Bourget - EOS Canada [@abourget]: Yes.. not ones participating in yesterday,s chain..
- 04.06.2018 12:12:36, Alexandre Bourget - EOS Canada [@abourget]: The testnet chain_id is 050ca7388c8b3a394e8c5cf817600cec5f2235c31dc37a2596522e934cf988c0
- 04.06.2018 12:12:53, Joshua Qiu @EOSeco [@Joshuaqiu]: validation.eoseco.com:10020
- 04.06.2018 12:12:54, Alexandre Bourget - EOS Canada [@abourget]: ok, local validation passed.. I'll peer that out now
- 04.06.2018 12:12:55, Alexandre Bourget - EOS Canada [@abourget]: Anyone ready ?
- 04.06.2018 12:13:00, Alexandre Bourget - EOS Canada [@abourget]: Great! Thanks Joshua!
- 04.06.2018 12:13:11, Joshua Qiu @EOSeco [@Joshuaqiu]: your address?
- 04.06.2018 12:13:14, Jacky - JEDA [@newbdez33]: m.jeda.one:3322
- 04.06.2018 12:13:53, Jacky - JEDA [@newbdez33]: just restart my node with new genesis
- 04.06.2018 12:15:39, Alexandre Bourget - EOS Canada [@abourget]: ready ?
- 04.06.2018 12:15:43, Jacky - JEDA [@newbdez33]: yes
- 04.06.2018 12:15:53, Nate D - Aloha EOS [@n8dee]: I'm ready too: peer2.test.alohaeos.com:9876
- 04.06.2018 12:17:13, Jacky - JEDA [@newbdez33]: {"server_version":"0961a560","chain_id":"050ca7388c8b3a394e8c5cf817600cec5f2235c31dc37a2596522e934cf988c0","head_block_num":1,"last_irreversible_block_num":0,
- 04.06.2018 12:17:34, Niclas Hedhman - EOS Germany [@niclash]: I now require "known IP numbers" on the p2p port.
- 04.06.2018 12:18:34, Nate D - Aloha EOS [@n8dee]: allow my host above and I'll peer to you
- 04.06.2018 12:19:02, Nate D - Aloha EOS [@n8dee]: I'm just using a test host
- 04.06.2018 12:20:20, Alexandre Bourget - EOS Canada [@abourget]: Tell me when you get blocks.
- 04.06.2018 12:20:33, Nate D - Aloha EOS [@n8dee]: getting blocks 👍
- 04.06.2018 12:20:45, Jacky - JEDA [@newbdez33]: getting it
- 04.06.2018 12:20:46, Alexandre Bourget - EOS Canada [@abourget]: it's a locked-down chain now..I'll post the private key when you're all past the "pre-resignation" block :)
- .....04.06.2018 12:20:49, Jack - AcroEOS [@gjacklee]: >>invited Myoungchul Park [@sps818]<<
- 04.06.2018 12:21:58, James Sutherland - Cypherglass.com [@JamesSutherland]: 12:00UTC is 5am Pacific time, those of us on the west coast of the US woudl appricate 13:00 or later.
- 04.06.2018 12:22:20, Joshua Qiu @EOSeco [@Joshuaqiu]: make sure , we will give up this chain after validation
- 04.06.2018 12:22:22, Joshua Qiu @EOSeco [@Joshuaqiu]: ?
- 04.06.2018 12:26:52, James Sutherland - Cypherglass.com [@JamesSutherland]: My node is up @ publicnode.cypherglass.com:9876
- 04.06.2018 12:26:57, James Sutherland - Cypherglass.com [@JamesSutherland]: It's syncing
- 04.06.2018 12:27:28, Joshua Qiu @EOSeco [@Joshuaqiu]: syncing
- 04.06.2018 12:27:46, James Sutherland - Cypherglass.com [@JamesSutherland]: 1000
- 04.06.2018 12:29:44, Jacky - JEDA [@newbdez33]: cleos.sh get info
- {
- "server_version": "0961a560",
- "chain_id": "579a649aae8f660aa1abbab262437596d1f388f709b0b94a9fd6bba479889ea5",
- "head_block_num": 6110,
- "last_irreversible_block_num": 6109,
- "last_irreversible_block_id": "000017dd169a1f50a4808f3563517cc1df881a16776a6d9a91e544e0fc31e782",
- "head_block_id": "000017dee519f1785e571480fe72989ac2dc7e4d00ecd0aa925aba6e7af414e4",
- "head_block_time": "2018-06-03T17:36:41",
- "head_block_producer": "eosio",
- "virtual_block_cpu_limit": 735929,
- "virtual_block_net_limit": 472963532,
- "block_cpu_limit": 199900,
- "block_net_limit": 1048576
- }
- 04.06.2018 12:30:18, Nate D - Aloha EOS [@n8dee]: that's a different chain id than I have
- 04.06.2018 12:30:58, Nate D - Aloha EOS [@n8dee]: {
- "server_version": "1314e925",
- "chain_id": "050ca7388c8b3a394e8c5cf817600cec5f2235c31dc37a2596522e934cf988c0",
- "head_block_num": 4791,
- "last_irreversible_block_num": 4790,
- "last_irreversible_block_id": "000012b6de601b2863cf74cec413781b00b671d2ed0b5d72f721b5955656abe5",
- "head_block_id": "000012b7bd12ee04ba08fba8e2f2ac0ece9154cbfcdc8607feb1ee1aec56d63b",
- "head_block_time": "2018-06-04T16:14:37",
- "head_block_producer": "eosio",
- "virtual_block_cpu_limit": 325576,
- "virtual_block_net_limit": 126388596,
- "block_cpu_limit": 199900,
- "block_net_limit": 1048576
- }
- 04.06.2018 12:31:00, Jacky - JEDA [@newbdez33]: Sorry. mistake. the correct one is : 050ca7388c8b3a394e8c5cf817600cec5f2235c31dc37a2596522e934cf988c0
- 04.06.2018 12:31:07, Nate D - Aloha EOS [@n8dee]: oh ok, cool
- 04.06.2018 12:31:19, Nate D - Aloha EOS [@n8dee]: and we seem to have paused at 4791
- 04.06.2018 12:31:24, James Sutherland - Cypherglass.com [@JamesSutherland]: Chain Id
- 050ca7388c8b3a394e8c5cf817600cec5f2235c31dc37a2596522e934cf988c0
- 04.06.2018 12:35:59, Alexandre Bourget - EOS Canada [@abourget]: it's not paused, it's flowing though..
- 04.06.2018 12:36:00, Alexandre Bourget - EOS Canada [@abourget]: 7338
- 04.06.2018 12:36:29, Nate D - Aloha EOS [@n8dee]: I must have lost connectivity then
- 04.06.2018 12:36:55, James Sutherland - Cypherglass.com [@JamesSutherland]: Yea me too, been stuck for a while ~1000
- 04.06.2018 12:37:32, Joshua Qiu @EOSeco [@Joshuaqiu]: yes , first sycing is slow
- 04.06.2018 12:37:44, Joshua Qiu @EOSeco [@Joshuaqiu]: you can see that cpu usage is 100%
- 04.06.2018 12:39:47, Alexandre Bourget - EOS Canada [@abourget]: Aloha, you still connected ?
- 04.06.2018 12:40:05, Alexandre Bourget - EOS Canada [@abourget]: I'll stop production.. it,s not useful to move it forward anyway
- 04.06.2018 12:40:17, Nate D - Aloha EOS [@n8dee]: I'm still at 4791 height
- 04.06.2018 12:41:26, Alexandre Bourget - EOS Canada [@abourget]: Aloha I lost connection to you for a moment...
- 04.06.2018 12:41:37, Alexandre Bourget - EOS Canada [@abourget]: connection failed to peer2.test.alohaeos.com:9876: Connection refused
- 04.06.2018 12:41:46, Nate D - Aloha EOS [@n8dee]: I restarted breifly to add peers but it's back now
- 04.06.2018 12:43:11, James Sutherland - Cypherglass.com [@JamesSutherland]: 3000
- 04.06.2018 12:43:19, Jacky - JEDA [@newbdez33]: i'm on 4791
- 04.06.2018 12:43:56, Joshua Qiu @EOSeco [@Joshuaqiu]: me too
- 04.06.2018 12:46:36, James Sutherland - Cypherglass.com [@JamesSutherland]: 4791
- 04.06.2018 12:46:51, James Sutherland - Cypherglass.com [@JamesSutherland]: CPU down to idle now
- 04.06.2018 12:47:25, Joshua Qiu @EOSeco [@Joshuaqiu]: maybe stop producing
- 04.06.2018 12:49:40, Jacky - JEDA [@newbdez33]: if stop producing we can still sync to the latest blocks right?
- 04.06.2018 12:53:31, Nate D - Aloha EOS [@n8dee]: I'm at 6110 now
- 04.06.2018 12:54:47, Robert DeWilder - Worbli [@BrownDonahue]: worbli is up @ validation-1.worbli.io:33981 and syncing
- 04.06.2018 12:55:14, Robert DeWilder - Worbli [@BrownDonahue]: Are you still stuck ~1000?
- 04.06.2018 12:55:22, Robert DeWilder - Worbli [@BrownDonahue]: I am seeing the same
- 04.06.2018 12:55:28, James Sutherland - Cypherglass.com [@JamesSutherland]: I'm holding at 4791 now
- 04.06.2018 12:55:33, Joshua Qiu @EOSeco [@Joshuaqiu]: just wait
- 04.06.2018 12:55:40, Joshua Qiu @EOSeco [@Joshuaqiu]: me too
- 04.06.2018 12:55:41, Jose Toriello [@josetoriello]: Maybe pin this?
- 04.06.2018 12:56:31, Nate D - Aloha EOS [@n8dee]: Sorry, I looked at the wrong window (too many open). I'm still at 4791 as well.
- 04.06.2018 12:57:22, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Let’s keep to 13:00 UTC for call tomorrow.
- I actually propose we have a standing call at that time for BPs through launch and possibly after also.
- BP Communication is going to be critical.
- 04.06.2018 12:59:00, deck - EOS New York [@deckw]: +1
- .....04.06.2018 13:01:23, Myoungchul Park [@sps818]: >>invited J Y [@y30nj]<<
- 04.06.2018 13:15:12, mini - EOS Argentina [@moonxor]: +1
- 04.06.2018 13:18:49, Stuardo Rodríguez 🇬🇹 [@stuardo_str]: I don't understand why we don't aks everyone in the eosio-bios-boot group to join? Some people have previous chain, some people have this chain. Is there a reason to maintain the old one?
- .....04.06.2018 13:21:54, Ben Sigman [@Bensig]: >>invited David Moss [@DavidMoss]<<
- .....04.06.2018 13:22:56, Ben Sigman [@Bensig]: >>invited Thomas "Disclaimer" Cox (all opinions mine) [@thomasbcox]<<
- 04.06.2018 13:47:47, mini - EOS Argentina [@moonxor]: [[Webpage]]https://www.reddit.com/r/eos/comments/8oikln/eos_nation_weve_booted_the_mainnet_and_entered/
- 04.06.2018 13:49:11, Ville Sundell [@SolariusFI]: @stuardo_str I have understood, that the 579a649aae8f660aa1abbab262437596d1f388f709b0b94a9fd6bba479889ea5 is still the candidate for the mainnet?
- 04.06.2018 13:49:33, David P - EOS42 [@GunnisonCap]: Hmm is that technically true? 🤔
- 04.06.2018 13:50:05, Rick Schlesinger - EOS New York [@unlockrs]: seems a bit premature...
- 04.06.2018 13:50:47, nahu [@naueh]: find it disrespectful to the community, at least
- 04.06.2018 13:51:00, Timothy Lewis [@trustlayer]: I thought we had agreed to coordinate comms and work collaboratively.
- 04.06.2018 13:51:20, mini - EOS Argentina [@moonxor]: I'm not on EOS naton :)
- 04.06.2018 13:51:42, mini - EOS Argentina [@moonxor]: Just Saw that and put it here
- 04.06.2018 13:51:51, Timothy Lewis [@trustlayer]: copy
- 04.06.2018 13:52:00, mini - EOS Argentina [@moonxor]: And paste :)
- 04.06.2018 13:52:10, Timothy Lewis [@trustlayer]: Who is the EOS Nation contact?
- 04.06.2018 13:52:18, mini - EOS Argentina [@moonxor]: @DenisCarriere
- 04.06.2018 13:54:26, mini - EOS Argentina [@moonxor]: @stuardo_str Also i think
- 04.06.2018 13:54:33, Denis | EOS Nation [@DenisCarriere]: This was a miscommunication, we've removed this Tweet immediately.
- This has been copied by another user.
- It should of said => "We're testing v1.0 and are currently in the validation stage"
- 😢
- The wording was misused, the intent of EOS Nation was never to say "we've launched mainnet", we have ZERO endpoints opened to the public.
- 04.06.2018 13:55:19, nahu [@naueh]: thanks for the clarification
- 04.06.2018 13:55:27, Timothy Lewis [@trustlayer]: We agreed in validation to coordinate comms
- 04.06.2018 13:55:37, Timothy Lewis [@trustlayer]: Thank you for cleaning it up quickly.
- 04.06.2018 13:56:39, Denis | EOS Nation [@DenisCarriere]: I've been radio silence with our "social media/marketing" team, however I guess they misunderstood the message.
- Everyone does know we are still in the "testing/validation" stages, we shouldn't of mentioned the word "mainnet"
- 04.06.2018 13:57:33, Denis | EOS Nation [@DenisCarriere]: [[Photo]]On our homepage, we have "Testing in progress"
- https://eosnation.io
- 04.06.2018 13:57:48, Timothy Lewis [@trustlayer]: Crafting a voice from the combined efforts of the BPC community is important. we will release a guideline for testing expectations.
- 04.06.2018 14:01:14, Denis | EOS Nation [@DenisCarriere]: Another issue is the EOSGo countdown, we might want to remove this as well:
- http://eosgo.io/6626-2/
- 04.06.2018 14:02:32, Denis | EOS Nation [@DenisCarriere]: [[Photo]]09 hours will not be when voting happens (not to my knowledge from this group)
- 04.06.2018 14:03:59, David P - EOS42 [@GunnisonCap]: No problem and thanks for removing. Let’s keep the communication coordinated all so we don’t confuse the community inadvertently.
- 04.06.2018 14:04:07, Denis | EOS Nation [@DenisCarriere]: The more we drag our feet with this validation, the MORE incorrect messages will arise in the future.
- Let's make sure at our next call we've got some solid validation results
- .....04.06.2018 14:07:25, Timothy Lewis [@trustlayer]: >>invited Kedar Iyer | LibertyBlock [@k26dr]<<
- 04.06.2018 14:16:27, Mathieu Boulianne - EOS Canada [@mathieubo]: They now fixed the countdown on their website
- 04.06.2018 14:18:29, Denis | EOS Nation [@DenisCarriere]: Perfect! 👍we are all working together on this, one big team
- 04.06.2018 14:19:38, Thomas "Disclaimer" Cox (all opinions mine) [@thomasbcox]: Indeed!
- 04.06.2018 14:20:55, David Moss [@DavidMoss]: {{FWD: David Moss [@DavidMoss], 04.06.2018 14:16:51}}
- Hey everyone! I'm excited about the launch and have been following all the different groups working on getting the EOSIO mainnet up and running. We at least need to get a test mainnet up and running as the world is watching. At this point I'd recommend that everyone connect to the EOS Core aka Ghostbusters chain so we can at least have a working mainnet even if it is a test. They are posting chain-specific information like the genesis info in their keybase group.
- 04.06.2018 14:21:01, David Moss [@DavidMoss]: Cross-posting this.
- 04.06.2018 14:26:47, Ville Sundell [@SolariusFI]: Is the Alex chain from yesterday (579a649aae8f660aa1abbab262437596d1f388f709b0b94a9fd6bba479889ea5) is still the/a candidate for the mainnet?
- 04.06.2018 14:28:52, Craig Murray - EOS BlockSmith [@craigmurray100]: i suggest wording is "main net candidate" in all cases, or better, candidate
- 04.06.2018 14:29:10, Timothy Lewis [@trustlayer]: Absolutely agreed.
- 04.06.2018 14:29:16, Craig Murray - EOS BlockSmith [@craigmurray100]: we have a mainnet *candidate* right now, paused from about 28 hours ago, undergoing testing
- 04.06.2018 14:29:54, Craig Murray - EOS BlockSmith [@craigmurray100]: we have further created a full test version, which is NOT a candidate, and where the god account is still active. This is being used by multiple groups to test features that are only present once we hit 15%, which we cannot test on the candidate directly
- 04.06.2018 14:30:36, Craig Murray - EOS BlockSmith [@craigmurray100]: finally we are also creating a 2nd test version (not a candidate) where the unstaking timer is very small, so that it can be tested as well for unstaking after 15% without waiting 3 days to run the test
- 04.06.2018 14:31:19, Craig Murray - EOS BlockSmith [@craigmurray100]: i would also suggest we stop talking about "eos-bios" and "ghostbusters" and "core". We are the mainnet launch group
- 04.06.2018 14:31:25, Craig Murray - EOS BlockSmith [@craigmurray100]: we need to act like it
- 04.06.2018 14:33:15, Sharif Bouktila - EOS Dublin [@sharifbouktila]: +1
- 04.06.2018 14:35:18, Ville Sundell [@SolariusFI]: Thank you, this is the clarification I needed 🙂
- 04.06.2018 14:37:26, Brett Jeffery @ EOS the World [@EOStheWorld]: 👏👏👏✔️
- 04.06.2018 14:39:20, Denis | EOS Nation [@DenisCarriere]: 👏agreed, there should be no difference from the multiple launch groups, we're all going to be peering from the same chain_id. We can have different topologies, but at the end of the day, we're standing up 1 mainnet with the multiple launch groups strengthening the overall network
- 04.06.2018 14:49:59, Zhao Yu@ EOSLaoMao.com [@zhaoyu]: [[Webpage]]Guys, we are managed to test the unstaking mechanism in like 30 minutes.
- We changed the unstaking delay from 3 days to 5 minutes.
- We made a repo, to make sure that everyone can run an unstake test using your laptop.
- and of course, the unstaking functionality works pretty well.
- just give it a try: https://github.com/eoslaomao/eos-cluster-generator
- 04.06.2018 14:50:49, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Great stuff 👍
- 04.06.2018 14:52:40, Thomas "Disclaimer" Cox (all opinions mine) [@thomasbcox]: Very smart! Love it
- 04.06.2018 14:52:48, Zhao Yu@ EOSLaoMao.com [@zhaoyu]: time to bed, bye bye
- 04.06.2018 15:00:27, Rick Schlesinger - EOS New York [@unlockrs]: +1
- 04.06.2018 15:00:33, Rick Schlesinger - EOS New York [@unlockrs]: +1
- 04.06.2018 15:01:35, Jiri Jetmar | EOS Germany | [email protected] [@jirijetmar]: +1 !! Well done...
- 04.06.2018 15:04:43, deck - EOS New York [@deckw]: Super confused about what this article is trying to say? I thought we agreed to run software validation tests today and collectively communicate that out. This seems to imply that EOS Core is somehow ahead of the curve here.
- 04.06.2018 15:04:43, deck - EOS New York [@deckw]: {{FWD: Ben Sigman [@Bensig], 04.06.2018 14:24:13}}
- [[Webpage]]https://medium.com/@bensig/letter-to-eos-block-producer-candidates-184ef59a0748
- 04.06.2018 15:48:52, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Guys I've cross posted our validation plan to the EOS core channel to give teams a framework for testing.
- The objective is for multiple BPs to test and satisfy themselves we have a chain which can be proposed as Mainnet.
- I don't believe that has changed since our call earlier.
- 04.06.2018 15:53:56, deck - EOS New York [@deckw]: +1
- 04.06.2018 16:00:16, mini - EOS Argentina [@moonxor]: great
- 04.06.2018 17:51:24, nahu [@naueh]: May I ask if eos-bios and ghostbusters are still trying to rise the same blockchain?
- 04.06.2018 17:52:53, Stuardo Rodríguez 🇬🇹 [@stuardo_str]: Yes
- 04.06.2018 18:21:19, Daniel Larimer [@daniellarimer]: [[Photo]]Are we there yet?
- 04.06.2018 18:22:03, David Moss [@DavidMoss]: Getting closer...
- 04.06.2018 18:22:21, Ben Sigman [@Bensig]: I would estimate that validation is going to take a few days - we added security validation and standards are being developed.
- 04.06.2018 18:22:39, David Moss [@DavidMoss]: Ghostbusters is working on standing up a validated "test" mainnet.
- 04.06.2018 18:22:40, mini - EOS Argentina [@moonxor]: it dosent take days :), takes minutes
- 04.06.2018 18:23:26, Ben Sigman [@Bensig]: haha yes, on EOS Core/ghostbusters it's pretty straightforward
- 04.06.2018 18:23:51, mini - EOS Argentina [@moonxor]: [[Photo]]
- 04.06.2018 18:24:09, mini - EOS Argentina [@moonxor]: is the same for any chain., Tools to validate the state should be agnostic
- 04.06.2018 18:24:30, mini - EOS Argentina [@moonxor]: this one is from core, a couple of minutes ago
- 04.06.2018 18:24:46, Alexandre Bourget - EOS Canada [@abourget]: Did chain_id 579a649aae8f660aa1abbab262437596d1f388f709b0b94a9fd6bba479889ea5 validate this ?
- The chain we all have been working on since yesterday ?
- 04.06.2018 18:24:48, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: Why is stake different
- 04.06.2018 18:25:08, mini - EOS Argentina [@moonxor]: i dont know, i think they have a rounding issue while injecting.
- 04.06.2018 18:25:16, mini - EOS Argentina [@moonxor]: @deckw was the first to validate it
- 04.06.2018 18:32:04, nahu [@naueh]: I'm not sure I understood. I'd support a main testnet though
- 04.06.2018 19:19:32, Igor Lins e Silva - EOS Rio [@igorls]: @rbmarliere found it too
- 04.06.2018 19:19:48, Ricardo Marlière - EOS Brazil [@rbmarliere]: hahaha
- 04.06.2018 19:19:54, Igor Lins e Silva - EOS Rio [@igorls]: we are injecting again, real-time checks shows no error for now
- 04.06.2018 19:20:17, Ricardo Marlière - EOS Brazil [@rbmarliere]: let us know when its up
- 04.06.2018 19:31:13, Chitty 🅑 - EOS Argentina [@JChitty]: +1
- .....04.06.2018 19:31:18, Timothy Lewis [@trustlayer]: >>unsupported service message type: messageActionChatAddUsers<<
- .....04.06.2018 19:55:10, Timothy Lewis [@trustlayer]: >>invited Rafael LaVerde [@VamosRaf]<<
- 04.06.2018 20:12:19, mini - EOS Argentina [@moonxor]: it seems that the idea of coordinating the communication didnt work.
- 04.06.2018 20:14:50, Nathan Rempel | GenerEOS Sydney [@nsrempel]: I think I need a status update of where everthing is at
- 04.06.2018 20:15:10, mini - EOS Argentina [@moonxor]: to be honest me too
- 04.06.2018 20:20:25, An Byung Chan - EOSYS [@Anbyungchan]: +1
- 04.06.2018 20:25:50, Joshua Qiu @EOSeco [@Joshuaqiu]: it seem we can find guys in GB to talk?
- 04.06.2018 20:26:18, Timothy Lewis [@trustlayer]: EOS Core channel
- .....04.06.2018 21:19:35, J Y [@y30nj]: >>invited Wayne Fu [@waynefu]<<
- .....04.06.2018 21:49:35, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: >>invited Luke Stokes [@lukestokes]<<
- .....04.06.2018 21:51:03, David Moss [@DavidMoss]: >>invited Christo Boshoff [@ChristoBoshoff]<<
- 04.06.2018 22:24:39, nahu [@naueh]: I think one of of the core premises from the p2p movement is to leverage cooperation against competition. Building tools to reduce the risk of trusting eack other brings an oportunity to disrupt the feartul<==>aggresive-competition feedback loop. A ridiculously destructive social mindset I'd say
- 04.06.2018 22:28:18, nahu [@naueh]: I guess that this fear blinds the ability to identify common objectives, which are evidently more valuable and sustainable than all the individual ones together.
- Matter of fact is that this is only obtainable through a convergence of agreements and efforts.
- Isn't that what consensus is about after all?
- 04.06.2018 22:29:18, mini - EOS Argentina [@moonxor]: +1000
- 04.06.2018 22:29:43, Timothy Lewis [@trustlayer]: Absolutely.
- 04.06.2018 22:31:01, nahu [@naueh]: Demoracy as we know it, seems to me just as an attempt that has not been able to scale properly to fulfill those requirements, so far
- 04.06.2018 22:33:26, nahu [@naueh]: Random thoughts, sorry for the OT. I think my point is that we will have to commonly face some fucking big responsabilities sooner or later. I find what's going on here an unseen oportunity to try
- 04.06.2018 22:35:48, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Let’s keep this channel for BP teams to report progress on Validation.
- 04.06.2018 22:40:42, nahu [@naueh]: of course. Borught this thoughts here because I find this channel the most collaborative one right now, and want to highlight the value I find on that
- 04.06.2018 22:57:17, Ricardo Marlière - EOS Brazil [@rbmarliere]: % cleos get code eosio
- shouldn't this output match with $(sha256sum /path/to/eos/build/contracts/eosio.system/eosio.system.wasm)?
- @abourget what am I missing? this is true for eosio.msig but not for eosio.token or eosio... currently checking on chain_id 579a649aae8f660aa1abbab262437596d1f388f709b0b94a9fd6bba479889ea5 with launch-rc branch contracts...
- 04.06.2018 23:27:19, Alexandre Bourget - EOS Canada [@abourget]: [[Webpage]]that's the reference contract used:
- - name: eosio.system.wasm
- ref: /ipfs/QmVTwpwBSWEWYT1dWHhgueZf2MmjsTuMRQXygvh7VxJHUE
- comment: "~v1.0.1 from github.com/EOS-Mainnet/eos"
- From the EOS-Mainnet/eos repo, just after the B1 tag v1.0.1
- 04.06.2018 23:29:54, Alexandre Bourget - EOS Canada [@abourget]: Can you run a WAST comparison ? That would be helpful. Please report back soon.
- 04.06.2018 23:34:50, Zhao Yu@ EOSLaoMao.com [@zhaoyu]: [[Webpage]]If anyone want to test voting, staking, unstaking, refund, etc. You can use our little tool to do that, would love to help. https://github.com/eoslaomao/eos-cluster-generator
- 04.06.2018 23:34:53, Ricardo Marlière - EOS Brazil [@rbmarliere]: hold on I'll rebuild v1.0.1
- 04.06.2018 23:36:28, Alexandre Bourget - EOS Canada [@abourget]: Contracts have not changed since v1.0.0
- 04.06.2018 23:36:31, Jacky - JEDA [@newbdez33]: I will try it later. Thanks!
- 04.06.2018 23:37:49, mini - EOS Argentina [@moonxor]: awesome!
- .....04.06.2018 23:38:59, Emily Bush - shEOS [@emmybush]: >>unsupported service message type: messageActionChatAddUsers<<
- 04.06.2018 23:47:56, Alexandre Bourget - EOS Canada [@abourget]: Ok, I just checked multiple builds with clean environments. v1.0.0 from eosio/eos, v1.0.1 from eosio/eos and v1.0.1 from EOS-Mainnet/eos .. they all match the contract that is on-chain. If you have any concerns, please bring them forward.
- I can also confirm that the builds from EOS Amsterdam have the same output. https://hub.docker.com/r/eosamsterdam/eos on the tag v1.0.1
- Make sure you build with the EOS and not SYS, try with a clean build environment if needed. You'll get there :)
- 04.06.2018 23:50:09, Patrick B - Aloha EOS: 👍
- 04.06.2018 23:50:23, Stuardo Rodríguez 🇬🇹 [@stuardo_str]: So, I'm checking the document "EOS Mainnet Validation Plan", and I do not see any progress on any task. I don't see anyone flagging "verified" any task. Are we at the same place we were 2 days ago?
- 04.06.2018 23:51:06, mini - EOS Argentina [@moonxor]: no
- 04.06.2018 23:51:54, Sharif Bouktila - EOS Dublin [@sharifbouktila]: That should now be used as a guide but each BP is responsible for their own testing to whatever scope and depth the feel necessary.
- 04.06.2018 23:51:57, mini - EOS Argentina [@moonxor]: actually please read the laomao message :)
- 04.06.2018 23:51:59, Stuardo Rodríguez 🇬🇹 [@stuardo_str]: I'm worried that we start a call in ~8h and we start hearing again "we haven't reached a consensus of what to test", like we did 2 days ago, like we did 12 hours ago
- 04.06.2018 23:52:22, Timothy Lewis [@trustlayer]: Quick question... What was the workaround for the RAM cost issue?
- 04.06.2018 23:53:39, Stuardo Rodríguez 🇬🇹 [@stuardo_str]: Which message?
- 04.06.2018 23:53:53, mini - EOS Argentina [@moonxor]: this one
- 04.06.2018 23:54:05, mini - EOS Argentina [@moonxor]: and the other one where they said they test unstacking :)
- 04.06.2018 23:55:47, Stuardo Rodríguez 🇬🇹 [@stuardo_str]: That is post is for a tool to help testing, but again, there is no checkmark in the validation document.
- 04.06.2018 23:56:06, Stuardo Rodríguez 🇬🇹 [@stuardo_str]: So we will be in a call tomorrow with no progress
- 04.06.2018 23:56:09, mini - EOS Argentina [@moonxor]: you can test it if you want with that tool :)
- 04.06.2018 23:56:19, mini - EOS Argentina [@moonxor]: i trust they test it
- 04.06.2018 23:57:00, Stuardo Rodríguez 🇬🇹 [@stuardo_str]: OK, so the plan is "as nobody reported any issues, we'll asume everything is OK" which is a perfect plan IMHO
- 04.06.2018 23:57:11, Stuardo Rodríguez 🇬🇹 [@stuardo_str]: is that the plan?
- 04.06.2018 23:57:16, mini - EOS Argentina [@moonxor]: if you want to help, please help :)
- 04.06.2018 23:57:43, Stuardo Rodríguez 🇬🇹 [@stuardo_str]: I want to know if we will be stuck tomorrow as we were today
- 04.06.2018 23:58:56, Timothy Lewis [@trustlayer]: +1
- 04.06.2018 23:59:09, Sharif Bouktila - EOS Dublin [@sharifbouktila]: We will be asking BP teams to confirm that have tested and are happy , if enough do we are good.
- Decentralized QA.
- Validation plan should be used as a guide. But feel free to test more or less.
- 04.06.2018 23:59:34, mini - EOS Argentina [@moonxor]: +1 that was the idea all along
- 05.06.2018 00:00:00, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Just took us a while to get there :)
- 05.06.2018 00:00:44, David Moss [@DavidMoss]: Exactly
- 05.06.2018 00:10:04, Ricardo Marlière - EOS Brazil [@rbmarliere]: I'll drop it. Of course different platforms would produce different binaries (tested on another distro just now), thanks for your time
- 05.06.2018 00:11:02, Timothy Lewis [@trustlayer]: Quick question... What was the workaround for the RAM cost issue? @marcantoineross
- 05.06.2018 00:11:41, Timothy Lewis [@trustlayer]: @sharifbouktila ?
- 05.06.2018 00:12:00, Sharif Bouktila - EOS Dublin [@sharifbouktila]: ?
- 05.06.2018 00:12:14, Timothy Lewis [@trustlayer]: The account generation issue
- 05.06.2018 00:12:28, Timothy Lewis [@trustlayer]: with ever so slightly increasing ram costs.
- 05.06.2018 00:13:29, Sharif Bouktila - EOS Dublin [@sharifbouktila]: I think everyone was using Alex’s approach , create funds and burn any remaining.
- 05.06.2018 00:14:08, Craig Murray - EOS BlockSmith [@craigmurray100]: the way they solved it is to over-fund eosio, then give out all the acconts in the snapshot, add up how much it cost, and then you know the total. Now do it again, funding eosio from total. Pretty sure this is how it was done
- 05.06.2018 00:14:29, Timothy Lewis [@trustlayer]: Is that what was used in the genesis? Increased token supply to run and then burn?
- 05.06.2018 00:14:33, mini - EOS Argentina [@moonxor]: are you talking about the snapshot? is not an issue if you use buy ram bytes,
- 05.06.2018 00:17:06, Timothy Lewis [@trustlayer]: Could you explain this resolution?
- 05.06.2018 00:17:19, Timothy Lewis [@trustlayer]: Did we have to increase the initial supply?
- 05.06.2018 00:17:27, mini - EOS Argentina [@moonxor]: [[Webpage]]https://github.com/EOSIO/eos/blob/5d60cfbeff32576742d9a28d6278c13b524111cc/contracts/eosio.system/delegate_bandwidth.cpp#L88
- 05.06.2018 00:17:43, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: Yes the initial supply is over 1B
- 05.06.2018 00:18:23, Timothy Lewis [@trustlayer]: Was this the solution used?
- 05.06.2018 00:18:25, Zhao Yu@ EOSLaoMao.com [@zhaoyu]: in the end, there are like only 2.034 EOS extra left to eosio account
- 05.06.2018 00:18:52, Zhao Yu@ EOSLaoMao.com [@zhaoyu]: the extra EOS is used to inject the snapshot, because you have to buy ram for new accounts.
- 05.06.2018 00:19:09, Timothy Lewis [@trustlayer]: What was the initial token supply used?
- 05.06.2018 00:19:34, Timothy Lewis [@trustlayer]: Anyone recall?
- 05.06.2018 00:19:45, Zhao Yu@ EOSLaoMao.com [@zhaoyu]: we dont have to recall
- 05.06.2018 00:19:58, Zhao Yu@ EOSLaoMao.com [@zhaoyu]: you can see that in the boot sequence file
- 05.06.2018 00:20:39, Zhao Yu@ EOSLaoMao.com [@zhaoyu]: https://ipfs.io/ipfs/QmXsZX3LVnisnDoaVYCK215fQorDcnjJpx5XDaBppzeERQ
- 05.06.2018 00:21:14, Zhao Yu@ EOSLaoMao.com [@zhaoyu]: [[Photo]]
- 05.06.2018 00:21:26, Timothy Lewis [@trustlayer]: Ok just wanted to confirm.
- 05.06.2018 00:21:35, Timothy Lewis [@trustlayer]: I have a problem with this that I would like to raise.
- 05.06.2018 00:21:47, Timothy Lewis [@trustlayer]: There is an unexpected consequence.
- 05.06.2018 00:22:09, Timothy Lewis [@trustlayer]: Increasing the initial supply increases BP rewards.
- .....05.06.2018 00:22:16, Maja V [@mvparadigm]: >>invited Benjamin Vujinovic [@benjaminvu]<<
- 05.06.2018 00:22:45, Timothy Lewis [@trustlayer]: Something to keep in mind.
- 05.06.2018 00:22:59, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: by the end of boot, the supply is like 1 000 000 002 EOS
- 05.06.2018 00:25:58, Timothy Lewis [@trustlayer]: taxation is theft? 😉 I can live with that.
- 05.06.2018 00:34:40, mini - EOS Argentina [@moonxor]: [[Webpage]]yes you are right
- https://github.com/EOSIO/eos/blob/master/contracts/eosio.system/producer_pay.cpp#L84
- 05.06.2018 00:35:01, Timothy Lewis [@trustlayer]: #justsaying
- 05.06.2018 01:10:57, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Regardless of how the next few hours play out. I believe everyone here only wants to see One EOS Mainnet
- There is no winners and losers in this. Each team has contributed enormously to get us here.
- This has got to be the largest Technology Project ever delivered. It has been a privilege to be apart of it.
- It is a block producers duty to protect the interests of EOS token holders, and that duty is there before during and after launch.
- I wrote in our BP announcement Post , being a block producer is akin to the Janitor in NASA helping put a man on the moon. Certainly feels that way now.
- Lets not rush the final stages here. Take your time and lets get this right. and also lets ensure a level playing field for all teams, hopefully we will be working as partners for a long time to come.
- 05.06.2018 01:52:20, Timothy Lewis [@trustlayer]: HERE HERE!
- 05.06.2018 01:52:32, Timothy Lewis [@trustlayer]: Or HEAR HEAR!
- 05.06.2018 01:52:41, Timothy Lewis [@trustlayer]: Whatever it is, I am proud of you all.
- 05.06.2018 02:19:25, Timothy Lewis [@trustlayer]: Anyone else think we should decrease the time in between validation and check up calls?
- 05.06.2018 02:19:47, Timothy Lewis [@trustlayer]: @marcantoineross @abourget @sharifbouktila ?
- 05.06.2018 02:20:54, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Validation is 13:00 UTC , i know most teams have that in their diary.
- What check up call ?
- 05.06.2018 02:23:44, Timothy Lewis [@trustlayer]: I get that.
- 05.06.2018 02:23:54, Timothy Lewis [@trustlayer]: It would be cool to increase that to twice a day.
- 05.06.2018 02:24:09, Timothy Lewis [@trustlayer]: Just to come together a bit more often.
- 05.06.2018 02:24:28, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Sure, what ever is necessary we will do it.
- 05.06.2018 03:12:39, Crystal Rose | ⚡️SENSE | 💜shEOS [@crystalrose]: Let's call it Mainnet.one
- 05.06.2018 03:24:28, Chitty 🅑 - EOS Argentina [@JChitty]: [[Webpage]]We found an issue that doesn't allow us to validate the eos.system , we have reported on GitHub: https://github.com/EOSIO/eos/issues/3824
- 05.06.2018 03:26:28, David P - EOS42 [@GunnisonCap]: Agree with that, we should have 12hr check in calls after the 13:00 from now til mainnet launch and for the 24hrs immediately after.
- 05.06.2018 03:27:17, David P - EOS42 [@GunnisonCap]: Who owns the domain name is the big question 😉
- 05.06.2018 03:30:54, Crystal Rose | ⚡️SENSE | 💜shEOS [@crystalrose]: decentralize it. submitted .eos for gTLD / SLD consideration to ICANN just for kicks.
- 05.06.2018 03:32:24, David P - EOS42 [@GunnisonCap]: Yeah I think it’s expensive so was thinking one of the first worker proposal funds should be that, but all the better if you work some magic now 😁
- 05.06.2018 03:43:00, mini - EOS Argentina [@moonxor]: so if we want to validate a eosio.system, we cant do it automatically, we need to decode the wasm to wast and validate it manually.
- 05.06.2018 03:50:54, Zhao Yu@ EOSLaoMao.com [@zhaoyu]: maybe there’s something dynamic during building process?
- 05.06.2018 03:51:19, Zhao Yu@ EOSLaoMao.com [@zhaoyu]: not a C writer, forgive me if Im wrong
- 05.06.2018 03:52:17, WinLin-EOSBIXIN [@hiwinlin]: [[Photo]]
- 05.06.2018 03:53:18, WinLin-EOSBIXIN [@hiwinlin]: Hi all, sorry for late, I have finish the smart contract set & replace/upgrade test. It succeed and test pass
- 05.06.2018 03:53:19, mini - EOS Argentina [@moonxor]: @zhaoyu i just read C++ i dont write it. I dont really know what is going on :( . but you can find the leak information in your eosio.system.wast file inside the build folder.
- 05.06.2018 03:54:22, Timothy Lewis [@trustlayer]: 👏
- 05.06.2018 03:56:03, Jiri Jetmar | EOS Germany | [email protected] [@jirijetmar]: +1
- 05.06.2018 03:57:38, Zhao Yu@ EOSLaoMao.com [@zhaoyu]: cool
- 05.06.2018 04:04:29, Leonard Dorlöchter - EOS network by Advanced Blockchain AG [@EOSnetwork]: happy to support that. I researched the process in detail previously. It takes a while and is expensive so it probably has to be one of the first worker proposal funds like David says 🙂
- .....05.06.2018 04:26:10, Jack - AcroEOS [@gjacklee]: >>invited 경수 민 [@Ryuk0]<<
- 05.06.2018 05:31:34, Ville Sundell [@SolariusFI]: That would be cool 🙂
- 05.06.2018 06:36:22, Sharif Bouktila - EOS Dublin [@sharifbouktila]: For this issue.
- It looks like there is a fix suggested. If correct can this be applied to the chains currently?
- Who in our community can check/ merge and test this ??
- I don't believe we need to wait for Block.one and @daniellarimer to fix small issues. If our fix is correct I'm sure they would incorporate into their codebase also.
- What does everyone else think ???
- .....05.06.2018 06:37:29, Lucas @EOS Beijing [@LucasEosBeijing]: >>invited Allen@EOS Beijing [@fightinglion]<<
- .....05.06.2018 06:40:03, Rafael LaVerde [@VamosRaf]: >>invited Jeff B [@jberwick]<<
- 05.06.2018 06:40:38, Nathan Rempel | GenerEOS Sydney [@nsrempel]: Seems pretty minor, its not even a fix its just an eosio_build change
- 05.06.2018 06:44:43, Stuardo Rodríguez 🇬🇹 [@stuardo_str]: You can do a pull request to the eos-Mainet repository. That one is managed by the community, not by B1.
- 05.06.2018 06:46:14, mini - EOS Argentina [@moonxor]: You need to relaunch the chain
- 05.06.2018 06:46:39, mini - EOS Argentina [@moonxor]: Or at least. A hardfork
- 05.06.2018 06:47:16, Nathan Rempel | GenerEOS Sydney [@nsrempel]: Hey mini btw ran firestarter against bios, got thi error while validating consitution
- 05.06.2018 06:47:26, Nathan Rempel | GenerEOS Sydney [@nsrempel]: FAIL
- ##########################################################
- 'ascii' codec can't encode character u'\u2019' in position 723: ordinal not in range(128)
- Traceback (most recent call last):
- File "validator.py", line 27, in <module>
- getattr(validator,'validate')(args)
- File "/Users/nathan.rempel/firestarter/validator/vanilla_validator.py", line 606, in validate
- if not validate_system_accounts(eos_snapshot): return
- File "/Users/nathan.rempel/firestarter/validator/vanilla_validator.py", line 345, in validate_system_accounts
- f.write(current_abi["ricardian_clauses"][0]["body"])
- UnicodeEncodeError: 'ascii' codec can't encode character u'\u2019' in position 723: ordinal not in range(128)
- 05.06.2018 06:47:35, Alexandre Bourget - EOS Canada [@abourget]: The one in 579a is one agreed upon by many members of the community. It's not a blocker imo. And it can be upgraded.
- 05.06.2018 06:47:59, Nathan Rempel | GenerEOS Sydney [@nsrempel]: is the hash issue when running "get code"
- 05.06.2018 06:48:13, Nathan Rempel | GenerEOS Sydney [@nsrempel]: will return different results for different set contracts?
- 05.06.2018 06:49:12, Alexandre Bourget - EOS Canada [@abourget]: The .wasm output is crippled, but the hash displayed on screen from get code is correct.
- 05.06.2018 06:49:16, mini - EOS Argentina [@moonxor]: This Is because i think ' Is not únicode, need to check the parsing
- 05.06.2018 06:49:51, Nathan Rempel | GenerEOS Sydney [@nsrempel]: Agreed not blocker
- 05.06.2018 06:49:58, mini - EOS Argentina [@moonxor]: Is correct with the versión you build in your system
- 05.06.2018 07:01:11, Zhao Yu@ EOSLaoMao.com [@zhaoyu]: [[Webpage]]We built a tool to validate unregistered eos holders and balances: https://github.com/EOSLaoMao/eos-unregistered-validator
- by the way, the unregistered onchain data of 579a649aae8f660aa1abbab262437596d1f388f709b0b94a9fd6bba479889ea5 is VALID. 🙂
- pretty simple tool, feel free to use it.
- 05.06.2018 07:04:54, WinLin-EOSBIXIN [@hiwinlin]: 👍👍👍👍👍👍
- 05.06.2018 07:05:10, Tyee Noprom | EOSBlock.co | EOSpace.io | EOS Committer [@tyeenoprom]: Awesome!
- 05.06.2018 07:05:21, Roman CryptoLions [@romanskaskiw]: great service to the community!
- 05.06.2018 07:08:01, Zhao Yu@ EOSLaoMao.com [@zhaoyu]: All holders are created equal, even they are [unregistered], right?
- 05.06.2018 07:18:45, Tyee Noprom | EOSBlock.co | EOSpace.io | EOS Committer [@tyeenoprom]: Have we clarified how to handle the [unregistered] accounts?
- 05.06.2018 07:27:38, Nathan Rempel | GenerEOS Sydney [@nsrempel]: unreg are those not even saved by ETH key conversion correct?
- 05.06.2018 07:27:51, Nathan Rempel | GenerEOS Sydney [@nsrempel]: like 0.2% or whatever?
- 05.06.2018 07:29:48, [email protected] [@crazybits]: unregistered fund would be sent to a EOS system account for the holder to claim later
- 05.06.2018 07:29:54, Zhao Yu@ EOSLaoMao.com [@zhaoyu]: yeah less than 20 thousand i think
- 05.06.2018 07:30:18, Nathan Rempel | GenerEOS Sydney [@nsrempel]: what was the xplanation why they couldn't be saved? did those balances get sent into ETH smart contracts?
- 05.06.2018 07:31:24, Tyee Noprom | EOSBlock.co | EOSpace.io | EOS Committer [@tyeenoprom]: I know it will be sent to a EOS system account, in what way can the ETH holders claim their tokens?
- 05.06.2018 08:03:20, Alexandre Bourget - EOS Canada [@abourget]: EVERYONE:
- I propose, as we agreed yesterday, that at 1315 UTC today, the 579a chain starts accepting transactions.
- It will allow for the next stage of tests..
- 05.06.2018 08:04:25, Craig Murray - EOS BlockSmith [@craigmurray100]: 😅👍
- 05.06.2018 08:05:10, deck - EOS New York [@deckw]: {{FWD: deck - EOS New York [@deckw], 05.06.2018 08:05:01}}
- Not sure if anyone has done this yet but I ran the firestart validation tool against the EOS Core network. It looked good with only one failure because they add an extra account. I believe this is expected:
- Validating EOS blockchain snapshot against "EOS Vanilla validator (v0.1)"
- Loading EOS snapshot OK
- Loading ETH snapshot OK
- Load EOS genesis OK
- Validating genesis OK
- Verifying global params vs genesis OK
- Verifying user account creation OK
- Verifying user account stake OK
- Verifying user account code OK
- Verifying user account permission OK
- Verifying privileged accounts OK
- Verifying extra accounts FAIL
- ** > Extra account found genesisblock
- 05.06.2018 08:05:57, Jae Chung - HKEOS [@jaechung]: 👍
- 05.06.2018 08:07:21, Alexandre Bourget - EOS Canada [@abourget]: Do we want to have eosio.unregd ? It's 3M EOS /cc @nsrempel
- 05.06.2018 08:12:37, Nathan Rempel | GenerEOS Sydney [@nsrempel]: I think we must
- 05.06.2018 08:12:58, Nathan Rempel | GenerEOS Sydney [@nsrempel]: whether or not its ever recoverable, it should stay in circulation until we know what to do with it
- 05.06.2018 08:13:38, Denis | EOS Nation [@DenisCarriere]: 👍👌
- 05.06.2018 08:17:17, Nathan Rempel | GenerEOS Sydney [@nsrempel]: nathan.rempel$ cleos -u http://mainnet.genereos.io get currency balance eosio.token eosio.unregd
- 3301220.3641 EOS
- 👍
- 05.06.2018 08:18:21, Jiri Jetmar | EOS Germany | [email protected] [@jirijetmar]: 👍
- 05.06.2018 08:33:30, Joshua Qiu @EOSeco [@Joshuaqiu]: test or validation?
- 05.06.2018 08:34:40, Alexandre Bourget - EOS Canada [@abourget]: [[Webpage]]Regarding: https://github.com/EOSIO/eos/issues/3824
- It seems that some build paths sneaked into the .wasm output. All Docker builds I saw (ours, EOS Amsterdam, EOS BIXIN) probably use a default path, so gave the exact same output.
- If you have built the eosio.system contract from a different path, comparison will fail. Easiest way to compare, is run the Docker build, which can more easily reproduce the environments.
- This does not change any functionality in the contract, and we can clean-up later without risks.
- If we reset the chain, I fear we're back on validation for a few days.
- What do you think ?
- 05.06.2018 08:37:35, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Can we discuss this at 13:00 UTC - 24 minutes from now. If no risk I don't see why it can't be in a later maintenance release.
- 05.06.2018 08:37:49, Nathan Rempel | GenerEOS Sydney [@nsrempel]: Agree, if theres a comfirmed standardized build that yeilds same result, carry on
- 05.06.2018 08:38:10, Jose Toriello [@josetoriello]: +1
- 05.06.2018 08:39:54, Craig Murray - EOS BlockSmith [@craigmurray100]: 👍
- 05.06.2018 08:43:13, Daniel Larimer [@daniellarimer]: [[Photo]]
- 05.06.2018 08:44:27, Craig Murray - EOS BlockSmith [@craigmurray100]: 😂
- 05.06.2018 08:44:43, Tyee Noprom | EOSBlock.co | EOSpace.io | EOS Committer [@tyeenoprom]: 😂
- 05.06.2018 08:45:19, Jacky - JEDA [@newbdez33]: 🤣
- 05.06.2018 08:45:33, Denis | EOS Nation [@DenisCarriere]: 👍 if no risk we move it on to maintenance release
- 05.06.2018 08:45:48, Daniel Larimer [@daniellarimer]: Just merged a fix for inclusion of local build system paths
- 05.06.2018 08:46:11, David Moss [@DavidMoss]: Thanks @daniellarimer
- 05.06.2018 08:46:18, Daniel Larimer [@daniellarimer]: https://github.com/EOSIO/eos/pull/3834/files
- 05.06.2018 08:46:32, Alexandre Bourget - EOS Canada [@abourget]: Does it impact any functionality of previously compiled contract ? Is it more cosmetic ?
- 05.06.2018 08:46:40, Daniel Larimer [@daniellarimer]: more cosmetic
- 05.06.2018 08:46:42, Alexandre Bourget - EOS Canada [@abourget]: Docker builds give a deterministic output, so can be easily compared.
- 05.06.2018 08:46:47, Alexandre Bourget - EOS Canada [@abourget]: Ok, thanks.
- 05.06.2018 08:50:33, Sharif Bouktila - EOS Dublin [@sharifbouktila]: @daniellarimer fancy coming on our Validation call in 10 mins ?
- 05.06.2018 08:53:13, Ken (EOSphere) [@ken_eosphere]: 👍
- 05.06.2018 08:53:14, Thomas "Disclaimer" Cox (all opinions mine) [@thomasbcox]: What link?
- 05.06.2018 08:54:37, James Sutherland - Cypherglass.com [@JamesSutherland]: Yes link me please Seymour.
- 05.06.2018 08:54:39, Sharif Bouktila - EOS Dublin [@sharifbouktila]: [[Webpage]]https://zoom.us/j/692622137
- 05.06.2018 08:55:13, James Sutherland - Cypherglass.com [@JamesSutherland]: Those EOS Dublin guys, always on the ball! I like it!
- .....05.06.2018 09:32:38, lo chaiching [@lochaiching]: >>invited Yvonne (EOShenzhen) [@yvonnekama]<<
- 05.06.2018 09:33:50, Jae Chung - HKEOS [@jaechung]: [[Photo]]
- 05.06.2018 09:34:02, Jae Chung - HKEOS [@jaechung]: as requested in the Zoom call
- 05.06.2018 09:34:49, Stuardo Rodríguez 🇬🇹 [@stuardo_str]: the orange box is the "you are here" in the map?
- 05.06.2018 09:37:39, Thomas "Disclaimer" Cox (all opinions mine) [@thomasbcox]: Is there yet a published protocol for coordinating and communicating tests, e.g. who is going to run which test, and what results they got? If not I’ll help create one; if there is I’ll help share it.
- 05.06.2018 09:40:51, Michael Yeates [@michaelyeates]: all tests should be scripted and put in a collective github
- 05.06.2018 09:41:10, Michael Yeates [@michaelyeates]: [[Webpage]]theres a skeleton test runner here https://github.com/michaeljyeates/ghostbustertests
- 05.06.2018 09:41:34, Michael Yeates [@michaelyeates]: we will start adding tests today which will be runnable by anyone
- 05.06.2018 09:42:56, Thomas "Disclaimer" Cox (all opinions mine) [@thomasbcox]: Good, I see the spreadsheet. I don’t see a protocol (list of idiot-proof steps to follow) to ID which test I will perform, register myself as the one testing, perform the test, save my test script into a repo, and log my results.
- 05.06.2018 09:44:00, Thomas "Disclaimer" Cox (all opinions mine) [@thomasbcox]: If there's any agreement that a protocol woudl be useful, I'll convene a small group to tackle this, in 90 min from now
- 05.06.2018 09:47:33, Jose Toriello [@josetoriello]: 👍
- 05.06.2018 09:47:47, Michael Yeates [@michaelyeates]: i think it would be useful to coordinate test writers so that we make sure we have full coverage
- 05.06.2018 09:48:53, Craig Murray - EOS BlockSmith [@craigmurray100]: @DavidMoss Thank you for joining us this morning, and for speaking up. That was extremely helpful
- 05.06.2018 09:49:04, Sam - EOS Dublin [@samnoble]: And to make sure we're all absoultely certain we're testing the same version of nodeos
- 05.06.2018 09:49:38, Jiri Jetmar | EOS Germany | [email protected] [@jirijetmar]: same here..
- 05.06.2018 09:50:23, Nathan Rempel | GenerEOS Sydney [@nsrempel]: Can the "accelerated validation tool suite" be revised to use to use existing nodes rather than docker cluster (i assume yes) and drop the unstake test. Is it comprehensive enough that people can agree to run it individually against the cloned candidate net fullnodes
- .....05.06.2018 09:50:41, Benjamin Wang | EOShenzhen [@iBenjaminWang]: >>invited Sheldon | EOShenzhen [@SheldonXP]<<
- 05.06.2018 09:50:47, Nathan Rempel | GenerEOS Sydney [@nsrempel]: how many of the proposed acceptance tests will it cover?
- 05.06.2018 09:50:53, Ville Sundell [@SolariusFI]: Thanks everyone for the call!
- 05.06.2018 09:51:18, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: The public endpoints for each chain should be all in one place too
- 05.06.2018 09:51:27, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: So people can run tests through all chains
- 05.06.2018 09:52:08, Nathan Rempel | GenerEOS Sydney [@nsrempel]: 👍
- 05.06.2018 09:52:11, Jae Chung - HKEOS [@jaechung]: {{FWD: Michael Yeates [@michaelyeates], 05.06.2018 08:47:10}}
- Public endpoints are now open, please use this genesis file{
- "initial_timestamp": "2018-06-04T00:00:00.001",
- "initial_key": "EOS7NjFmKkHsCTdT56RihFzhBVqdBhp9m2W6KpTKAFcVdvd18VKAE",
- "initial_configuration": {
- "max_block_net_usage": 1048576,
- "target_block_net_usage_pct": 1000,
- "max_transaction_net_usage": 524288,
- "base_per_transaction_net_usage": 12,
- "net_usage_leeway": 500,
- "context_free_discount_net_usage_num": 20,
- "context_free_discount_net_usage_den": 100,
- "max_block_cpu_usage": 200000,
- "target_block_cpu_usage_pct": 1000,
- "max_transaction_cpu_usage": 150000,
- "min_transaction_cpu_usage": 100,
- "max_transaction_lifetime": 3600,
- "deferred_trx_expiration_window": 600,
- "max_transaction_delay": 3888000,
- "max_inline_action_size": 4096,
- "max_inline_action_depth": 4,
- "max_authority_depth": 6
- }
- }
- 05.06.2018 09:52:15, Jae Chung - HKEOS [@jaechung]: {{FWD: Michael Yeates [@michaelyeates], 05.06.2018 08:47:50}}
- These are the endpoints
- nodetest.eosio.africa:9886
- nodetest0.eosio.africa:9895
- nodetest1.eosio.africa:9895
- 40.85.223.230:9875
- 40.85.223.230:9874
- 40.85.223.230:9873
- 05.06.2018 09:52:25, Sharif Bouktila - EOS Dublin [@sharifbouktila]: [[Webpage]]Next Validation Call - 18:00 UTC 5th June 2018
- Zoom Link - https://zoom.us/j/911102470
- Sheet is here: https://docs.google.com/spreadsheets/d/1t4p1tWs0fXf28nF2w4m5wwzO1cAknYs5XQGJLAIf1pU/edit#gid=0
- .....05.06.2018 09:52:29, Sharif Bouktila - EOS Dublin [@sharifbouktila]: >>unsupported service message type: messageActionPinMessage<<
- 05.06.2018 09:52:39, Jae Chung - HKEOS [@jaechung]: Here is the genesis and endpoints for the chain open for public validation
- 05.06.2018 09:53:13, Nathan Rempel | GenerEOS Sydney [@nsrempel]: Lets be explicit about which chain is which when sharing these? unless we're testing how the fork functionality works?
- 05.06.2018 09:53:16, Ross (EOSphere) [@rossco99]: Is this an accelerated chain for testing running contracts etc.?
- 05.06.2018 09:53:21, Nathan Rempel | GenerEOS Sydney [@nsrempel]: i guess genesis will keep them seperate
- 05.06.2018 09:53:56, Craig Murray - EOS BlockSmith [@craigmurray100]: p2p-peer-address = peer.eosn.io:9876
- p2p-peer-address = 91.121.177.21:9876
- p2p-peer-address = node1.eosgreen.io:19876
- p2p-peer-address = boot.eostitan.com:9876
- p2p-peer-address = mainnet.eoseco.com:10010
- p2p-peer-address = m.jeda.one:3322
- p2p-peer-address = p2p-1.eosnetwork.io:9876
- p2p-peer-address = 23.22.71.127:9876
- p2p-peer-address = abp-node.eoscalgary.io:5222
- p2p-peer-address = eosbp.eosvillage.io:8181
- p2p-peer-address = p2p-seoul.mainnet.eosnodeone.io:9976
- p2p-peer-address = 185.159.128.146:9876
- p2p-peer-address = 54.95.202.14:9876
- p2p-peer-address = 35.234.53.243:9876
- p2p-peer-address = eos.nodepacific.com:443
- p2p-peer-address = fullnode.eoslaomao.com:443
- p2p-peer-address = nl.eu-west.eosamsterdam.net:9867
- p2p-peer-address = node2.blockeos.io:9987
- p2p-peer-address = 18.188.38.75:9876
- p2p-peer-address = 18.188.4.97:9876
- p2p-peer-address = mainnet-prod.eossv.org:443
- p2p-peer-address = 52.9.143.150:9876
- p2p-peer-address = p2p.genereos.io:9876
- p2p-peer-address = rvk1.sheos.org:1876
- p2p-peer-address = 18.234.6.119:80
- p2p-peer-address = 13.71.188.181
- p2p-peer-address = 130.211.59.178:9876
- p2p-peer-address = node1.eosphere.io:9876
- p2p-peer-address = node2.eosphere.io:9876
- p2p-peer-address = mainnet.eoseco.com:10010
- p2p-peer-address = bp.superone.io:443
- p2p-peer-address = 52.79.113.58:9876
- p2p-peer-address = 167.114.173.19:9996
- p2p-peer-address = peer.eosio.sg:9876
- p2p-peer-address = new.eoshenzhen.io:10010
- p2p-peer-address = p2p.meet.one:9876
- p2p-peer-address = peer1.eosthu.com
- p2p-peer-address = 40.67.252.236:9999
- p2p-peer-address = 35.197.190.234:19878
- p2p-peer-address = eosapi.blockmatrix.network:13546
- p2p-peer-address = publicnode.cypherglass.com:9876
- This is a peer list for *9ea5 chain_id, the one Canada spun up 48 hours ago
- 05.06.2018 09:54:05, Jae Chung - HKEOS [@jaechung]: yes genesis will keep them separate. It doesn't matter which chain you're testing really they are all equal candidates right?
- 05.06.2018 09:54:44, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: @abourget have you cloned this chain already to let transactions run through?
- 05.06.2018 09:54:49, Craig Murray - EOS BlockSmith [@craigmurray100]: if they have different genesis you should test both, no?
- 05.06.2018 09:54:52, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: or will transactions only run through on EOS Core
- 05.06.2018 09:55:02, Nathan Rempel | GenerEOS Sydney [@nsrempel]: correct, but i think people would rather know they aren't peered with chain a and running api tests on chain b? or that doesn't matter either?
- 05.06.2018 09:55:33, Jae Chung - HKEOS [@jaechung]: that's what I'm saying. It doesn't matter which chain your testing, we don't need to give them names. Just take the genesis and endpoints and test these candidates and let's get through it
- 05.06.2018 09:55:37, Craig Murray - EOS BlockSmith [@craigmurray100]: yes people should know who they are peered to, and who they are testing, 100%
- 05.06.2018 09:56:01, Craig Murray - EOS BlockSmith [@craigmurray100]: yeah, so *0ea5 is one candidate. I gave a peer list for that
- 05.06.2018 09:56:05, Sam - EOS Dublin [@samnoble]: Can I propose we create a tab for each of the two candidate nets in the spreadsheet. Including the genesis and a list of peers? @sharifbouktila can you update the pinned message to include a link to the spreadsheet again please.
- 05.06.2018 09:56:07, Jae Chung - HKEOS [@jaechung]: yep
- 05.06.2018 09:56:15, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: BIOS Chain ID: 579a649aae8f660aa1abbab262437596d1f388f709b0b94a9fd6bba479889ea5
- EOS Core Chain ID: 0d6c11e66db1ea0668d630330aaee689aa6aa156a27d39419b64b5ad81c0a760
- 05.06.2018 09:56:28, Nathan Rempel | GenerEOS Sydney [@nsrempel]: 👍
- 05.06.2018 09:56:56, James Sutherland - Cypherglass.com [@JamesSutherland]: Could you post the genesis for this?
- 05.06.2018 09:57:15, Sam - EOS Dublin [@samnoble]: @craigmurray100 and add it to the spreadsheet please
- 05.06.2018 09:57:20, Sam - EOS Dublin [@samnoble]: @jaechung can you do the same for EOS CORE?
- 05.06.2018 09:57:47, Jae Chung - HKEOS [@jaechung]: sure, could you link us the spreadsheet again?
- 05.06.2018 09:58:09, Sharif Bouktila - EOS Dublin [@sharifbouktila]: spreadsheet pinned also
- 05.06.2018 09:58:11, Sam - EOS Dublin [@samnoble]: [[Webpage]]https://docs.google.com/spreadsheets/d/1t4p1tWs0fXf28nF2w4m5wwzO1cAknYs5XQGJLAIf1pU/edit#gid=0
- 05.06.2018 09:58:16, Sam - EOS Dublin [@samnoble]: Thanks @sharifbouktila
- 05.06.2018 09:58:21, Nathan Rempel | GenerEOS Sydney [@nsrempel]: I like Bios and Core, can get really pedantic and say CandidateNetB(ios) and CandidateNetC(ore)
- 05.06.2018 09:58:44, Sharif Bouktila - EOS Dublin [@sharifbouktila]: B Chain / C Chain
- 05.06.2018 09:58:49, Nathan Rempel | GenerEOS Sydney [@nsrempel]: ^^
- 05.06.2018 09:58:49, Sam - EOS Dublin [@samnoble]: +1
- 05.06.2018 09:59:19, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: I like this differentiator
- 05.06.2018 09:59:21, Sam - EOS Dublin [@samnoble]: We could go all out and use UUIDs 😉
- 05.06.2018 10:00:03, Tim Weston - GenerEOS (BP Sydney) [@timweston]: I agree that ssteps like this need to be added to the spreadsheet so it can be quantified
- 05.06.2018 10:01:15, Sam - EOS Dublin [@samnoble]: And are we all in agreement that we're running nodeos version 1603685708?
- nodeos --version
- 1603685708
- 05.06.2018 10:01:32, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: +1
- 05.06.2018 10:01:33, Tim Weston - GenerEOS (BP Sydney) [@timweston]: We dont want to have the testing being lost along the way otherwise we will go around in cirlces
- 05.06.2018 10:01:37, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: Running 1603685708
- 05.06.2018 10:01:43, James Sutherland - Cypherglass.com [@JamesSutherland]: We should add that to the spreadsheet, nodeOS version
- 05.06.2018 10:01:54, Thomas "Disclaimer" Cox (all opinions mine) [@thomasbcox]: care to help craft the testing protocol?
- 05.06.2018 10:01:56, James Sutherland - Cypherglass.com [@JamesSutherland]: For each chain
- 05.06.2018 10:02:02, Thomas "Disclaimer" Cox (all opinions mine) [@thomasbcox]: Who else has the passion?
- 05.06.2018 10:02:27, Nathan Rempel | GenerEOS Sydney [@nsrempel]: We also agreed then both Chain meshes will stay active? A rep from each will clone and pause/shutdown the chain in a state that can be restarted, remeshed, resynced? Should that backup be loaded into repo?
- 05.06.2018 10:02:52, Nathan Rempel | GenerEOS Sydney [@nsrempel]: i.e. the active chain meshes will go "beta" and start transaction testing, then get reverted
- 05.06.2018 10:02:59, Tim Weston - GenerEOS (BP Sydney) [@timweston]: Can help for a bit but its getting late over in Australia 😳
- 05.06.2018 10:03:10, Sam - EOS Dublin [@samnoble]: I've added a 'nodeos' sheet with the current version in. We can update that if we need to switch to a new version
- 05.06.2018 10:04:55, Nathan Rempel | GenerEOS Sydney [@nsrempel]: which git tag/version is needed for 1603685708
- 05.06.2018 10:06:17, deck - EOS New York [@deckw]: @sharifbouktila Can I request that 1 rep from each team is present. Having multiples can exclude other teams if we only have 100 slots
- 05.06.2018 10:07:32, Jose Toriello [@josetoriello]: I can help coordinate
- 05.06.2018 10:08:46, Joshua Qiu @EOSeco [@Joshuaqiu]: snapshot you use has md5 e9e3dd52360efb13e6ba0fd0070dcefb?
- 05.06.2018 10:09:10, Nathan Rempel | GenerEOS Sydney [@nsrempel]: Ive added API lists, if anyone wants to test running traffic against other node http apis
- 05.06.2018 10:09:33, Nathan Rempel | GenerEOS Sydney [@nsrempel]: or if we share with non BP groups to run tests
- 05.06.2018 10:10:14, Nathan Rempel | GenerEOS Sydney [@nsrempel]: I dont have a list of these, so please self populate if you want to share
- 05.06.2018 10:10:15, Ren Kaidi - EOSWTZ [@FoxMr]: which git tag/version is needed for 1603685708 ?
- 05.06.2018 10:10:22, Sharif Bouktila - EOS Dublin [@sharifbouktila]: +1
- 05.06.2018 10:10:27, James Sutherland - Cypherglass.com [@JamesSutherland]: +1
- 05.06.2018 10:13:06, James Sutherland - Cypherglass.com [@JamesSutherland]: I tried building launch-rc and it wans't that version
- 05.06.2018 10:13:46, Sam - EOS Dublin [@samnoble]: [[Webpage]]https://github.com/eosdublin/nodeos-template/tree/eos-mainnet/Ubuntu_16_04_LTS/launch-rc-5f964d4
- 05.06.2018 10:14:05, Sam - EOS Dublin [@samnoble]: Commit 5f964d4cef7efaa7b76221c33616810afa8332e8 from the EOS Mainnet repo gives nodeos -v = 1603685708
- 05.06.2018 10:14:26, deck - EOS New York [@deckw]: I think we should look to use launch-1.0.1
- 05.06.2018 10:17:40, James Sutherland - Cypherglass.com [@JamesSutherland]: Confirmed commit 5f964d4cef7efaa7b76221c33616810afa8332e8 = 1603685708
- 05.06.2018 10:17:50, Alexandre Bourget - EOS Canada [@abourget]: lol :)
- 05.06.2018 10:17:58, Alexandre Bourget - EOS Canada [@abourget]: you switch to Dvorak mid-line ? :)
- 05.06.2018 10:18:10, James Sutherland - Cypherglass.com [@JamesSutherland]: LMAO! Almost
- 05.06.2018 10:20:10, Sam - EOS Dublin [@samnoble]: [[Webpage]]Ok, here's the diff https://github.com/EOS-Mainnet/eos/compare/launch-1.0.1...EOS-Mainnet:launch-rc
- Does anyone object to upgrading to the launch-1.0.1 tag? https://github.com/EOS-Mainnet/eos/tree/launch-1.0.1
- .....05.06.2018 10:22:53, Igor Lins e Silva - EOS Rio [@igorls]: >>invited Domi - EOS Rio [@domiscd]<<
- 05.06.2018 10:40:12, WinLin-EOSBIXIN [@hiwinlin]: @jaechung please tell what accounts has been created in the net?
- 05.06.2018 10:40:33, WinLin-EOSBIXIN [@hiwinlin]: I default with accounts:
- ['eosio','eosio.msig','eosio.token','eosio.burned','eosio.names',
- 'eosio.saving','eosio.bpay','eosio.vpay','eosio.disco',
- 'eosio.ramfee','eosio.ram','eosio.unregd','eosio.stake']
- 05.06.2018 10:40:35, Jae Chung - HKEOS [@jaechung]: one account: genesisblock
- 05.06.2018 10:40:52, WinLin-EOSBIXIN [@hiwinlin]: does that lose someone ?
- 05.06.2018 10:40:52, Jae Chung - HKEOS [@jaechung]: no eosio.unregd
- 05.06.2018 10:41:17, Jae Chung - HKEOS [@jaechung]: no eosio.disco
- 05.06.2018 10:41:26, WinLin-EOSBIXIN [@hiwinlin]: it's ok, the redundant account will make no harm
- 05.06.2018 10:41:44, Jae Chung - HKEOS [@jaechung]: okay
- 05.06.2018 10:41:51, WinLin-EOSBIXIN [@hiwinlin]: what is the use of genesisblock account ?
- 05.06.2018 10:42:02, Jae Chung - HKEOS [@jaechung]: the one ABP
- 05.06.2018 10:42:11, Jae Chung - HKEOS [@jaechung]: eosio —> genesisblock produces
- 05.06.2018 10:42:16, WinLin-EOSBIXIN [@hiwinlin]: ok
- 05.06.2018 10:42:58, WinLin-EOSBIXIN [@hiwinlin]: I will adjust my validation tool to support eos-bios and ghostbuster both
- 05.06.2018 10:43:20, Zhao Yu@ EOSLaoMao.com [@zhaoyu]: is there a detailed boot sequence? @jaechung
- 05.06.2018 10:43:37, Zhao Yu@ EOSLaoMao.com [@zhaoyu]: would like to run some validation test
- 05.06.2018 10:45:03, Joshua Qiu @EOSeco [@Joshuaqiu]: you dont do any thing with snapshot_unregistered.csv?
- 05.06.2018 10:45:09, WinLin-EOSBIXIN [@hiwinlin]: [[Webpage]]@jaechung I will use snapshot file: https://github.com/eosauthority/genesis/blob/master/snapshot-files/final/2/snapshot.csv
- 05.06.2018 10:49:09, Joshua Qiu @EOSeco [@Joshuaqiu]: snapshot is all right ,pass my simple validation , but just let's unreg balance gone?
- 05.06.2018 10:49:30, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: unreg is not deployed yet I believe
- 05.06.2018 10:49:44, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: @michaelyeates
- 05.06.2018 10:50:25, Jae Chung - HKEOS [@jaechung]: [[Webpage]]https://github.com/CryptoLions/EOS-Boot-Steps-dawn4
- A variation of this, we changed a lot of stuff and ran bash scripts
- 05.06.2018 10:50:40, Jae Chung - HKEOS [@jaechung]: @igorls knows more about that
- 05.06.2018 10:51:04, Jae Chung - HKEOS [@jaechung]: sure thanks
- 05.06.2018 10:51:09, Charles H - EOS 42 [@ankh2054]: @Joshuaqiu @syed_jafri contract will be implemented after community approval.
- 05.06.2018 10:51:30, Michael Yeates [@michaelyeates]: we are waiting for community discussions on this contract but dont see a reason why we couldnt load the contract and populate it
- 05.06.2018 10:52:17, Joshua Qiu @EOSeco [@Joshuaqiu]: which account hold these 3M EOS?
- 05.06.2018 10:52:35, Stuardo Rodríguez 🇬🇹 [@stuardo_str]: p2p-peer-address = node.eosmeso.io:9876 missing in that list :)
- 05.06.2018 10:53:09, Michael Yeates [@michaelyeates]: who has some software acceptance testing scripts?
- 05.06.2018 10:54:08, Igor Lins e Silva - EOS Rio [@igorls]: eosio still has them, they can be then transfered after approval using multisig
- 05.06.2018 10:54:15, Marc-Antoine Ross - EOS Canada [@marcantoineross]: @zhaoyu does
- 05.06.2018 10:55:04, Michael Yeates [@michaelyeates]: github link?
- 05.06.2018 10:55:06, Zhao Yu@ EOSLaoMao.com [@zhaoyu]: [[Webpage]]https://github.com/eoslaomao/eos-cluster-generator
- 05.06.2018 10:55:56, Joshua Qiu @EOSeco [@Joshuaqiu]: eosio will not resign?and pulish private key?
- 05.06.2018 10:56:08, Zhao Yu@ EOSLaoMao.com [@zhaoyu]: it helps to setup a multp-bp local testnet quickly, and generate a bunch of scripts to test voting/staking/unstaking/ etc
- 05.06.2018 10:56:34, Zhao Yu@ EOSLaoMao.com [@zhaoyu]: once it’s booted up, kinda up to you to apply any test cases
- 05.06.2018 10:58:01, Michael Yeates [@michaelyeates]: great, can we take the produced scripts to run against an existing network?
- 05.06.2018 11:01:46, Zhao Yu@ EOSLaoMao.com [@zhaoyu]: if you run your node with docker, it’s would be just change a few lines of code
- 05.06.2018 11:07:36, Thomas "Disclaimer" Cox (all opinions mine) [@thomasbcox]: Super. Testing protocol breakout group will convene in 23 minutes, at 1130 AM EST / 1530 UTC
- 05.06.2018 11:08:09, Michael Yeates [@michaelyeates]: do you have a link to the code?
- 05.06.2018 11:09:32, Zhao Yu@ EOSLaoMao.com [@zhaoyu]: do you guys have any thoughts about the cases that, EOS holders that missed the registration, also have no outgoing transcation from their ETH wallet, so the fallback mechanism couldnt cover that, either.
- 05.06.2018 11:09:45, Zhao Yu@ EOSLaoMao.com [@zhaoyu]: endup with no trace in the final snapshot
- 05.06.2018 11:10:22, Thomas "Disclaimer" Cox (all opinions mine) [@thomasbcox]: I had heard that all 1 billion tokens were accounted for, which implies that your concern is misplaced.
- 05.06.2018 11:10:36, Thomas "Disclaimer" Cox (all opinions mine) [@thomasbcox]: Best to bring that up in the Snapshot group. Or ask Sandwich directly.
- 05.06.2018 11:11:17, David Moss [@DavidMoss]: All but .1337 of one EOS token has been accounted for...
- 05.06.2018 11:11:28, Thomas "Disclaimer" Cox (all opinions mine) [@thomasbcox]: Join here
- https://t.me/joinchat/GgxZkRD5IzcLY8xNlu8ksg
- 05.06.2018 11:11:38, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: Need to store their ETH public key as a memo for easier claiming later imo
- 05.06.2018 11:12:17, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: He means the content in the unregd account
- 05.06.2018 11:12:26, Thomas "Disclaimer" Cox (all opinions mine) [@thomasbcox]: I stand corrected, 999,999,999.8663 tokens are accounted for, and 0.1337 of a token is unaccounted for.
- 05.06.2018 11:12:54, Roman CryptoLions [@romanskaskiw]: 1337 😊
- 05.06.2018 11:14:33, Roshan - EOS Authority [@telepound]: [[Webpage]]Yes unregd account holds holds 0.3% or 3,301,220.3641 EOS is on a smart contract and not in the user account.
- https://eosauthority.com/genesis
- 05.06.2018 11:15:10, Michael Yeates [@michaelyeates]: do we have an agreement on where the ram fee should come from?
- 05.06.2018 11:15:59, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: Issue a higher initial supply
- 05.06.2018 11:16:41, Zhao Yu@ EOSLaoMao.com [@zhaoyu]: that’s what eos-bios do, with like 2.124 EOS extra token left to eosio acccount
- 05.06.2018 11:16:53, Michael Yeates [@michaelyeates]: that will affect the pay to producers wouldnt it?
- 05.06.2018 11:17:04, Michael Yeates [@michaelyeates]: ie producers would be paid more
- 05.06.2018 11:17:43, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: By the end of boot, there would be a supply of 1 000 000 002 EOS, very close to 1 billion
- 05.06.2018 11:17:53, Michael Yeates [@michaelyeates]: have we checked that?
- 05.06.2018 11:17:59, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: Yes, in BIOS
- 05.06.2018 11:18:01, Craig Murray - EOS BlockSmith [@craigmurray100]: yes
- 05.06.2018 11:18:03, Michael Yeates [@michaelyeates]: the supply was 1bn when i checked
- 05.06.2018 11:18:04, Craig Murray - EOS BlockSmith [@craigmurray100]: with multiple tools
- 05.06.2018 11:19:54, Denis | EOS Nation [@DenisCarriere]: How can someone fully validate/test the boot sequence if changes are still being made to the boot sequence? This type of information should be public, this is the reason why I would not trust any chain that is not being launched with eos-bios
- What type of changes are still being added?
- 05.06.2018 11:20:31, Michael Yeates [@michaelyeates]: we agreed that state was the important thing to validate
- 05.06.2018 11:21:28, Craig Murray - EOS BlockSmith [@craigmurray100]: hrm, actually
- 05.06.2018 11:21:36, Craig Murray - EOS BlockSmith [@craigmurray100]: ./cleos.sh get currency stats eosio.token "EOS"
- {
- "EOS": {
- "supply": "1000011821.0000 EOS",
- "max_supply": "10000000000.0000 EOS",
- "issuer": "eosio"
- }
- }
- 05.06.2018 11:22:51, Michael Yeates [@michaelyeates]: thats a bit more than 2 eos
- 05.06.2018 11:23:08, Michael Yeates [@michaelyeates]: will lead to excessive inflation
- 05.06.2018 11:24:48, Zhao Yu@ EOSLaoMao.com [@zhaoyu]: maybe there should be an action log generated after launch in this way. bios-boot does a great job doing this. after all, it’s just a bunch of data, but we need to be verifying that really easily.
- 05.06.2018 11:24:51, Denis | EOS Nation [@DenisCarriere]: Both state and boot sequence is important, why would hidding the boot sequence benefit anyone?
- 05.06.2018 11:25:18, Michael Yeates [@michaelyeates]: nobody is hiding
- 05.06.2018 11:25:26, Michael Yeates [@michaelyeates]: read the actions from the chain
- 05.06.2018 11:25:38, Zhao Yu@ EOSLaoMao.com [@zhaoyu]: after all accounts created, it will have 2 extra EOS left, which is not a big deal I think?
- 05.06.2018 11:25:59, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: Is this on 579a649aae8f660aa1abbab262437596d1f388f709b0b94a9fd6bba479889ea5?
- 05.06.2018 11:26:02, Michael Yeates [@michaelyeates]: we are worried about total supply affecting block producer pay
- 05.06.2018 11:26:13, Zhao Yu@ EOSLaoMao.com [@zhaoyu]: lets test that
- 05.06.2018 11:26:22, Michael Yeates [@michaelyeates]: people may think we added it so that we would benefit
- 05.06.2018 11:27:03, Nathan Rempel | GenerEOS Sydney [@nsrempel]: What number do you get on latest? I'm running v1.0.1
- 05.06.2018 11:27:12, Denis | EOS Nation [@DenisCarriere]: Only 2 EOS remains in eosio
- $ cleos get currency balance eosio.token eosio EOS
- 2.0312 EOS
- Maybe @abourget can expand
- 05.06.2018 11:27:30, Craig Murray - EOS BlockSmith [@craigmurray100]: yes.
- 05.06.2018 11:27:52, Michael Yeates [@michaelyeates]: its the supply that is being questioned, not eosio balance
- 05.06.2018 11:27:54, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: Why is there 11000 EOS excess then, the accounts are already created
- 05.06.2018 11:27:55, Thomas "Disclaimer" Cox (all opinions mine) [@thomasbcox]: starting in 3 min
- 05.06.2018 11:28:11, Thomas "Disclaimer" Cox (all opinions mine) [@thomasbcox]: (Testing protocol workgroup)
- 05.06.2018 11:29:46, Zhao Yu@ EOSLaoMao.com [@zhaoyu]: you have to buy memery when u create new accounts, the extra 11000 EOS is used to pay that
- 05.06.2018 11:29:49, Denis | EOS Nation [@DenisCarriere]: Is it from the creation of accounts?
- 05.06.2018 11:30:20, Denis | EOS Nation [@DenisCarriere]: 👍Exactly what I thought, creating all those accounts cost RAM and eosio pays for that
- 05.06.2018 11:31:17, Zhao Yu@ EOSLaoMao.com [@zhaoyu]: so that users will have exact amount of EOS as they registered.
- 05.06.2018 11:31:23, Craig Murray - EOS BlockSmith [@craigmurray100]: yes, it has all been accounted for over the last 48 hours
- 05.06.2018 11:31:46, Denis | EOS Nation [@DenisCarriere]: eosio purchased 38.84 MB worth of RAM
- $ cleos get account eosio
- 05.06.2018 11:31:57, WinLin-EOSBIXIN [@hiwinlin]: [[Webpage]]@all I have finished the validation of Ghosterbuster testnet, you can download the lastest eos_onchain_validator.py from https://github.com/EOSBIXIN/eostoolkit/tree/master/eos_validator and to check by yourself
- 05.06.2018 11:32:14, Zhao Yu@ EOSLaoMao.com [@zhaoyu]: great
- 05.06.2018 11:32:46, Zhao Yu@ EOSLaoMao.com [@zhaoyu]: have you run that
- 05.06.2018 11:32:47, Zhao Yu@ EOSLaoMao.com [@zhaoyu]: ?
- 05.06.2018 11:34:17, WinLin-EOSBIXIN [@hiwinlin]: he validation result is :
- snapshot compare、account、balance validation is succeed.
- But the codehash of eosio compare filed, please check @jaechung
- 05.06.2018 11:34:29, Zhao Yu@ EOSLaoMao.com [@zhaoyu]: I do think the proposal that we use bios-boot to inject all the actions, then hand it over to GB to run an ABP node producing blocks, is a smart one.
- 05.06.2018 11:35:31, Zhao Yu@ EOSLaoMao.com [@zhaoyu]: since bios-boot does a great job executing these actions and with proper logging of these actions. Also, they are plenty of tools to validate that, really fast.
- 05.06.2018 11:35:51, WinLin-EOSBIXIN [@hiwinlin]: [[Photo]]
- 05.06.2018 11:35:59, WinLin-EOSBIXIN [@hiwinlin]: [[Photo]]
- 05.06.2018 11:36:26, Zhao Yu@ EOSLaoMao.com [@zhaoyu]: multisig contract error?
- 05.06.2018 11:36:36, WinLin-EOSBIXIN [@hiwinlin]: the eosio.system and eosio.bios contracts have some differences
- 05.06.2018 11:37:00, Nathan Rempel | GenerEOS Sydney [@nsrempel]: is this related to the sha256 thing that was fixed/described?
- 05.06.2018 11:37:05, WinLin-EOSBIXIN [@hiwinlin]: eosio.system and eosio.bios contracts has error
- 05.06.2018 11:37:23, Zhao Yu@ EOSLaoMao.com [@zhaoyu]: that only affects system contract, no?
- 05.06.2018 11:37:54, Nathan Rempel | GenerEOS Sydney [@nsrempel]: i think so
- 05.06.2018 11:39:41, WinLin-EOSBIXIN [@hiwinlin]: [[Webpage]]that maybe just because of compile issue casue by https://github.com/EOSIO/eos/issues/3824
- 05.06.2018 11:40:30, WinLin-EOSBIXIN [@hiwinlin]: but some prove should be supplied else I think there should be something wrong in the chain @jaechung
- 05.06.2018 11:42:22, Ken (EOSphere) [@ken_eosphere]: [[Webpage]]Some snapshot injection validation results for registered accounts against the two chains - https://github.com/eosphere/candidate-chain-snapshot-verification thanks to @deckw tool eospy - .Total accounts: 163930.
- 05.06.2018 11:42:24, Igor Lins e Silva - EOS Rio [@igorls]: multisig is working fine
- 05.06.2018 11:42:30, Igor Lins e Silva - EOS Rio [@igorls]: @hiwinlin
- 05.06.2018 11:43:23, WinLin-EOSBIXIN [@hiwinlin]: yes, multisig compare successfully
- 05.06.2018 11:43:39, Igor Lins e Silva - EOS Rio [@igorls]: where is the error?
- 05.06.2018 11:43:47, WinLin-EOSBIXIN [@hiwinlin]: but eosio.bios and eosio.system have difference
- 05.06.2018 11:44:05, Igor Lins e Silva - EOS Rio [@igorls]: eosio.bios don't exist anymore
- 05.06.2018 11:44:17, Igor Lins e Silva - EOS Rio [@igorls]: i don't understand the issue
- 05.06.2018 11:44:32, WinLin-EOSBIXIN [@hiwinlin]: [[Photo]]
- 05.06.2018 11:46:48, Igor Lins e Silva - EOS Rio [@igorls]: @hiwinlin are you checking state or history?
- 05.06.2018 11:47:16, WinLin-EOSBIXIN [@hiwinlin]: I just comper the cosehash of eosio account ,you can get the info by :
- cleos get code eosio
- 05.06.2018 11:47:33, Igor Lins e Silva - EOS Rio [@igorls]: ba8f337948f958209ebf6aa7d3e4d699d909181a52a155dcac3ca4dd6afe0c52
- 05.06.2018 11:48:21, WinLin-EOSBIXIN [@hiwinlin]: just supply some evidences that the eosio.system will not be modified and that will be ok
- 05.06.2018 11:48:59, WinLin-EOSBIXIN [@hiwinlin]: [[Webpage]]hat maybe just because of compile issue casue by https://github.com/EOSIO/eos/issues/3824
- 05.06.2018 11:49:24, Igor Lins e Silva - EOS Rio [@igorls]: [[Photo]]
- 05.06.2018 11:49:26, Igor Lins e Silva - EOS Rio [@igorls]: please check this
- 05.06.2018 11:49:39, Igor Lins e Silva - EOS Rio [@igorls]: it gives the exact same output here
- 05.06.2018 11:51:17, WinLin-EOSBIXIN [@hiwinlin]: well I get :
- $ sha256sum eosio.system.wasm
- fdb35bf92a4e518739e90a1bc9ffed6d1aef94b31a279b52876d84025931af88 eosio.system.wasm
- 05.06.2018 11:51:56, WinLin-EOSBIXIN [@hiwinlin]: [[Photo]]
- 05.06.2018 11:52:08, WinLin-EOSBIXIN [@hiwinlin]: I build with docker and cmdline is :
- 05.06.2018 11:52:22, WinLin-EOSBIXIN [@hiwinlin]: docker build -t winlin/eosmainnet:v1.0.1 --build-arg branch=v1.0.1 --build-arg symbol=EOS .
- 05.06.2018 11:52:23, Igor Lins e Silva - EOS Rio [@igorls]: yes.. thats why your paths are different
- 05.06.2018 11:53:13, Jae Chung - HKEOS [@jaechung]: you must use the same path @hiwinlin
- 05.06.2018 11:53:34, Igor Lins e Silva - EOS Rio [@igorls]: also for the boost libs
- 05.06.2018 11:54:13, Igor Lins e Silva - EOS Rio [@igorls]: [[Photo]]
- 05.06.2018 11:54:20, WinLin-EOSBIXIN [@hiwinlin]: so, do you have write some artical about your boot sequances and describe what the desgin of your steps
- that will make it more clear for the comminity I think
- 05.06.2018 11:54:22, Igor Lins e Silva - EOS Rio [@igorls]: hardcoded paths on the wast file
- 05.06.2018 11:54:57, WinLin-EOSBIXIN [@hiwinlin]: I hope to see the detail of your boot sequacnes, that will be more helpful
- 05.06.2018 11:55:09, Zhao Yu@ EOSLaoMao.com [@zhaoyu]: yeah, would love to have some sort of auto generated action log or something
- 05.06.2018 11:55:57, WinLin-EOSBIXIN [@hiwinlin]: I need go to bad ,it is 00:00 AM, if you have any info please @ me tks
- 05.06.2018 11:56:12, Craig Murray - EOS BlockSmith [@craigmurray100]: the boot sequence is not published for chain C? Seems like it should be
- 05.06.2018 11:59:43, Igor Lins e Silva - EOS Rio [@igorls]: validation tools should not follow sequences, they should be able to validade state
- 05.06.2018 12:01:15, Zhao Yu@ EOSLaoMao.com [@zhaoyu]: good point
- 05.06.2018 12:01:38, Zhao Yu@ EOSLaoMao.com [@zhaoyu]: actually, you can fetch all the actions out from the chain
- 05.06.2018 12:03:18, Craig Murray - EOS BlockSmith [@craigmurray100]: yes you can fetch the tx, and you can validate state, but it would also be nice to review the boot sequence that created them. Is there some reason to not publish this information?
- 05.06.2018 12:04:35, Igor Lins e Silva - EOS Rio [@igorls]: all data is on chain
- 05.06.2018 12:04:35, Jae Chung - HKEOS [@jaechung]: it hasn't been published yet, @igorls will do so now
- 05.06.2018 12:04:56, Marc-Antoine Ross - EOS Canada [@marcantoineross]: 👍
- 05.06.2018 12:05:19, Craig Murray - EOS BlockSmith [@craigmurray100]: thx! 👍
- 05.06.2018 12:05:32, Marc-Antoine Ross - EOS Canada [@marcantoineross]: That will surely help reach full consensus on the content we want on chain, 1 more ✔️!
- 05.06.2018 12:05:47, Jae Chung - HKEOS [@jaechung]: Along with the injector script too.
- 05.06.2018 12:10:09, Igor Lins e Silva - EOS Rio [@igorls]:
- run_cmd "set contract eosio "$CONTRACTS/eosio.bios" -p eosio";
- run_cmd "create account eosio b1 EOS5cujNHGMYZZ2tgByyNEUaoPLFhZVmGXbZc9BLJeQkKZFqGYEiQ";
- run_cmd "create account eosio eosio.msig $EOSIO_PUB"
- run_cmd "create account eosio eosio.token $EOSIO_PUB"
- run_cmd "create account eosio eosio.ram $EOSIO_PUB"
- run_cmd "create account eosio eosio.ramfee $EOSIO_PUB"
- run_cmd "create account eosio eosio.names $EOSIO_PUB"
- run_cmd "create account eosio eosio.stake $EOSIO_PUB"
- run_cmd "create account eosio eosio.saving $EOSIO_PUB"
- run_cmd "create account eosio eosio.bpay $EOSIO_PUB"
- run_cmd "create account eosio eosio.vpay $EOSIO_PUB"
- run_cmd "push action eosio setpriv "'["eosio.msig",1]'" -p eosio"
- run_cmd "set contract eosio.msig "$CONTRACTS/eosio.msig" -p eosio.msig"
- run_cmd "set contract eosio.token "$CONTRACTS/eosio.token" -p eosio.token"
- cleos push action eosio.token create '["eosio","10000000000.0000 EOS"]' -p eosio.token
- cleos get currency stats eosio.token EOS
- cleos push action eosio.token issue '["eosio", "1000000000.0000 EOS", "initial issuance"]' -p eosio
- cleos get currency stats eosio.token EOS
- run_cmd "set contract eosio "$CONTRACTS/eosio.system" -p eosio"
- cleos push action eosio buyrambytes '["eosio","b1",8192]' -p eosio
- cleos transfer eosio b1 "10.0000 EOS" "eosacknowledgments.io" -p eosio
- cleos system delegatebw --transfer eosio b1 "49999995.0050 EOS" "49999995.0050 EOS"
- ## inject tokens
- ## revert eosio to system
- ## resign eosio
- 05.06.2018 12:10:51, Igor Lins e Silva - EOS Rio [@igorls]: here is some boot sequence
- 05.06.2018 12:13:06, jem [@jemxpat]: For GB core paused net:
- $ cleos get currency stats eosio.token "EOS"
- {
- "EOS": {
- "supply": "1000000000.0000 EOS",
- "max_supply": "10000000000.0000 EOS",
- "issuer": "eosio"
- }
- }
- 05.06.2018 12:17:37, Marc-Antoine Ross - EOS Canada [@marcantoineross]: can this be on github please?
- 05.06.2018 12:18:17, Michael Yeates [@michaelyeates]: we are very busy validating the software at the moment
- 05.06.2018 12:18:51, Marc-Antoine Ross - EOS Canada [@marcantoineross]: we'd like to help @michaelyeates. sharing code helps paralelize work
- 05.06.2018 12:19:48, Igor Lins e Silva - EOS Rio [@igorls]: this draft? is totally irrelevant for validation
- 05.06.2018 12:19:57, Zhao Yu@ EOSLaoMao.com [@zhaoyu]: how did you manage to do this?
- 05.06.2018 12:23:44, jem [@jemxpat]: As mentioned before, the fundamental cause of the difference in hash value is the inclusion of the absolute path of the build directory as part of the hashed data. Many have been using "/home/eos" as the base directory from which the code is built, including the boost libraries.
- I guess that the base directory used in the Docker build is something different. Perhaps if /home/eos was used, it would give the same hash as we see.
- While the hash and data in the files should be independent of the source code location, currently it is not. This is a bug, but overall just an annoyance when understood.
- 05.06.2018 12:27:41, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: Change path or have these changes: https://github.com/EOSIO/eos/pull/3834/files
- 05.06.2018 12:30:07, Rohan - EOSauthority [@RohanWillAnswer]: [[Photo]]
- 05.06.2018 12:30:34, Craig Murray - EOS BlockSmith [@craigmurray100]: should this be in github?
- .....05.06.2018 12:43:07, Branden Espinoza [@espinbrand]: >>unsupported service message type: messageActionChatJoinedByLink<<
- 05.06.2018 12:48:07, Ricardo Marlière - EOS Brazil [@rbmarliere]: 🙏
- 05.06.2018 12:56:18, Thomas "Disclaimer" Cox (all opinions mine) [@thomasbcox]: What time and link for the next meeting please?
- 05.06.2018 12:59:18, Thomas "Disclaimer" Cox (all opinions mine) [@thomasbcox]: Report from the Test Protocol Workgroup:
- 05.06.2018 12:59:18, Thomas "Disclaimer" Cox (all opinions mine) [@thomasbcox]: {{FWD: Thomas "Disclaimer" Cox (all opinions mine) [@thomasbcox], 05.06.2018 12:46:35}}
- Test Protocol group will be preparing a suggested protocol for two things:
- 1. How to propose a UA test
- 2. How to sign up for & run a UA test and report results
- Work is being done by Branden Espinoza with assist from Sharif and Ross and at least 1 other
- 05.06.2018 12:59:18, Thomas "Disclaimer" Cox (all opinions mine) [@thomasbcox]: {{FWD: Thomas "Disclaimer" Cox (all opinions mine) [@thomasbcox], 05.06.2018 12:48:00}}
- We have a draft statement of UA Testing Exit Criteria we'll be releasing.
- Work on the draft is being done by Sharif and several others
- 05.06.2018 12:59:19, Thomas "Disclaimer" Cox (all opinions mine) [@thomasbcox]: {{FWD: Thomas "Disclaimer" Cox (all opinions mine) [@thomasbcox], 05.06.2018 12:49:28}}
- The test protocol will be accompanied by a tool or set of tools to make it easier to keep the data accurate and complete
- 05.06.2018 12:59:55, Thomas "Disclaimer" Cox (all opinions mine) [@thomasbcox]: Timeline is - in 60 minutes we will get a progress report from @espinbrand and an estimated time to completion
- 05.06.2018 13:00:25, Thomas "Disclaimer" Cox (all opinions mine) [@thomasbcox]: @sharifbouktila will share the draft Exit Criteria statement in I believe that same timeframe
- 05.06.2018 13:54:26, Joshua Qiu @EOSeco [@Joshuaqiu]: do we still have meeting in 18 utc ?
- 05.06.2018 13:54:31, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Proposed Validation Exit Criteria
- 1. Completed Test Plan verified by multiple BP teams
- 2. No new proposals of critical tests outstanding
- 3. Confirmation from the EMLG that Validation is Complete (Go/No Go)
- 4. Decision on which chain to enable for staking
- 05.06.2018 13:54:52, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Yes in 6 minutes
- 05.06.2018 13:55:03, Sharif Bouktila - EOS Dublin [@sharifbouktila]: [[Webpage]]https://zoom.us/j/911102470
- 05.06.2018 13:55:56, mini - EOS Argentina [@moonxor]: Love this idea
- 05.06.2018 13:56:30, mini - EOS Argentina [@moonxor]: +1
- 05.06.2018 14:19:09, Joshua Qiu @EOSeco [@Joshuaqiu]: agree on issue
- 05.06.2018 14:19:12, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Issue more tokens for accounts
- 05.06.2018 14:19:17, Jae Chung - HKEOS [@jaechung]: Option 1: Take From Unregistered Accounts, Return them later on fallback registration
- Option 2: Issue more tokens
- 05.06.2018 14:19:59, Thomas "Disclaimer" Cox (all opinions mine) [@thomasbcox]: Option 3: Borrow from 'unregistered' balance, pay back loan from producer pay over first N days
- 05.06.2018 14:20:13, Thomas "Disclaimer" Cox (all opinions mine) [@thomasbcox]: What's the total number of tokens required, roughly?
- 05.06.2018 14:20:14, James Sutherland - Cypherglass.com [@JamesSutherland]: Please say the amount
- 05.06.2018 14:20:27, Nate D - Aloha EOS [@n8dee]: Yes, what is the amount? @abourget
- 05.06.2018 14:21:58, Michael Yeates [@michaelyeates]: 11821
- 05.06.2018 14:22:01, Domi - EOS Rio [@domiscd]: Option 1: Take From Unregistered Accounts, Return them later:
- $ cleos get currency stats eosio.token "EOS"
- {
- "EOS": {
- "supply": "1000000000.0000 EOS",
- "max_supply": "10000000000.0000 EOS",
- "issuer": "eosio"
- }
- }
- Option 2: Issue more tokens:
- {
- "EOS": {
- "supply": "1000011821.0000 EOS",
- "max_supply": "10000000000.0000 EOS",
- "issuer": "eosio"
- }
- }
- 05.06.2018 14:22:02, Josh Kauffman - EOS Canada [@joshkauff]: {{FWD: WinLin-EOSBIXIN [@hiwinlin], 03.06.2018 05:52:37}}
- and I run with issued '1,000,011,818.7401 EOS' but failed with error in record_unregsited:
- assertion failure with message: overdrawn balance
- so I change issue amount into "1000012000.0000 EOS" and run again, and successed, eosio balnace is 181.0276 EOS, so the final EOS issue should be :
- 1000012000.0000 - 181.0276 = 1,000,011,818.9724
- and the eosio.ram and eosio.ramfee may change 1-2 EOS,
- so we can issue once and use 1,000,011,820.0000 EOS, and after BIOS, the balance of eosio is just 1-2 EOS left, which I think is ok
- 05.06.2018 14:23:42, Josh Kauffman - EOS Canada [@joshkauff]: so if we round up and call that 12,000 extra tokens. that affects the supply by 0.024% . It also only creates 600 tokens per year extra in inflation. Or only 120 tokens FOR THE WHOLE YEAR that goes towards BP pay.
- 05.06.2018 14:38:20, Sharif Bouktila - EOS Dublin [@sharifbouktila]: We can adjust inflation later to reduce supply , or repay from a Worker Proposal
- 05.06.2018 14:54:37, Ravi (Blockpro.one) [@ravincal]: Are you guys only validating the snapshot or also if things like regProducer works?
- 05.06.2018 14:56:37, James Sutherland - Cypherglass.com [@JamesSutherland]: Both. Data validation and Software function validation
- 05.06.2018 14:57:54, Ravi (Blockpro.one) [@ravincal]: My regProducer call is returning no active private key error.. though wallet is open and private keys are visible there
- 05.06.2018 14:58:20, Ravi (Blockpro.one) [@ravincal]: Check my messages in bios boot channel
- 05.06.2018 14:59:04, Ravi (Blockpro.one) [@ravincal]: New account creation worked without a hitch
- 05.06.2018 14:59:27, Ravi (Blockpro.one) [@ravincal]: But regProducer not working
- .....05.06.2018 15:01:36, Josh Kauffman - EOS Canada [@joshkauff]: >>invited Ovi - eosromania22 - BP [@iamveritas]<<
- 05.06.2018 15:04:12, James Sutherland - Cypherglass.com [@JamesSutherland]: Very busy right now @ravincal but I wil ltry to hook up with you later.
- 05.06.2018 15:05:48, Timothy Lewis [@trustlayer]: If your BP is on this list, MSG me:
- boot.eostitan.com
- ec2-18-188-4-97.us-east-2.compute.amazonaws.com
- m.jeda.one
- api.eosuk.io
- node.eosmeso.io
- prod.mainnet.eos.cybex.io
- ec2-18-188-38-75.us-east-2.compute.amazonaws.com
- fullnode.eoslaomao.com
- 35.234.53.243
- eosboot.chainrift.com
- boot.eostitan.com
- ec2-54-95-202-14.ap-northeast-1.compute.amazonaws.com
- 05.06.2018 15:07:00, Rohan - EOSauthority [@RohanWillAnswer]: Is being on the list a good thing or a bad thing? 🙂
- 05.06.2018 15:07:14, Timothy Lewis [@trustlayer]: no comment.
- 05.06.2018 15:07:20, Shaheen Counts - EOS BlockSmith [@SeanPaul19]: Haha
- 05.06.2018 15:07:25, Rohan - EOSauthority [@RohanWillAnswer]: 😂
- 05.06.2018 15:09:15, Marshall - SaltBlock [@marshallslc]: I can stop sweating now
- 05.06.2018 15:09:20, Andrey Marchuk [@madz7z]: 🙂
- 05.06.2018 15:09:58, Ravi (Blockpro.one) [@ravincal]: Sent you Dm
- 05.06.2018 15:20:43, Craig Murray - EOS BlockSmith [@craigmurray100]: woot! my name didnt show up... 😅
- 05.06.2018 15:21:07, Craig Murray - EOS BlockSmith [@craigmurray100]: also thank you for doing this, very important
- 05.06.2018 15:21:39, Luke Stokes [@lukestokes]: As requested on the Zoom call, here are the screenshots I mentioned of the conversation with Dan. His criteria five days ago:
- 1. Does someone lose money?
- 2. Is there down time?
- 3. Does the chain just pause for 24 hours while we fix it?
- He goes on to say his primary concern is the loss of private keys since we do not have account recovery for these early accounts generated via the snapshot. If we're not dealing with a category 1 or 2 issue, we should be in a pretty good place to launch. Category 3 would be very unfortunate, but something we can recover from. As Dan said 5 days ago "I don't think the current test network is actually invalid" (which would imply it's valid).
- 05.06.2018 15:22:01, Luke Stokes [@lukestokes]: [[Photo]]
- 05.06.2018 15:22:07, James Sutherland - Cypherglass.com [@JamesSutherland]: +1
- 05.06.2018 15:22:14, Luke Stokes [@lukestokes]: [[Photo]]
- 05.06.2018 15:24:40, Luke Stokes [@lukestokes]: Now for my personal opinion: if we don't have mission critical bugs, we should move forward. There will be more bugs discovered in the future in production, regardless of how much sanitized testing we do under lab conditions. Continuing to find bugs this late in the process is certainly concerning, but if they aren't show stoppers, we should be able to deal with it. The best penetration testers I've ever seen come from unconvention sources, not from structured security firms. The sooner we get this out there live, the sooner we'll know what we're really dealing with.
- 05.06.2018 15:28:04, mini - EOS Argentina [@moonxor]: +1 to @lukestokes
- 05.06.2018 15:28:10, Marshall - SaltBlock [@marshallslc]: Agreed. Everyone is anxious to get it launched and tensions are high. I think there may be a misconception that people are searching for bugs blindly, but really are just making sure the software works as intended through the entire process. Once we can be sure everything works as intended I dont think anyone is calling for testing beyond that.
- 05.06.2018 15:28:31, Michael Yeates [@michaelyeates]: 👍
- 05.06.2018 15:28:41, deck - EOS New York [@deckw]: +1
- 05.06.2018 15:28:47, Denis | EOS Nation [@DenisCarriere]: 👍 @lukestokes Thanks for sharing, very valuable information
- 05.06.2018 15:29:05, Joshua Qiu @EOSeco [@Joshuaqiu]: agreed
- 05.06.2018 15:29:55, Luke Stokes [@lukestokes]: I've shared the above message in various channels, but feel free to share further if it's helpful. I'm excited to see this thing launch and then enjoy the wild ride we find ourselves on after that. 🙂
- 05.06.2018 15:30:27, Jae Chung - HKEOS [@jaechung]: 👍
- 05.06.2018 15:37:16, Thomas "Disclaimer" Cox (all opinions mine) [@thomasbcox]: Excellent thank you
- 05.06.2018 15:42:00, Marc-Antoine Ross - EOS Canada [@marcantoineross]: Just to recap my recommendation:
- A) We continuously improve security of the network (multiple initiatives going on now to help, thanks to @trustlayer @dftle)
- In parallel:
- 1. We agree on the optimal boot sequence (is RAM buying the only issue to solve?)
- 2. We code it (if not already done) and review open source implementation
- 3. We inject it (if not already done) on a swarm of nodes that includes everyone
- 4. We test it (insert Thomas' test report spreadsheet here)
- 5. We select an ABP to hold custody of the network until the 15% vote is done
- 6. We make sure we are all meshed together and help hide the ABP
- 7. We announce validation and voting starts.
- Please note: No one can brag about the software used, being the boot node or the ABP. We just don't care!!
- We are all working together to make this better for the token holders.
- 05.06.2018 15:45:07, mini - EOS Argentina [@moonxor]: +1
- 05.06.2018 15:46:36, Luke Stokes [@lukestokes]: Sounds good to me, though I have had concerns with the "code it" part in the past if it could lead to any possible security concerns. To me, the key step is the validation of the chain. However that chain is created doesn't matter to me as much as it being a valid chain. Once we have a valid chain (regardless of the process), I'm happy to go live with it without caring who created it or how it was created.
- 05.06.2018 15:46:40, Patrick B - Aloha EOS: 👍
- 05.06.2018 15:47:11, Marc-Antoine Ross - EOS Canada [@marcantoineross]: +1
- 05.06.2018 15:53:51, Luke Stokes [@lukestokes]: [[Webpage]]Related to the concerns here Dan mentioend about web portals for voting, I'd like to make sure people know about this desktop voting tool @jestagram created: https://steemit.com/eos/@greymass/announcing-eos-voter-an-eos-block-producer-voting-tool-and-light-wallet
- I've used his Vessel wallet before on STEEM along with many other products he's built over the last (almost) two years. They are soild. I helped him test this on the Jungle Testnet when he first launched it, and it worked well.
- I much prefer independent blockchain applications over the old model of centralized points of failure (website servers which can be hacked, DNS which can be hacked, domains which can be phished, etc). To me, moving our community away from "use this website" to "use this open-source, trusted tool" is a better approach long term.
- 05.06.2018 15:53:58, Thomas "Disclaimer" Cox (all opinions mine) [@thomasbcox]: [[Webpage]]BPs doing testing, please join the Zoom testing call at the top of the current hour (in 6 minutes)
- https://zoom.us/j/5788946547
- .....05.06.2018 16:37:52, Timothy Lewis [@trustlayer]: >>invited ndoy |EOS Cafe: Calgary | [@n_doy]<<
- 05.06.2018 16:38:10, ndoy |EOS Cafe: Calgary | [@n_doy]: Thanks @trustlayer !
- 05.06.2018 16:41:13, Roman CryptoLions [@romanskaskiw]: Link to document?
- 05.06.2018 16:41:42, ndoy |EOS Cafe: Calgary | [@n_doy]: Brandon is updating
- 05.06.2018 16:42:38, Thomas "Disclaimer" Cox (all opinions mine) [@thomasbcox]: Summary of testing meeting:
- 1. All test DEFINITIONS to be entered within 49 minutes of now
- 2. Links to the data entry forms (2) and spreadsheet (1) will be posted here as soon as Branden makes the requested changes
- 3. All test RESULTS to be added ASAP, by 2300 UTC preferred.
- 4. If you have a bunch of results in a file, you can paste a link to the file in the 'results' field -- you don't have to laboriously re-enter old data
- 05.06.2018 16:46:40, Rohan - EOSauthority [@RohanWillAnswer]: Thanks Branden for putting together the document so quickly
- 05.06.2018 16:47:24, Jose Toriello [@josetoriello]: Word.
- 05.06.2018 16:51:22, Thomas "Disclaimer" Cox (all opinions mine) [@thomasbcox]: Question was posed, "how do I show a user the Constitution? How do I display the relevant Ricardian Contract for an action?"
- Answer for web clients - the web api that lets you serialize JSON to binary should retrieve the JSON ABI of the command, which contains a section [ricardian clauses] which contains what you need. It'll be ugly as-is, so format it using "handlebar" or "moustache" method
- I'm trying to get some sample code to share.
- There's a plan to upgrade CLEOS to do this at the command line and show the user the contract.
- 05.06.2018 16:54:15, Roman CryptoLions [@romanskaskiw]: What's the framing of this whole effort? Is it critical testing before going public with the launch?
- Asking because we don't know how to reconcile the hard deadline with continuous testing we've been doing since April in the Jungle Testnet.
- 05.06.2018 16:56:24, Thomas "Disclaimer" Cox (all opinions mine) [@thomasbcox]: Please focus on critical tests
- 05.06.2018 16:58:26, Denis | EOS Nation [@DenisCarriere]: 🔔REQUEST: Been in touch with a few Asian BP, they are asking if we can push our 2300 UTC (decision meeting) by +2 hours.
- Most of them are not even awake to receive this message, Mao (EOSReal) is still monitoring Telegram and it's 5am for him.
- EOS Sillicon Valley also mentioned it would be better if we pushed a couple hours.
- ❓PROPOSAL: Have the decision meeting pushed to 0100 UTC Wednesday (June 6)
- 👍/ 👎for moving +2 hours?
- 05.06.2018 17:00:45, Marc-Antoine Ross - EOS Canada [@marcantoineross]: 👍
- 05.06.2018 17:00:55, deck - EOS New York [@deckw]: 👍
- 05.06.2018 17:01:04, Nate D - Aloha EOS [@n8dee]: 👍
- 05.06.2018 17:02:04, Nate D - Aloha EOS [@n8dee]: telegram needs a reaction counter like slack and discord
- 05.06.2018 17:02:56, Branden Espinoza [@espinbrand]: Could I get one person to test the forms really quickly?
- 05.06.2018 17:03:05, deck - EOS New York [@deckw]: tag
- 05.06.2018 17:03:40, Rick Schlesinger - EOS New York [@unlockrs]: 👍 @maoyifeng is a good human
- 05.06.2018 17:03:57, Patrick B - Aloha EOS: 👍
- 05.06.2018 17:05:48, Jose Toriello [@josetoriello]: 👍
- 05.06.2018 17:10:09, deck - EOS New York [@deckw]: Tagging @DavidMoss
- 05.06.2018 17:10:13, Sharif Bouktila - EOS Dublin [@sharifbouktila]: @marcantoineross I've made you an Admin here so you can pin next call details.
- 05.06.2018 17:10:54, Marc-Antoine Ross - EOS Canada [@marcantoineross]: Thanks
- 05.06.2018 17:11:22, Denis | EOS Nation [@DenisCarriere]: [[Photo]]😴
- 05.06.2018 17:11:52, Thomas "Disclaimer" Cox (all opinions mine) [@thomasbcox]: [[Webpage]]Enter your tests!
- https://docs.google.com/spreadsheets/d/1dWqr9Izk4b2xJnoS94nMD0aKR6UU3Xssy7iowwtcxBQ/edit?usp=sharing
- 05.06.2018 17:11:59, Rohan - EOSauthority [@RohanWillAnswer]: 👍I won’t be able to make it, but I feel Asian BPs need to be kept in the loop
- 05.06.2018 17:16:35, Craig Murray - EOS BlockSmith [@craigmurray100]: 👍
- 05.06.2018 17:18:28, Thomas "Disclaimer" Cox (all opinions mine) [@thomasbcox]: @sharifbouktila the 'next validation' call message pinned in this channel is obsolete
- 05.06.2018 17:19:25, Rohan - EOSauthority [@RohanWillAnswer]: Are we ok to have cleos commands for create steps?
- 05.06.2018 17:20:16, Branden Espinoza [@espinbrand]: So long as the procedure to use them is clear to the audience.
- 05.06.2018 17:20:18, deck - EOS New York [@deckw]: @marcantoineross I believe you are admin now
- 05.06.2018 17:21:17, Rohan - EOSauthority [@RohanWillAnswer]: Ok, something like this
- > cleos create account ACC_NAME_GENESIS NEW_ACC EOS_PUB EOS_PUB
- That should be fairly easy for most people to understand.. the capital letters need to be replaced in..
- 05.06.2018 17:21:17, Branden Espinoza [@espinbrand]: So long as the procedure to use them is clear to the audience. The idea is to keep the Steps as process oriented as possible.
- 05.06.2018 17:21:36, Sharif Bouktila - EOS Dublin [@sharifbouktila]: @marcantoineross will pin next one, I won’t make it.
- 05.06.2018 17:21:44, Rohan - EOSauthority [@RohanWillAnswer]: Yes, I like to have ready to execute commands, much easier to compelte them all..
- 05.06.2018 17:21:55, Branden Espinoza [@espinbrand]: Out of my world of expertise.
- 05.06.2018 17:22:48, Rohan - EOSauthority [@RohanWillAnswer]: If we say create an account, its not clear what is being tested.. Specific commands will be easier? I will keep adding it that way.. maybe a note on what it is intended to do followed by the command.
- 05.06.2018 17:23:15, Branden Espinoza [@espinbrand]: So long as the procedure to use them is clear to the audience.
- 05.06.2018 17:23:24, Branden Espinoza [@espinbrand]: sorry... didn't mean to post that again.
- 05.06.2018 17:23:25, Branden Espinoza [@espinbrand]: lol
- 05.06.2018 17:24:00, Branden Espinoza [@espinbrand]: I sincerely hope the tool helps. I'm logging off for about 2 hours. I will check back as soon as I can to address any issues.
- 05.06.2018 17:30:44, Thomas "Disclaimer" Cox (all opinions mine) [@thomasbcox]: [[Webpage]]Testing Zoom call in 5 minutes, at 21:35 UTC
- https://zoom.us/j/5788946547
- 05.06.2018 17:31:58, Thomas "Disclaimer" Cox (all opinions mine) [@thomasbcox]: Folks engaged in testing, please hop on -- I have a question from Bart and Arhag about the testing method, there's a concern that certain approaches might create spurious errors
- 05.06.2018 17:34:33, deck - EOS New York [@deckw]: @abourget @michaelyeates @ eric
- 05.06.2018 17:35:16, Roman CryptoLions [@romanskaskiw]: @xebb82
- 05.06.2018 17:35:23, deck - EOS New York [@deckw]: yea that one haha
- 05.06.2018 17:35:26, Roman CryptoLions [@romanskaskiw]: 😊
- 05.06.2018 17:35:45, deck - EOS New York [@deckw]: @igorls
- .....05.06.2018 17:42:10, Marc-Antoine Ross - EOS Canada [@marcantoineross]: >>unsupported service message type: messageActionPinMessage<<
- 05.06.2018 17:52:16, Sam - EOS Dublin [@samnoble]: Are we all in agreement that we are moving THE CALL from 2300 UTC to 0100 UTC? I haven't seen any objections yet.
- 05.06.2018 17:53:36, Gustavo Mick - EOS Argentina [@gustavomick]: Eos argentina agrees
- 05.06.2018 17:53:37, David Moss [@DavidMoss]: 9am Wednesday China time. Definitely want Asia to participate here
- 05.06.2018 17:56:38, Branden Espinoza [@espinbrand]: After you submit a case, you are taken to a page that includes a link to edit your case. You should bookmark that link for each case since it is the only means you have to edit a case once it is submitted.
- 05.06.2018 17:56:44, James Sutherland - Cypherglass.com [@JamesSutherland]: Works for me. Please pin that if it's for real.
- 05.06.2018 17:56:54, James Sutherland - Cypherglass.com [@JamesSutherland]: If it becomes real I mean
- 05.06.2018 17:59:48, Marc-Antoine Ross - EOS Canada [@marcantoineross]: ALL: Next zoom call at 0100UTC to review test results and agree on plan
- .....05.06.2018 17:59:57, Marc-Antoine Ross - EOS Canada [@marcantoineross]: >>unsupported service message type: messageActionPinMessage<<
- 05.06.2018 18:00:04, James Sutherland - Cypherglass.com [@JamesSutherland]: +1
- 05.06.2018 18:07:45, mini - EOS Argentina [@moonxor]: +1
- 05.06.2018 18:27:35, deck - EOS New York [@deckw]: Validation call with Aarin Hagerty(edited for correct spelling)
- Pausing the chain is dangerous because there a potential to fork, if keys are shared and/or we have a rogue ABP. After 15%, everyone can fix a checkpoint at the first schedule, making that key irrelevant for eternity.
- Having defined process for injection and validation on a testnet is sufficient to get high degree of confidence that a mainnet using the same process, validation being rerun, is good.
- Two options:
- Throw all current mainnet candidates out after testing is complete and recreate using the deterministic approaches we have. We would validate the boot sequence and the state right after. (recommended)
- Reuse the chain we have been testing and allow this potential issue to remain
- Testing delayed transactions would be interesting.
- 05.06.2018 18:47:43, Stuardo Rodríguez 🇬🇹 [@stuardo_str]: Throw all current mainnet candidates out after testing is complete and recreate using the deterministic approaches we have.
- <- 👏👏👏👏👏👏👏
- 05.06.2018 18:48:48, Stuardo Rodríguez 🇬🇹 [@stuardo_str]: that would throw all "ego" away about any opinion of bios/gb chain, and launch a unified mainnet together.
- 05.06.2018 19:02:22, Thomas "Disclaimer" Cox (all opinions mine) [@thomasbcox]: [[Webpage]]Arhag, often goes by Aarin Hagerty, various spellings. Github is "https://github.com/arhag"
- 05.06.2018 19:03:30, Thomas "Disclaimer" Cox (all opinions mine) [@thomasbcox]: [[Webpage]]Please enter test results!
- Please enter test results!
- Please enter test results!
- https://docs.google.com/spreadsheets/d/1dWqr9Izk4b2xJnoS94nMD0aKR6UU3Xssy7iowwtcxBQ/edit#gid=1501553294
- 05.06.2018 19:07:40, Nathan Rempel | GenerEOS Sydney [@nsrempel]: Honestly I thought this was all done during all those dry runs everyone was doing, but can't turn back the clock now. I agree if we're happy with acceptance testing and have an agreed validation test, it makes sense if the code hasn't changed, re-start, validate, open tx
- 05.06.2018 19:11:27, deck - EOS New York [@deckw]: The 'nice' thing about restarting the chain is it gives us the oppurtunity to decided on the method of paying for RAM at launch. One group or the other can launch, freeze for N hours, allow for everyone to mesh, validate (much like the snapshot), and report results. It might just work out :)
- 05.06.2018 19:11:47, Timothy Lewis [@trustlayer]: +1
- 05.06.2018 19:12:10, Craig Murray - EOS BlockSmith [@craigmurray100]: we already have one ready for tx though, why does that need restarted?
- 05.06.2018 19:12:35, Nathan Rempel | GenerEOS Sydney [@nsrempel]: Have we run the UAT against a copy of that chain?
- 05.06.2018 19:12:52, Timothy Lewis [@trustlayer]: Have we had a well done voting tool at all during this?
- 05.06.2018 19:13:06, Nathan Rempel | GenerEOS Sydney [@nsrempel]: Just need to make sure we tighten up the timelines
- 05.06.2018 19:13:39, Nathan Rempel | GenerEOS Sydney [@nsrempel]: like, UAT is done, we have 2 hours to boot, 2 hours to validate, if everyone checks in good, send it
- 05.06.2018 19:14:46, deck - EOS New York [@deckw]: I believe that eosportal was tested against a Core testnet
- 05.06.2018 19:15:06, Thomas "Disclaimer" Cox (all opinions mine) [@thomasbcox]: Only 13 test cases added so far.
- ZERO test results added so far.
- Please pick up the pace.
- 05.06.2018 19:15:06, Nathan Rempel | GenerEOS Sydney [@nsrempel]: Can i throw some of the latest firestarter results in here i should ido that elsewhere?
- 05.06.2018 19:15:42, Thomas "Disclaimer" Cox (all opinions mine) [@thomasbcox]: check your PMs
- 05.06.2018 19:16:08, mini - EOS Argentina [@moonxor]: @zhaoyu did some testing, dont know if they put it in the sheet, but they have already test a lot of things.
- 05.06.2018 19:16:40, Craig Murray - EOS BlockSmith [@craigmurray100]: yeah the chain B has 2.5 days of testing already, there was a spreadsheet for that, not sure what happened to all that 😢
- 05.06.2018 19:17:08, Nathan Rempel | GenerEOS Sydney [@nsrempel]: Does the DaoMao test incude any/many/all of those 13 cases?
- 05.06.2018 19:18:11, Thomas "Disclaimer" Cox (all opinions mine) [@thomasbcox]: If you can find it, paste the link in to the test results form. No reinventing wheels here please
- 05.06.2018 19:18:31, Thomas "Disclaimer" Cox (all opinions mine) [@thomasbcox]: They may not know about the form yet.
- 05.06.2018 19:19:35, mini - EOS Argentina [@moonxor]: [[Webpage]]https://github.com/eoslaomao/eos-cluster-generator
- 05.06.2018 19:19:35, Craig Murray - EOS BlockSmith [@craigmurray100]: Dublin would know they were doing all the record keeping. I'm fighting with this private network wireguard mesh situation trying to find the correct fork for each of my nodes
- 05.06.2018 19:19:55, mini - EOS Argentina [@moonxor]: Conclusion
- 1 The voting works as expected.
- 2 The delegatebw and undelegatebw works as expected.
- 3 The network needs to reach 15% votings, until then, you cannot unstake all tokens.
- 4 staked tokens weather be used for voting or not, both can be unstaked after 15% voting readched.
- 05.06.2018 19:20:00, mini - EOS Argentina [@moonxor]: from their repo
- 05.06.2018 19:20:21, Thomas "Disclaimer" Cox (all opinions mine) [@thomasbcox]: Outstanding. Please PM me the link to their repo
- 05.06.2018 19:20:22, mini - EOS Argentina [@moonxor]: yes probably
- 05.06.2018 19:22:58, Nathan Rempel | GenerEOS Sydney [@nsrempel]: [[Webpage]]how quickly can someone update https://github.com/eoslaomao/eos-cluster-generator to cover the remaning UATs?
- 05.06.2018 19:24:25, Nathan Rempel | GenerEOS Sydney [@nsrempel]: i can start writing an eosjs for the non-time sensitive test cases too if no one has yet, or are we running manual commands?
- 05.06.2018 19:25:58, mao EOSREAL [@maoyifeng]: chinese bpcs confirming the call 1.5 hours ahead and are discussing their stands.
- 05.06.2018 19:37:02, Jose Toriello [@josetoriello]: Please specify what time you mean?
- 05.06.2018 19:37:26, Nathan Rempel | GenerEOS Sydney [@nsrempel]: I'm currently writing a supplment document with all the cleos commands to run to match the test cases and what the expected results should be
- 05.06.2018 19:37:35, Nathan Rempel | GenerEOS Sydney [@nsrempel]: BPs can manually run them if they wish and add to the form
- 05.06.2018 19:37:42, mao EOSREAL [@maoyifeng]: Roughly 1.5 hours ahead. 9pm est.
- 05.06.2018 19:42:09, WinLin-EOSBIXIN [@hiwinlin]: 👍
- 05.06.2018 19:51:54, WinLin-EOSBIXIN [@hiwinlin]: to be honest most case in the execel doc, many of Chinese BPs have finished since Dawn-v4.2 and many times test in v1.0.1
- 05.06.2018 19:52:10, Nathan Rempel | GenerEOS Sydney [@nsrempel]: The UAT document seems a little schewed, like Case RAM market has name bidding..
- 05.06.2018 19:57:17, Nathan Rempel | GenerEOS Sydney [@nsrempel]: [[Webpage]]Ok guys here's a supplement document: https://docs.google.com/document/d/1qFEUt3Z-sy8-vWrzmoH-TjXE6850_4lCpdzBoDaj4A0/edit?usp=sharing
- 05.06.2018 19:57:57, Nathan Rempel | GenerEOS Sydney [@nsrempel]: This is so individual BPs or "anyone" can run the UAT tests by command line. Please continue to add commands for each test case as appropriate
- 05.06.2018 19:58:48, Nathan Rempel | GenerEOS Sydney [@nsrempel]: some of the latest UATs will require scripts to be run, however I see several are from @BohdanCryptoLions
- 05.06.2018 19:59:09, Nathan Rempel | GenerEOS Sydney [@nsrempel]: If @BohdanCryptoLions has been running against release-rc or v1.0.1 perhaps you can publish those results
- 05.06.2018 19:59:23, mini - EOS Argentina [@moonxor]: awesome work!
- .....05.06.2018 20:01:43, Nate D - Aloha EOS [@n8dee]: >>invited Peter Bue | EOSAmericas<<
- .....05.06.2018 20:02:43, Peter Bue | EOSAmericas: >>invited Ed Silva [@edeosamericas]<<
- 05.06.2018 20:20:03, Thomas "Disclaimer" Cox (all opinions mine) [@thomasbcox]: 👍😁
- 05.06.2018 20:20:23, Thomas "Disclaimer" Cox (all opinions mine) [@thomasbcox]: 👍😁❤️
- 05.06.2018 20:20:50, Thomas "Disclaimer" Cox (all opinions mine) [@thomasbcox]: Wonderful.
- 05.06.2018 20:23:23, WinLin-EOSBIXIN [@hiwinlin]: yes, Asia bp teams have done lots things but someone always choose to ignore this
- 05.06.2018 20:23:35, WinLin-EOSBIXIN [@hiwinlin]: this is unfair
- 05.06.2018 20:24:22, Nathan Rempel | GenerEOS Sydney [@nsrempel]: I don't think the community is ignoring, do you have documentation that test was run? Please, PLEASE share!
- 05.06.2018 20:28:15, Thomas "Disclaimer" Cox (all opinions mine) [@thomasbcox]: I don't think it's ignoring, so much as cluelessness. I hang my head in shame
- 05.06.2018 20:30:01, WinLin-EOSBIXIN [@hiwinlin]: [[Webpage]]thanks all
- the validation tool :https://github.com/EOSBIXIN/eostoolkit/tree/master/eos_validator
- 05.06.2018 20:30:29, WinLin-EOSBIXIN [@hiwinlin]: this tool support eos-bios and manula startup chain
- 05.06.2018 20:30:32, WinLin-EOSBIXIN [@hiwinlin]: both
- 05.06.2018 20:30:33, Nathan Rempel | GenerEOS Sydney [@nsrempel]: for all the chinese BP that have used this tool, do you have a document showing the output?
- 05.06.2018 20:30:51, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: Yes it would be great to see validation from Asia
- 05.06.2018 20:30:55, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: So we can record that
- 05.06.2018 20:31:16, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: Right now, its only BIXIN and a handful that we have seen
- 05.06.2018 20:31:22, WinLin-EOSBIXIN [@hiwinlin]: [[Photo]]
- 05.06.2018 20:31:22, WinLin-EOSBIXIN [@hiwinlin]: like this
- 05.06.2018 20:31:33, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: That is run by you though
- 05.06.2018 20:31:38, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: We mean other Asian BPCs
- 05.06.2018 20:31:47, WinLin-EOSBIXIN [@hiwinlin]: [[Photo]]
- 05.06.2018 20:32:22, WinLin-EOSBIXIN [@hiwinlin]: wait please, it is too early I will find some bp to prove
- 05.06.2018 20:32:30, WinLin-EOSBIXIN [@hiwinlin]: @zhaoyu from EOSLAOMAO
- 05.06.2018 20:33:06, WinLin-EOSBIXIN [@hiwinlin]: @Joshuaqiu from ESOSeco
- 05.06.2018 20:33:17, WinLin-EOSBIXIN [@hiwinlin]: @dftle from eosAsia
- 05.06.2018 20:33:31, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: Yes we have those 3, they are vocal in calls as well
- 05.06.2018 20:33:36, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: We were wondering if theres anyone else
- 05.06.2018 20:33:40, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: We have missed?
- 05.06.2018 20:38:33, Alexandre Bourget - EOS Canada [@abourget]: WinLin was the first to write a validation tool, first to publish reports and I think he has the most complete validation took to date.
- He published his results time and time again and helped identify issues in the first 3 iterations of a mainnet candidate.
- I am truly grateful for your work WinLin
- 05.06.2018 20:39:18, Nathan Rempel | GenerEOS Sydney [@nsrempel]: 👍👍👍
- 05.06.2018 20:41:16, WinLin-EOSBIXIN [@hiwinlin]: we all support EOS and want to make her more valueable, so let's move on and introdcue her into the world
- 05.06.2018 20:41:21, Luke Stokes [@lukestokes]: Has anyone else created a currency contract yet? I tried created the eosDAC one on Jungle and it blew up. Might just be our contract is bad, but I wanted to ask here before I ask in Jungle.
- 05.06.2018 20:41:45, Nate D - Aloha EOS [@n8dee]: 👍
- 05.06.2018 20:42:01, WinLin-EOSBIXIN [@hiwinlin]: yes, I have created some token in dawn v-4.2
- 05.06.2018 20:42:08, Peter Bue | EOSAmericas: 👍👍
- 05.06.2018 20:42:30, WinLin-EOSBIXIN [@hiwinlin]: so there should be some wrong in you env,please chekc again
- 05.06.2018 20:42:36, Luke Stokes [@lukestokes]: Thanks. I'll dig into main.cpp at line 2584 to see what's broken. 🙂
- 05.06.2018 20:43:02, Luke Stokes [@lukestokes]: I had an earlier issue with boost, so it surely could be my local env. Thanks.
- 05.06.2018 20:44:39, Tyee Noprom | EOSBlock.co | EOSpace.io | EOS Committer [@tyeenoprom]: Tyee from EOSpace
- 05.06.2018 20:44:53, Nathan Rempel | GenerEOS Sydney [@nsrempel]: 👍
- 05.06.2018 20:44:58, WinLin-EOSBIXIN [@hiwinlin]: [[Webpage]]you can take a look about the EOSLaoMao tool:https://github.com/EOSLaoMao/eos-cluster-generator
- 05.06.2018 20:45:35, Luke Stokes [@lukestokes]: thanks. I haven't played with that yet, but it looks really interesting
- 05.06.2018 20:46:03, WinLin-EOSBIXIN [@hiwinlin]: yes, that's very intreasting
- .....05.06.2018 20:47:07, Gordo | HelloEOS [@gordoor]: >>unsupported service message type: messageActionChatJoinedByLink<<
- 05.06.2018 20:51:46, Branden Espinoza [@espinbrand]: Many of you know my role in this. Is there any value in me attending the next zoom call?
- 05.06.2018 20:52:57, Branden Espinoza [@espinbrand]: Maybe a better question for @thomasbcox
- 05.06.2018 20:53:43, Tyee Noprom | EOSBlock.co | EOSpace.io | EOS Committer [@tyeenoprom]: You can use this tool to validate yourself.
- 05.06.2018 20:53:55, Thomas "Disclaimer" Cox (all opinions mine) [@thomasbcox]: No
- 05.06.2018 20:53:59, Joshua Qiu @EOSeco [@Joshuaqiu]: i have another , but only check balance
- 05.06.2018 20:54:14, Branden Espinoza [@espinbrand]: TY.
- 05.06.2018 20:54:27, Gordo | HelloEOS [@gordoor]: We've been using WinLin's tool to do the validation
- 05.06.2018 20:55:31, Tyee Noprom | EOSBlock.co | EOSpace.io | EOS Committer [@tyeenoprom]: That's fine. You can use your own tool to validate both chains.
- 05.06.2018 20:56:50, Joshua Qiu @EOSeco [@Joshuaqiu]: [[Webpage]]https://github.com/qiushaoxi/eos-validation snapshot balance are fine on both side
- 05.06.2018 20:57:41, Nathan Rempel | GenerEOS Sydney [@nsrempel]: Ok guys, in advance of this call I have created a survey so we can get some "real documented concensus" rather than everyone speaking up in a call
- 05.06.2018 20:57:42, Nathan Rempel | GenerEOS Sydney [@nsrempel]: [[Webpage]]https://goo.gl/forms/D8M7c5QFyMCwmCaF2
- 05.06.2018 20:58:27, Nathan Rempel | GenerEOS Sydney [@nsrempel]: I'd appreciate if someone could review it quickly if there's anything that should be added
- 05.06.2018 20:58:51, deck - EOS New York [@deckw]: I believe the suggested time to launch was tomorrow at 1300 utc
- 05.06.2018 20:59:06, Nathan Rempel | GenerEOS Sydney [@nsrempel]: oops my bad
- 05.06.2018 20:59:15, deck - EOS New York [@deckw]: np good work putting this together
- 05.06.2018 20:59:19, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: I dont think we can reverse chain probably, do you mean load from backup?
- 05.06.2018 20:59:22, Nathan Rempel | GenerEOS Sydney [@nsrempel]: mixing up my time zone
- 05.06.2018 20:59:48, deck - EOS New York [@deckw]: @marcantoineross Can you post the zoom link
- 05.06.2018 20:59:52, Marc-Antoine Ross - EOS Canada [@marcantoineross]: [[Webpage]]All, here the proposed flow for the call
- A) We continuously improve security of the network (multiple initiatives going on now to help, thanks to @trustlayer @dftle)
- In parallel:
- 1. We agree on the optimal boot sequence (is RAM buying the only issue to solve?)
- 2. We code it (if not already done) and review open source implementation
- 3. We inject it (if not already done) on a swarm of nodes that includes everyone
- 4. We test it (insert Thomas' test report spreadsheet here)
- 5. We select an ABP to hold custody of the network until the 15% vote is done
- 6. We make sure we are all meshed together and help hide the ABP
- https://zoom.us/j/681388230
- .....05.06.2018 20:59:59, Marc-Antoine Ross - EOS Canada [@marcantoineross]: >>unsupported service message type: messageActionPinMessage<<
- 05.06.2018 21:01:08, Nathan Rempel | GenerEOS Sydney [@nsrempel]: Yes
- 05.06.2018 21:01:19, Nathan Rempel | GenerEOS Sydney [@nsrempel]: basically "restart"
- 05.06.2018 21:13:41, Timothy Lewis [@trustlayer]: Guys the call is at max participants
- 05.06.2018 21:14:07, Denis | EOS Nation [@DenisCarriere]: 🤔 I'll let people know about this
- 05.06.2018 21:14:20, Craig Murray - EOS BlockSmith [@craigmurray100]: i thought someone with a 200 max was going to create?
- 05.06.2018 21:14:28, Nathan Rempel | GenerEOS Sydney [@nsrempel]: Are we having 1 BP rep in the call? I know there is a lot of BP candidates
- 05.06.2018 21:16:12, Timothy Lewis [@trustlayer]: No one from my team is in the call and I would like to be.
- 05.06.2018 21:16:28, Andrey Marchuk [@madz7z]: there are 2 spots
- 05.06.2018 21:16:29, Nathan Rempel | GenerEOS Sydney [@nsrempel]: they are working to expand to 200
- 05.06.2018 21:16:30, Andrey Marchuk [@madz7z]: try now
- 05.06.2018 21:16:31, Nathan Rempel | GenerEOS Sydney [@nsrempel]: some open spaces now
- 05.06.2018 21:16:53, Timothy Lewis [@trustlayer]: sweet
- 05.06.2018 21:16:54, Timothy Lewis [@trustlayer]: ty
- 05.06.2018 21:35:23, James Sutherland - Cypherglass.com [@JamesSutherland]: 200 people allowed now
- 05.06.2018 21:35:42, James Sutherland - Cypherglass.com [@JamesSutherland]: Nice job @marcantoineross !
- 05.06.2018 21:37:03, Denis | EOS Nation [@DenisCarriere]: 👨👩👧👧 more people allowed!
- 05.06.2018 21:37:34, James Sutherland - Cypherglass.com [@JamesSutherland]: Please mute yourself
- 05.06.2018 21:37:58, James Sutherland - Cypherglass.com [@JamesSutherland]: Stay muted until you need to speak
- .....05.06.2018 22:29:38, Ren Kaidi - EOSWTZ [@FoxMr]: >>invited Jialin Li - EOS WTZ [@Jialinli]<<
- 05.06.2018 22:31:06, Alex - EOS SOCAL [@gmory]: Did the call just die?
- 05.06.2018 22:31:13, Nathan Rempel | GenerEOS Sydney [@nsrempel]: non
- 05.06.2018 22:31:21, James Sutherland - Cypherglass.com [@JamesSutherland]: I am still on
- 05.06.2018 22:31:49, James Sutherland - Cypherglass.com [@JamesSutherland]: 138 people connected
- 05.06.2018 22:33:57, deck - EOS New York [@deckw]: @marcantoineross do it
- 05.06.2018 22:34:03, deck - EOS New York [@deckw]: I vote @marcantoineross
- 05.06.2018 22:39:15, Alexandre Bourget - EOS Canada [@abourget]: Great call tonight folks
- 05.06.2018 22:39:42, Gustavo Mick - EOS Argentina [@gustavomick]: @abourget could you summarize results from the meeting, thanks
- 05.06.2018 22:40:15, Nathan Rempel | GenerEOS Sydney [@nsrempel]: Can we get a group made for the Asian security audit coordination asap
- 05.06.2018 22:40:32, Thomas "Disclaimer" Cox (all opinions mine) [@thomasbcox]: Remarkably productive, all things considered. Many thanks to @marcantoineross for hosting
- 05.06.2018 22:40:43, mini - EOS Argentina [@moonxor]: +1 to this
- 05.06.2018 22:42:26, Nathan Rempel | GenerEOS Sydney [@nsrempel]: @bo_eossv
- 05.06.2018 22:42:50, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: @eosnewyork, we also need a global statement for this security delay
- .....05.06.2018 22:43:02, Nathan Rempel | GenerEOS Sydney [@nsrempel]: >>invited Bo - EOS Silicon Valley [@bo_eossv]<<
- 05.06.2018 22:43:48, Gustavo Mick - EOS Argentina [@gustavomick]: We need a summary of the meeting for those who have not been able to be there.
- 05.06.2018 22:43:52, Chitty 🅑 - EOS Argentina [@JChitty]: Did I understand correctly that EOS BIOS will launch and EOS Core will appoint the ABPs... right?
- 05.06.2018 22:44:05, Andrey Marchuk [@madz7z]: productive call, thank you!
- 05.06.2018 22:44:07, Nathan Rempel | GenerEOS Sydney [@nsrempel]: No..?
- 05.06.2018 22:44:18, Nathan Rempel | GenerEOS Sydney [@nsrempel]: what do you mean appoint?
- 05.06.2018 22:45:10, Chitty 🅑 - EOS Argentina [@JChitty]: mmm, I guess we will wait for the votes to be counted.. I believe the question was SHOULD WE RESTART A NEW CHAIN, INJECTED WITH THE CURRENT 579a CONTENTS. AND HAVE GHOSTBUSTER RUN ABP TO 15% UNLOCK ?
- 05.06.2018 22:45:33, Nathan Rempel | GenerEOS Sydney [@nsrempel]: i thought that question was discarded
- 05.06.2018 22:45:48, Chitty 🅑 - EOS Argentina [@JChitty]: why?
- 05.06.2018 22:46:20, Craig Murray - EOS BlockSmith [@craigmurray100]: because JEM declined on behalf of ghostbusters to do so, he was not particularly interested in that
- 05.06.2018 22:46:38, Chitty 🅑 - EOS Argentina [@JChitty]: Did Jem represent all of Ghostbuster?
- 05.06.2018 22:46:43, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: I believe the final was for everyone to join 579a chain once it passes security check phase
- 05.06.2018 22:47:24, Craig Murray - EOS BlockSmith [@craigmurray100]: Yes, the offer was made for ghostbusters to pick up that chain and be an ABP for it, which they declined. But there was overwhelming support of the 579a chain.
- 05.06.2018 22:47:47, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: They offerred to join as every other BPC
- 05.06.2018 22:48:08, Craig Murray - EOS BlockSmith [@craigmurray100]: yes, they offered to join as ordinary BPCs and not be the ABP
- 05.06.2018 22:48:09, Bo - EOS Silicon Valley [@bo_eossv]: Sure!
- 05.06.2018 22:48:09, Chitty 🅑 - EOS Argentina [@JChitty]: Ok, I must have missed that.. will wait on the report
- 05.06.2018 22:49:17, Luke Stokes [@lukestokes]: One person representing multiple BPCs seemed counter to DPOS to me, but I'm glad we worked through things and got consensus. Great meeting.
- 05.06.2018 22:49:50, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: You mean proxying votes is counter to DPOS?
- 05.06.2018 22:49:55, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: Thats a first
- 05.06.2018 22:50:03, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: They delegated their votes
- 05.06.2018 22:50:05, Luke Stokes [@lukestokes]: No, proxy is fine if it's on chain and known
- 05.06.2018 22:50:07, Nathan Rempel | GenerEOS Sydney [@nsrempel]: I think the idea is that, randomly assingning a new ABP does not serve much purpose at this point. Everyone regardless of security model is welcome to mesh
- .....05.06.2018 22:50:54, Nathan Rempel | GenerEOS Sydney [@nsrempel]: >>invited Hugo Yang - EOS Silicon Valley [@hugo_eossv]<<
- 05.06.2018 22:51:00, David P - EOS42 [@GunnisonCap]: Given the timezone of the vote, a proxy representative makes sense in the same way Bo was for multiple BP's
- 05.06.2018 22:51:08, Luke Stokes [@lukestokes]: Thanks for clarifying that's an arangement (via proxy) people are comfortable with
- 05.06.2018 22:53:14, Igor Lins e Silva - EOS Rio [@igorls]: Just for clarification, we see no problem in ghosbusters as the abp if the community decides so, the point jem was stating was simply that we don't have the absolute desire to be the ABP
- 05.06.2018 22:54:13, Nathan Rempel | GenerEOS Sydney [@nsrempel]: @bo_eossv @hugo_eossv Is there a testnet already existing or do you need the community to start one?
- 05.06.2018 22:54:28, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: could we use jungle?
- 05.06.2018 22:54:35, Igor Lins e Silva - EOS Rio [@igorls]: +1
- 05.06.2018 22:54:44, Nathan Rempel | GenerEOS Sydney [@nsrempel]: Well, I thought the idea would be to run it against a clone of our candidate
- 05.06.2018 22:54:51, Nathan Rempel | GenerEOS Sydney [@nsrempel]: but as long as the code is the code
- 05.06.2018 22:54:59, Nathan Rempel | GenerEOS Sydney [@nsrempel]: 🤷♂️
- 05.06.2018 22:56:26, Nathan Rempel | GenerEOS Sydney [@nsrempel]: I just don't want us to fall into the trap that we're all waiting for this security audit - but kept in the dark. Would like to see some action
- 05.06.2018 22:56:52, Nate D - Aloha EOS [@n8dee]: I thought the idea was to start one on production infrastructure
- 05.06.2018 22:56:58, Nate D - Aloha EOS [@n8dee]: But yes, needs clarification.
- 05.06.2018 22:57:20, Nathan Rempel | GenerEOS Sydney [@nsrempel]: That was my understanding also, so Im eager to swtich our production over to this test
- .....05.06.2018 22:57:26, Nathan Rempel | GenerEOS Sydney [@nsrempel]: >>invited Dr Ye EOS WashingtonDC - TChain [@doc309]<<
- 05.06.2018 22:57:51, Bo - EOS Silicon Valley [@bo_eossv]: Let me do a quick check
- 05.06.2018 22:58:05, Ren Kaidi - EOSWTZ [@FoxMr]: 👍
- .....05.06.2018 22:59:47, Ren Kaidi - EOSWTZ [@FoxMr]: >>invited Xueyuan She_EOS_WTZ [@xueyuanshe]<<
- 05.06.2018 22:59:55, Stuardo Rodríguez 🇬🇹 [@stuardo_str]: +1
- 05.06.2018 23:00:22, Nathan Rempel | GenerEOS Sydney [@nsrempel]: Is someone already preparing a joint statement to the community regarding the eosio issuance of tokens to gift the genesis accounts with free ram?
- 05.06.2018 23:01:51, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: Cypherglass offered to do a video
- 05.06.2018 23:01:54, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: @finchify
- 05.06.2018 23:02:54, Nathan Rempel | GenerEOS Sydney [@nsrempel]: It would be great if we could get a joint steemit/etc to include and share the video
- 05.06.2018 23:02:56, Dr Ye EOS WashingtonDC - TChain [@doc309]: Just to clarify, GB/core chain is pretty complicated due to an optional WG (not officially released) - not a serious VPN solution that could be hard to defend in security audits - it is wise to back out
- 05.06.2018 23:04:45, Dafeng Guo - EOS Asia [@dftle]: Details will be coming out soon. Things are rolling.
- 05.06.2018 23:04:57, Nathan Rempel | GenerEOS Sydney [@nsrempel]: Thanks for that
- 05.06.2018 23:05:02, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: Thanks Dafeng, keep us updated!
- 05.06.2018 23:06:13, Bo - EOS Silicon Valley [@bo_eossv]: 👍👍
- 05.06.2018 23:07:45, Dr Ye EOS WashingtonDC - TChain [@doc309]: please get tchain posted
- 05.06.2018 23:08:21, Kevin Rose - EOS New York [@eosnewyork]: We’ve already issued it
- 05.06.2018 23:08:36, Kevin Rose - EOS New York [@eosnewyork]: The statement today states the time needed
- 05.06.2018 23:08:55, Kevin Rose - EOS New York [@eosnewyork]: [[Webpage]]https://steemit.com/eos/@eosnewyork/the-eos-mainnet-launch-security-and-software-testing
- 05.06.2018 23:08:56, Dafeng Guo - EOS Asia [@dftle]: Are we referring to use jungle’s network mesh? I believe most nodes there don’t have prod level infra. We are happy to work with @BohdanCryptoLions if he is interested in helping to set up a new testnet that’s just for security testing.
- 05.06.2018 23:09:38, Timothy Lewis [@trustlayer]: Great post.
- 05.06.2018 23:09:43, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: I think Bohdan has one of the most experience in the community for launching testnets. He is also a relatively neutral party that can help launch a security testnet
- 05.06.2018 23:09:57, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: Would be nice to reach out to him @BohdanCryptoLions
- 05.06.2018 23:10:23, Kevin Rose - EOS New York [@eosnewyork]: Group effort. Thanks Tim. If you haven’t posted already please feel free to. This is a statement meant to represent us all.
- 05.06.2018 23:11:26, Nathan Rempel | GenerEOS Sydney [@nsrempel]: +1
- 05.06.2018 23:15:53, Jose Toriello [@josetoriello]: +1+1
- 05.06.2018 23:16:00, Jose Toriello [@josetoriello]: +1
- 05.06.2018 23:20:30, Luke Stokes [@lukestokes]: I was thinking the same thing. That post covered what we ended up deciding on the call quite nicely. Well done!
- 05.06.2018 23:21:17, Dr Ye EOS WashingtonDC - TChain [@doc309]: +1 happy to help if he Bohdan agrees to take the job
- 05.06.2018 23:23:31, Nathan Rempel | GenerEOS Sydney [@nsrempel]: Any other candidates to spin up this testnet if @BohdanCryptoLions isn't available?
- 05.06.2018 23:25:56, mini - EOS Argentina [@moonxor]: I think eos.host was quite good, dont know if those guys are up for it
- 05.06.2018 23:26:24, mini - EOS Argentina [@moonxor]: the monitor was superv, and they have account creation and a super nice voting system
- 05.06.2018 23:27:16, Dafeng Guo - EOS Asia [@dftle]: I contacted EOS.host yesterday
- 05.06.2018 23:27:51, Dafeng Guo - EOS Asia [@dftle]: They are happy to help
- 05.06.2018 23:28:11, Craig Murray - EOS BlockSmith [@craigmurray100]: nice job staying away from calling out these launch cohorts by name. Well done sir, thanks
- 05.06.2018 23:28:47, mini - EOS Argentina [@moonxor]: great!
- 05.06.2018 23:35:01, Crystal Rose | ⚡️SENSE | 💜shEOS [@crystalrose]: Throwing out some gratitude and kudos to everyone for working hard and giving yourself to this mission. It’s apparent how much time and care has been dedicated. I’m impressed by this group. Thank you! 🙏
- 05.06.2018 23:37:04, Branden Espinoza [@espinbrand]: I am too. I made the mistake of sticking my nose into some social media crypto groups today. They have no idea. The disconnect between what they are saying and what everyone here is doing is unreal. This is amazing to be a (small) part of.
- 05.06.2018 23:53:15, Bo - EOS Silicon Valley [@bo_eossv]: Thanks so much everyone for making EOS possible! Really appreciate it👍
- 05.06.2018 23:53:35, Bo - EOS Silicon Valley [@bo_eossv]: We are running full speed on security testing and will try our best to share as many updates as possible. This won't be a blackbox testing!
- 05.06.2018 23:54:09, Patrick B - Aloha EOS: Thanks to you as well Bo!
- .....06.06.2018 00:05:43, Mark-eosantpool [@mark_cike]: >>invited Mr.Z [@zlajoni]<<
- 06.06.2018 00:08:23, Marc-Antoine Ross - EOS Canada [@marcantoineross]: [[Webpage]]All,
- Thanks for the productive call.
- Find below the minutes, if you find anything not accurate, please suggest/comment and we will fix.
- The raw log is included at the bottom of the document.
- https://docs.google.com/document/d/1nHEA3rfWLh4uwKmFnY9ypuRnWbOjuSy0LdKUlTapS2c
- Next steps:
- June 6, 1300 UTC: to get consensus on remaining points not covered yet
- Test plan & timeline
- ABP selection
- Meshing and ABP protection
- More?
- June 7, 0100 UTC: the next go/no go discussion, hopefully with the initial security report in hand
- Thanks, good after-mor-night!
- .....06.06.2018 00:08:42, Marc-Antoine Ross - EOS Canada [@marcantoineross]: >>unsupported service message type: messageActionPinMessage<<
- 06.06.2018 00:09:18, Daniel Larimer [@daniellarimer]: That’s 1337.
- 06.06.2018 00:09:50, David Moss [@DavidMoss]: Ah, interesting. I heard it different.
- 06.06.2018 00:09:54, David Moss [@DavidMoss]: Still, a rounding error.
- 06.06.2018 00:10:11, David Moss [@DavidMoss]: Printing tokens for RAM startup is a rounding error as well.
- 06.06.2018 00:12:25, Timothy Lewis [@trustlayer]: @abourget @marcantoineross did anyone try to implement the RAM cost solution from K Kim of EOSeoul?
- 06.06.2018 00:12:26, Timothy Lewis [@trustlayer]: [[Webpage]]https://gist.github.com/redjade/c22b2d7fcab068af67c0de31724eb710
- 06.06.2018 00:13:43, Zhao Yu@ EOSLaoMao.com [@zhaoyu]: any post about the whole idea?
- 06.06.2018 00:14:24, Timothy Lewis [@trustlayer]: From the authoer @resvera: "It compiles well. I didn’t test it. I’m not quite expert on boot sequences. But a developer.
- test would be as follows. After Alex method which increases token supply and buys rams for genesis accounts, then input value of migratebiosramcost action will be the amount of token used to buy rams (not value of increased token amount). After migratebiosramcost with correct input, then bancor balance of EOS(CORE) will be clean as initial state of setcode of eosio.system contract. migratebiosramcost can be called only once as I implemented.
- This action is very delicate one. Test should include code reviews and actual push action test."
- 06.06.2018 00:22:11, Jose Toriello [@josetoriello]: I suggest "issuing" and not "printing" as a better vocabulary, especially given the sensitivity around the subject.
- 06.06.2018 00:22:15, James Sutherland - Cypherglass.com [@JamesSutherland]: Nice work on this. Thanks for taking the effort to put this document togeather.
- 06.06.2018 00:22:26, Jose Toriello [@josetoriello]: +1
- 06.06.2018 00:22:38, Jose Toriello [@josetoriello]: +1
- 06.06.2018 00:22:48, Jose Toriello [@josetoriello]: +1
- 06.06.2018 00:26:59, Timothy Lewis [@trustlayer]: There has been a potential well written solution to the whole problem that was painstakingly written by @resvera ... Has it been tested? Are we a community that doesn't try to fix issues?
- 06.06.2018 00:27:30, David Moss [@DavidMoss]: I agree it's a good solution. Would it require restarting the mainnet candidate>?
- 06.06.2018 00:27:50, Timothy Lewis [@trustlayer]: It would require a code review first, and then a test.
- 06.06.2018 00:28:10, Nathan Rempel | GenerEOS Sydney [@nsrempel]: Question do you want to solve the problem with smart contract edits? You could just load a new eosio.token and go "unissue 11890". then reload the original
- 06.06.2018 00:29:34, Timothy Lewis [@trustlayer]: If the test was successful then we would have the opportunity to walk through the boot process as a community.
- 06.06.2018 00:31:08, Nathan Rempel | GenerEOS Sydney [@nsrempel]: Have you run his code and tested it?
- 06.06.2018 00:31:21, Timothy Lewis [@trustlayer]: We are talking about 2-3 hours to complete a test. I have not run it. It needs to be tested.
- 06.06.2018 00:32:05, Timothy Lewis [@trustlayer]: I only learned about it recently, but from the quality of the code, it was well thought out and I was asking who if anyone had tried it.
- 06.06.2018 00:35:14, Marc-Antoine Ross - EOS Canada [@marcantoineross]: While the community has voted to move on & the cost/benefice is hard to justify, you are very welcome to propose this solution on the next call. Especially if it's been thoroughly tested by different teams.
- 06.06.2018 00:36:35, Nathan Rempel | GenerEOS Sydney [@nsrempel]: Yes, I think if you want to champion new solutions at this point in time the onus should be on you to test, prove, and provide evidence or get like minded individuals to do the same
- 06.06.2018 00:37:17, Nathan Rempel | GenerEOS Sydney [@nsrempel]: Otherwise we're in an endless loop of "hey what about this?"
- 06.06.2018 00:37:28, Timothy Lewis [@trustlayer]: Understood. I am now just trying to find out if anyone had tested it, or reasons they anyone may have not wanted to.
- 06.06.2018 00:37:44, Nathan Rempel | GenerEOS Sydney [@nsrempel]: Looking at the code my concern is that while it may be valid its a lot of change
- 06.06.2018 00:39:04, Nathan Rempel | GenerEOS Sydney [@nsrempel]: Considering validation tools already check every account balance i think something simpler like
- void token::unissue(quantity){
- statstable.modify( st, 0, [&]( auto& s ) {
- s.supply -= quantity;
- });
- }
- would be preferable in the eosio.token contract, then reconfirm balances = total supply
- 06.06.2018 00:39:29, Nathan Rempel | GenerEOS Sydney [@nsrempel]: Above is obviously invalid code, just making a point
- 06.06.2018 00:40:20, Nathan Rempel | GenerEOS Sydney [@nsrempel]: That said I agree with using tools that don't require contract changes... and burning the inflation through those pre-exisisting tools, personally
- .....06.06.2018 01:23:33, Emily Bush - shEOS [@emmybush]: >>invited Autumn Moss Penaloza / shEOS [@AutumnPenaloza]<<
- 06.06.2018 01:33:46, Yvonne (EOShenzhen) [@yvonnekama]: great.thank you
- 06.06.2018 01:42:05, Nathan Rempel | GenerEOS Sydney [@nsrempel]: I guess its getting late in North America, but Europe is starting to wake up, looking forward to some updates soon :)
- 06.06.2018 01:50:21, Luke Stokes [@lukestokes]: Is there a better place to discuss this tool? I had a question about why eosiocpp was removed. I can't compile my token contract to test it. cc @hiwinlin
- 06.06.2018 02:24:28, Sandwich [@sandw1ch]: [[Webpage]]https://github.com/EOSIO/eos/pull/2669/files ?
- 06.06.2018 02:27:31, WinLin-EOSBIXIN [@hiwinlin]: you can not find you eosiocpp?
- 06.06.2018 02:28:01, WinLin-EOSBIXIN [@hiwinlin]: maybe sudo make install again
- 06.06.2018 04:36:03, Nathan Rempel | GenerEOS Sydney [@nsrempel]: Any news guys?
- 06.06.2018 05:16:31, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Next call is 13:00 UTC 6th June 2018
- I’ll post Zoom details before here.
- 06.06.2018 06:21:52, Ross (EOSphere) [@rossco99]: [[Webpage]]Would be great if everyone could complete at least one test.. there don't seem to be many participants in our UAT google doc... https://docs.google.com/spreadsheets/d/1dWqr9Izk4b2xJnoS94nMD0aKR6UU3Xssy7iowwtcxBQ/edit?usp=sharing
- 06.06.2018 07:14:16, jem [@jemxpat]: Account creation RAM costs solution
- We have a total solution for RAM allocation, which rsults in no issuance of new tokens, no borrowing of tokens, and complete accurate final state of all accounts.
- During injection, we pay for RAM for each account creation using eosio. These payments go into the system accounts eosio.ram and eosio.ramfee. After injection of the full snapshot, 19,832.9750 EOS are in the eosio.ram system account, and 95.5916 EOS are in the eosio.ramfee system account.
- Just prior to the resignation of the eosio account during the boot process, the entire amount from both of these accounts are transfered to the eosio account, leaving both with a balance of zero.
- By this method, the correct amounts exist in all system account, no RAM fees are charged to user accounts, and the total supply at launch is set to exactly 1B EOS.
- We believe that this simple process is the correct way to create user accounts while maintaining precise suply of total EOS tokens.
- 06.06.2018 07:15:22, Sharif Bouktila - EOS Dublin [@sharifbouktila]: EVERYONE - We have our next Validation Call at 13:00 UTC
- If you are not directly involved in testing / validation, you do not need to attend the call.
- The focus on this call with be establishing what the current status is.
- Agreeing what the focus is for the next 6 hours.
- @samnoble or I will post clear minutes and actions to this group for awareness so nothing is lost in translation.
- 06.06.2018 07:19:10, mao EOSREAL [@maoyifeng]: This is something super.
- 06.06.2018 07:21:35, Jiri Jetmar | EOS Germany | [email protected] [@jirijetmar]: I don't want to interfere here, but I think the RAM cost issues is not the real problem. We have lived for centuries in economic systems where at no time is the available money supply known. With EOS we now have a very small and only initial delta - no one really cares. Nobody loses money, nobody gets cheated. But I'm fine with any solution that can be delivered ASAP. Thank you.
- 06.06.2018 07:22:06, jem [@jemxpat]: Kind of embarassing how easy the solution is. Since we have a snapshot of the chain just prior to resignation, applying it is a 1 minute job.
- .....06.06.2018 07:22:41, Laim Lais: >>unsupported service message type: messageActionChatJoinedByLink<<
- .....06.06.2018 07:26:39, Sharif Bouktila - EOS Dublin [@sharifbouktila]: >>unsupported service message type: messageActionPinMessage<<
- 06.06.2018 07:28:25, Nathan Rempel | GenerEOS Sydney [@nsrempel]: I feel like im missing something? Where did the EOS that goes into eosio.ram and eosio.ramfree come from? If you put it "back" in eosio... wasn't the total supply number still increased?
- 06.06.2018 07:29:46, jem [@jemxpat]: No
- 06.06.2018 07:30:00, Bohdan CryptoLions 🦁 [@BohdanCryptoLions]: I am here, need to sleep sometimes =)
- 06.06.2018 07:30:09, Nathan Rempel | GenerEOS Sydney [@nsrempel]: I feel ya brother
- 06.06.2018 07:30:16, Michael Yeates [@michaelyeates]: the eos in the ram and ramfee accounts is the exact amount that was spent for ram by eosio
- 06.06.2018 07:31:34, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Both solutions were fine with me, please think about how we collectively make a decision and close this one out.
- 06.06.2018 07:33:51, Jiri Jetmar | EOS Germany | [email protected] [@jirijetmar]: +1
- 06.06.2018 07:35:21, David P - EOS42 [@GunnisonCap]: This sounds much more elegant, how difficult is that to implement? As you’ll note from the global voting many BPs seem to just be at a stage of wanting to launch whatever provided there is no security compromise. If this can be done properly on a Core validated chain at the same equivalent stage (or ahead) then we can vote to just move and adopt that with this solution. Coming just with this I think will get a no if it implies any sort of delay. I much prefer this as a launch sequence though for many reasons.
- 06.06.2018 07:35:42, jem [@jemxpat]: It was 4 lines of script, total.
- 06.06.2018 07:36:14, Nathan Rempel | GenerEOS Sydney [@nsrempel]: I thought this was the process already done, theres still a *supply* surplus, your just moving numbers around. eosio issued 19,832.9750 EOS and 95.5916 EOS did it not? Did these EOS just appear - from where?
- 06.06.2018 07:36:25, David P - EOS42 [@GunnisonCap]: Fine so if you can implement and say “we have a chain with this solution and equally well validated” then we can vote directly between the chain to proceed with.
- 06.06.2018 07:38:22, jem [@jemxpat]: GB made a snapshot of the chain after injection and immediately before resignation of eosio (done to create base for Acceptance Test Nets), so they do these two transfers, resign, and are back to the same state as the existing paused chain.
- 06.06.2018 07:38:36, Bohdan CryptoLions 🦁 [@BohdanCryptoLions]: is testnet still needed? what data use to deploy?
- 06.06.2018 07:38:51, Sharif Bouktila - EOS Dublin [@sharifbouktila]: The question of which chain to launch is bigger.
- It doesn't feel right that we are making it on what I believe is such a small issue.
- How long would it take WP inflation to build to pay for this RAM? If we calculated that inflation from when each of the chains were established and manually inserted the inflation what would more than cover it and is in keeping with what everyone knew was coming.
- 06.06.2018 07:40:08, Nathan Rempel | GenerEOS Sydney [@nsrempel]: Would be nice if @bo_eossv and @dftle could provide an update, its been many hours
- 06.06.2018 07:40:29, jem [@jemxpat]: No one is arguing which chain to launch. It really doesn't matter.
- 06.06.2018 07:41:22, Michael Yeates [@michaelyeates]: we should reach community consensus on these issues
- 06.06.2018 07:41:39, David P - EOS42 [@GunnisonCap]: So apart from the global politics and niceties of including both bios/core in this process, we actually can just use an equally validated chain as the basis instead with a better solution? That’s what needs to be communicated on the call.
- 06.06.2018 07:41:55, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Can your approach be applied to B-Chain also.
- 06.06.2018 07:42:30, David P - EOS42 [@GunnisonCap]: I’d prefer not to have the same BP running the call and setting the questions every time. If Dublin ran a previous then last night was Canada so it’s for another to facilitate it.
- 06.06.2018 07:42:42, jem [@jemxpat]: Yes, if they made the snapshot which had been promised to GB before it was launched.
- 06.06.2018 07:42:51, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Would you like to take it?
- 06.06.2018 07:42:54, Sharif Bouktila - EOS Dublin [@sharifbouktila]: I agree
- 06.06.2018 07:42:55, Craig Murray - EOS BlockSmith [@craigmurray100]: Personally I think this issue was closed. I agree with Dan, we should just print tokens. We have a 72 hour old chain done this way, we should stick with it
- 06.06.2018 07:43:50, Sharif Bouktila - EOS Dublin [@sharifbouktila]: So they would need a new one? Or is it possible on what they have now?? Sorry I need to get this 100% clear.
- 06.06.2018 07:44:23, Michael Yeates [@michaelyeates]: they should be able to start a snapshot and fix it without rebooting
- 06.06.2018 07:44:40, David P - EOS42 [@GunnisonCap]: I would normally if people wished, but I’m in a hotel room here in SF with iffy WiFi, don’t trust it based on performance to date
- 06.06.2018 07:44:48, jem [@jemxpat]: I don't know. The snapshot had been promised, but was never delivered. I don't know if it was ever made.
- 06.06.2018 07:45:01, Joshua Qiu @EOSeco [@Joshuaqiu]: can user sell initail ram?
- 06.06.2018 07:45:51, Sharif Bouktila - EOS Dublin [@sharifbouktila]: I'll do this one then. And ask for another Chair for the next one on the call.
- 06.06.2018 07:46:17, David P - EOS42 [@GunnisonCap]: Here’s the point - Dan was asked without any other solution proposed. If we now say it’s slightly messy solution that will require political explanation and justification to community later on this chain, or cleaner solution without any such issues on this chain (and no downside of additional delays), which do you think Dan would choose?
- 06.06.2018 07:46:42, Craig Murray - EOS BlockSmith [@craigmurray100]: Dan already weighed in on this
- 06.06.2018 07:46:43, David P - EOS42 [@GunnisonCap]: I think he’d probably say use the better solution if there’s no delay
- 06.06.2018 07:46:49, Craig Murray - EOS BlockSmith [@craigmurray100]: there is delay
- 06.06.2018 07:47:01, David P - EOS42 [@GunnisonCap]: What’s the delay?
- 06.06.2018 07:47:17, Craig Murray - EOS BlockSmith [@craigmurray100]: restarting a chain, revalidating it, changing the boot sequence
- 06.06.2018 07:47:34, Joshua Qiu @EOSeco [@Joshuaqiu]: i am thinking about sell inital ram, who should pay for this sell
- 06.06.2018 07:48:05, David P - EOS42 [@GunnisonCap]: Thanks, I also think you’re trying to be politically neutral here too
- 06.06.2018 07:48:19, David P - EOS42 [@GunnisonCap]: Is that correct @jemxpat ?
- 06.06.2018 07:48:28, Zhao Yu@ EOSLaoMao.com [@zhaoyu]: if user can sell ram, then the only thing fixed by jem’s new solution is the extra 2 or 3 EOS I think.
- 06.06.2018 07:48:50, Michael Yeates [@michaelyeates]: they cannot sell their initial ram until someone else buys it
- 06.06.2018 07:49:31, Zhao Yu@ EOSLaoMao.com [@zhaoyu]: so they can, as sell means there are buyers and sellers.
- 06.06.2018 07:49:52, Craig Murray - EOS BlockSmith [@craigmurray100]: i cannot believe we're still debating this non-issue
- 06.06.2018 07:50:02, Michael Yeates [@michaelyeates]: if they sell too much of their initial ram then their account will not function well
- 06.06.2018 07:50:04, Craig Murray - EOS BlockSmith [@craigmurray100]: even Dan knows it's not a real issue
- 06.06.2018 07:50:22, Zhao Yu@ EOSLaoMao.com [@zhaoyu]: what about I buy some more then sell it?
- 06.06.2018 07:50:46, jem [@jemxpat]: Maybe Dan "knows" it. But, it is wrong, and will look very bad to some.
- 06.06.2018 07:50:59, Craig Murray - EOS BlockSmith [@craigmurray100]: glad you know more than Dan
- 06.06.2018 07:51:18, Craig Murray - EOS BlockSmith [@craigmurray100]: you're not even a BP, what is your stake in all this, a non BP who speaks for an entire launch group?
- 06.06.2018 07:51:22, jem [@jemxpat]: I'm not a syncophant.
- 06.06.2018 07:52:12, Craig Murray - EOS BlockSmith [@craigmurray100]: you think you know better than him though, clearly.
- 06.06.2018 07:52:41, jem [@jemxpat]: I think he is not omniscient.
- 06.06.2018 07:52:52, Craig Murray - EOS BlockSmith [@craigmurray100]: no one does
- 06.06.2018 07:53:10, Zhao Yu@ EOSLaoMao.com [@zhaoyu]: so your solution rounded it up to exactly 1B, and this 70 hours validated by multiple teams serveral times B-chain has a 3 extra EOS
- 06.06.2018 07:53:52, Craig Murray - EOS BlockSmith [@craigmurray100]: yes, it's just another attempt to delay, not a real issue, that's my take. Issue is resolved, closed, and yesterday's news. You'll need a better reason than that to stop progress I hope
- 06.06.2018 07:54:01, Jiri Jetmar | EOS Germany | [email protected] [@jirijetmar]: if we continue to deal with trivialities like this, then we see EOS trading at or below USD 5. That would destroy an enormous value. If EOS were a listed company, the shareholders (EOS Tokenholders) would fire the management that is acting so negligently.
- 06.06.2018 07:54:18, jem [@jemxpat]: There is no extra 3 EOS in the chain which ran this. There was precisely 1000000000.0000 EOS.
- 06.06.2018 07:55:22, Zhao Yu@ EOSLaoMao.com [@zhaoyu]: that I agree
- 06.06.2018 07:57:02, Michael Yeates [@michaelyeates]: if we dont deal with these issues yet then we will have to face them on a live running chain
- 06.06.2018 07:57:56, Zhao Yu@ EOSLaoMao.com [@zhaoyu]: so what kind of disasters we are talking about because of these 3 extra EOS?
- 06.06.2018 07:58:30, jem [@jemxpat]: It is curently 19k extra.
- 06.06.2018 07:58:46, Zhao Yu@ EOSLaoMao.com [@zhaoyu]: no
- 06.06.2018 07:58:50, Zhao Yu@ EOSLaoMao.com [@zhaoyu]: user can sell ram
- 06.06.2018 07:58:55, Michael Yeates [@michaelyeates]: public opinion backlash about the 20k tokens added
- 06.06.2018 07:58:57, Zhao Yu@ EOSLaoMao.com [@zhaoyu]: ram is EOS too
- 06.06.2018 07:59:02, Dr Ye EOS WashingtonDC - TChain [@doc309]: what is your most recent total block count that is increasing by the second?
- 06.06.2018 07:59:23, Jiri Jetmar | EOS Germany | [email protected] [@jirijetmar]: But this RAM issue is only an initial topic and the delta is very limited. I do not think that anybody cares..
- 06.06.2018 07:59:24, Zhao Yu@ EOSLaoMao.com [@zhaoyu]: there’s no 20K, there’s 3 EOS, user can sell ram, ram is EOS
- 06.06.2018 08:00:36, Michael Yeates [@michaelyeates]: the supply discrepency is 20k
- 06.06.2018 08:00:46, Sandwich [@sandw1ch]: General Information: Rounding a number after calculations, not before, can reduce accumulative dust.
- 06.06.2018 08:02:34, Zhao Yu@ EOSLaoMao.com [@zhaoyu]: we will tell public the truth. and the truth is, we give them some free ram. that’s all.
- 06.06.2018 08:03:54, Jiri Jetmar | EOS Germany | [email protected] [@jirijetmar]: +1 .. every bank on the planet has delta in their books. It doesn't bother anyone, because it's negligible.
- 06.06.2018 08:03:55, Stuardo Rodríguez 🇬🇹 [@stuardo_str]: Just to confirm, do we have a call right now? Or in 12 hours?
- 06.06.2018 08:04:21, jem [@jemxpat]: One hour later.
- 06.06.2018 08:05:38, Stuardo Rodríguez 🇬🇹 [@stuardo_str]: In 1h thank you
- 06.06.2018 08:06:44, Sharif Bouktila - EOS Dublin [@sharifbouktila]: GUYS - Lets keep it on topic and professional please!
- 06.06.2018 08:07:18, Dr Ye EOS WashingtonDC - TChain [@doc309]: may i ask who can turn a node to produce blocks on boot-net?
- 06.06.2018 08:09:42, Stuardo Rodríguez 🇬🇹 [@stuardo_str]: I think jem's solution is valid, and I think it is very easy to re run all automated tests in 30 mins. I think it is easy to launch a "net - d" chain, re run the tests and compare. I think this "net d" adds more value than the already running "net c". We also don't have much to do while the Chinese validation runs. I see it as a good thing to do to kill time during this 24h period anyway
- 06.06.2018 08:10:37, Dr Ye EOS WashingtonDC - TChain [@doc309]: tons of things going on if you stay hands on
- 06.06.2018 08:10:55, Stuardo Rodríguez 🇬🇹 [@stuardo_str]: "net d" won't stall more than what we are already waiting
- 06.06.2018 08:13:08, Dr Ye EOS WashingtonDC - TChain [@doc309]: anyone who got time think of ways of beefing up the security of the mainnet and the best practices of SDLC/Agile process and best practice on launching the mainnet (production deployment)
- 06.06.2018 08:14:12, Benjamin Wang | EOShenzhen [@iBenjaminWang]: EOS is not a stock. The value of EOS is based on trust and consensus.
- 06.06.2018 08:14:29, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: I have a feeling Dan just wants the chain up before the hackathon 😉
- 06.06.2018 08:19:32, Kyunghwan Kim - EOSeoul.io [@resvera]: [[Webpage]]I wrote some notes for documenting and understanding what is going on current ram cost approaches and its consequences technically. I hope this article helps to see the whole picture.
- https://gist.github.com/redjade/bdbcb10708e165b29d0301c5defa1d83
- 06.06.2018 08:25:10, Thomas "Disclaimer" Cox (all opinions mine) [@thomasbcox]: Longer term, a rotating chair is probably wise.
- 06.06.2018 08:26:58, Dr Ye EOS WashingtonDC - TChain [@doc309]: 👍
- 06.06.2018 08:27:03, Nathan Rempel | GenerEOS Sydney [@nsrempel]: Can we try to follow a similar format, whoever curates the meeting. It seemed to be highly effective
- 06.06.2018 08:31:08, Thomas "Disclaimer" Cox (all opinions mine) [@thomasbcox]: An agenda helps
- 06.06.2018 08:35:43, Thomas "Disclaimer" Cox (all opinions mine) [@thomasbcox]: What is the status of data entry of test cases and test results?
- 06.06.2018 08:36:17, David P - EOS42 [@GunnisonCap]: 😁 I think you’re right! See you in SF soon? I arrived last night.
- 06.06.2018 08:37:25, David P - EOS42 [@GunnisonCap]: @sharifbouktila want a hand with some thoughts on that? I’ll DM
- 06.06.2018 08:37:51, David P - EOS42 [@GunnisonCap]: Let’s just keep the call positive and productive all, we all want the same thing
- 06.06.2018 08:38:26, Craig Murray - EOS BlockSmith [@craigmurray100]: I'm very close to canceling my flight and staying home for tulip. I'm very concerned about the state of the launch and our inability to form consensus about validation, UAT, private networks, security testing, and when to allow a chain to process tx
- 06.06.2018 08:38:28, Jiri Jetmar | EOS Germany | [email protected] [@jirijetmar]: 100% agree!
- 06.06.2018 08:38:30, Syed Jafri | EOS Cafe / Calgary [@syed_jafri]: About to hop in a plane in 20 😁
- 06.06.2018 08:43:28, David P - EOS42 [@GunnisonCap]: Understand, no doubt as a senior technologist we need you focused on the launch so if it’s ongoing you can’t really do both.
- 06.06.2018 08:44:20, David P - EOS42 [@GunnisonCap]: Great, shout me when you’re around. I think Joe of Asia and Steve of Tribe amongst anothers are around later so maybe we can all get out
- 06.06.2018 08:47:34, Craig Murray - EOS BlockSmith [@craigmurray100]: Jerry is a far better systems admin than I am. Likely he can handle anything that comes up. I'm just concerned right now because we're still running multiple topologies simultaneously and so far the group of BPs are struggling to form consensus on many topics. It's the uncertainty of the entire launch that has me debating my fight later today
- 06.06.2018 08:48:47, Nathan Rempel | GenerEOS Sydney [@nsrempel]: I'm also concerned we haven't heard from our friends in China on a joint testnet to run against and have these audits performed against. I'll be hopeful for an update shortly
- 06.06.2018 08:49:14, Craig Murray - EOS BlockSmith [@craigmurray100]: a full 3rd party security audit takes a lot of time. YOu can't do that in a day, it takes weeks to test properly
- 06.06.2018 08:49:52, David P - EOS42 [@GunnisonCap]: Come over, I talk to Gee and all BP's will have access to an all day business work room so we can use it as much as needed once there
- 06.06.2018 08:50:50, David P - EOS42 [@GunnisonCap]: @maoyifeng @bo_eossv Are either of you able to help get an update on this?
- 06.06.2018 08:51:36, Nathan Rempel | GenerEOS Sydney [@nsrempel]: Yes, however being told we'll see results in 24-48 hours and we'll have whitebox testing everyone can participate in... my expectations are a bit high
- 06.06.2018 08:53:26, Sharif Bouktila - EOS Dublin [@sharifbouktila]: [[Webpage]]https://zoom.us/j/786777697
- 06.06.2018 08:53:33, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Call in 7 minutes
- 06.06.2018 08:56:49, mao EOSREAL [@maoyifeng]: I am still on plane. Can not join this call. Please update me your progress.
- 06.06.2018 09:03:28, Hyunwoong JI - EOSeoul.io [@EOSeoul_hyunwoong]: 👍
- 06.06.2018 09:11:49, Nathan Rempel | GenerEOS Sydney [@nsrempel]: @sharifbouktila on Jungle I *just* tested using my genesis account to successfully create and regproducer successfully
- 06.06.2018 09:13:07, Nathan Rempel | GenerEOS Sydney [@nsrempel]: using 8kb ram, 1EOS/1EOS cpu/net
- 06.06.2018 09:19:14, Brett Jeffery @ EOS the World [@EOStheWorld]: [[Webpage]]Morning meeting starting
- https://zoom.us/j/786777697
- 06.06.2018 09:32:27, - eoseoul.io DS Lim [@dslim]: 👍👍
- 06.06.2018 09:33:42, jem [@jemxpat]: Excellent! 👍
- 06.06.2018 09:35:59, Yongkil Kwon - EOSeoul.io: Awesome! 👍
- 06.06.2018 09:40:41, Dr Ye EOS WashingtonDC - TChain [@doc309]: with billions at stake, should we start a small pilot EOS holders group to run the mainnet in production for a while before a massive deployment?
- 06.06.2018 09:42:47, David Moss [@DavidMoss]: Yes, and there should be more than one.
- 06.06.2018 09:43:24, Nathan Rempel | GenerEOS Sydney [@nsrempel]: mainnet in production that's not a mainnet? need to clarify the semantics there
- 06.06.2018 09:44:29, Dr Ye EOS WashingtonDC - TChain [@doc309]: the boot-net for now
- 06.06.2018 09:45:02, Nathan Rempel | GenerEOS Sydney [@nsrempel]: anyone can already do this... it's just a public test net
- 06.06.2018 09:45:23, Nathan Rempel | GenerEOS Sydney [@nsrempel]: and should be
- 06.06.2018 09:45:35, Craig Murray - EOS BlockSmith [@craigmurray100]: cryptolions has been doing this for a month or more
- 06.06.2018 09:45:44, Craig Murray - EOS BlockSmith [@craigmurray100]: they are currently running latest tag
- 06.06.2018 09:46:05, Dr Ye EOS WashingtonDC - TChain [@doc309]: yeah... not with real money though
- 06.06.2018 09:46:15, Craig Murray - EOS BlockSmith [@craigmurray100]: you cannot run with real money and not be the mainnet
- 06.06.2018 09:46:33, David Moss [@DavidMoss]: Faucets
- 06.06.2018 09:46:35, Craig Murray - EOS BlockSmith [@craigmurray100]: once it's real money, it's real money and we've live, you cannot take it down and say "we've found issues we're going to start over"
- 06.06.2018 09:46:38, David Moss [@DavidMoss]: Or sys
- 06.06.2018 09:46:50, Craig Murray - EOS BlockSmith [@craigmurray100]: yes cryptolions has faucets
- 06.06.2018 09:47:23, Dr Ye EOS WashingtonDC - TChain [@doc309]: we are testing on jungle with faucets
- .....06.06.2018 10:00:34, Dr Ye EOS WashingtonDC - TChain [@doc309]: >>unsupported service message type: messageActionChatAddUsers<<
- 06.06.2018 10:56:53, Nathan Rempel | GenerEOS Sydney [@nsrempel]: Very productive call, thanks everyone. I'll look forward to seeing testing results appearing in the spreadsheet, lets not lose sight of that :)
- 06.06.2018 10:57:50, Nathan Rempel | GenerEOS Sydney [@nsrempel]: [[Webpage]]Perhaps we can add this document to the pin to keep eyes on it: https://docs.google.com/spreadsheets/d/1dWqr9Izk4b2xJnoS94nMD0aKR6UU3Xssy7iowwtcxBQ/edit?usp=sharing
- 06.06.2018 10:58:57, Craig Murray - EOS BlockSmith [@craigmurray100]: that felt like a very productive meeting
- 06.06.2018 10:59:02, David P - EOS42 [@GunnisonCap]: So the owner of the channel that was live streaming (and had previous BP meetings) without consent has taken down all of the previous videos. They appear to be named Malc Silberman
- 06.06.2018 10:59:44, David P - EOS42 [@GunnisonCap]: Could somebody confirm who on this channel represents them or 'EOSBschool', it seems they are not a Block Producer from an intial look
- 06.06.2018 11:00:15, David P - EOS42 [@GunnisonCap]: Yes, we need to all stay focused on the issues at hand and productive resolution now.
- .....06.06.2018 11:05:42, Benjamin Wang | EOShenzhen [@iBenjaminWang]: >>invited delin he [@hedelinok]<<
- 06.06.2018 11:15:50, Kyunghwan Kim - EOSeoul.io [@resvera]: And commented on gist above with accompaning patch for implementing token::unissue as PoC. It compiles but not tested and not quite sure before peer code review and someone's test.
- I'm not quite fluent in C++, but it was relatively not too difficult to implement. Curious why this simple and clean solution is not yet done. Anyway, it seems that to implement this solution, cleos subcommands and their unittests is super trivial to developers fluent in C++.
- 06.06.2018 11:16:18, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Does anyone know who in here is from the account who live-streamed that call?
- 06.06.2018 11:17:19, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Else I’ll start removing people who haven’t been active in the past 48 hours in this channel and lock it
- 06.06.2018 11:18:49, Mathieu Boulianne - EOS Canada [@mathieubo]: The Zoom link was shared in other public channels
- 06.06.2018 11:19:05, Hubert Chang-EOSeoul.io [@hjhc00]: All I heard was someone saying 'EOSBS'? But maybe it is best to contact Mark.
- 06.06.2018 11:19:56, David P - EOS42 [@GunnisonCap]: EOSBSchool - Malcolm Silberman
- 06.06.2018 11:20:01, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Crazy it’s why we post the link just before that call , if they aren’t in here already they should let be on the call
- 06.06.2018 11:20:03, David P - EOS42 [@GunnisonCap]: I don't know who they are on here
- 06.06.2018 11:20:20, David P - EOS42 [@GunnisonCap]: They probably are on here
- 06.06.2018 11:20:59, Josh Kauffman - EOS Canada [@joshkauff]: If we want to keep the calls private, maybe on the next call we take the time to first go thru all participants and remove those who should not be on the call.
- 06.06.2018 11:21:26, Sharif Bouktila - EOS Dublin [@sharifbouktila]: There was 140 today at one point
- 06.06.2018 11:21:49, David P - EOS42 [@GunnisonCap]: [[Photo]]This guy
- 06.06.2018 11:24:10, Hubert Chang-EOSeoul.io [@hjhc00]: Maybe our friends in US can help us out.
- 06.06.2018 11:24:41, Hubert Chang-EOSeoul.io [@hjhc00]: Kevin? :) He seems to know just about everyone
- 06.06.2018 11:25:44, Sharif Bouktila - EOS Dublin [@sharifbouktila]: I don’t think they are private , but hate the idea of someone recording and streaming without consent.
- Our position and decisions are market sensitive.
- 06.06.2018 11:25:49, Brett Jeffery @ EOS the World [@EOStheWorld]: Understood. But just because we don’t post doesn’t mean we’re not active. @hartparr is very active, but he may not post much. 👍
- 06.06.2018 11:27:19, David P - EOS42 [@GunnisonCap]: It's a breach of trust above all, I think we are all fine with the contents being released for historical record later but we don't need the distraction of fielding questions from countless others. Let's hope this doesn't happen again.
- 06.06.2018 11:27:21, Sukesh Tedla (EOSGreen) [@Sukeshtedla]: +1
- 06.06.2018 11:27:44, David P - EOS42 [@GunnisonCap]: Anyway let's get back to onto more important matters
- 06.06.2018 11:27:58, Sukesh Tedla (EOSGreen) [@Sukeshtedla]: 👍
- 06.06.2018 11:28:15, Sharif Bouktila - EOS Dublin [@sharifbouktila]: That’s fine , I don’t have the answer just yet.
- Reminds me of my post months ago about our mantra of DBAA - “Don’t be an a**hole”
- 06.06.2018 11:29:00, Ville Sundell [@SolariusFI]: So, is the consensus now, that the new selected ABP will launch their own chain?
- 06.06.2018 11:29:01, David P - EOS42 [@GunnisonCap]: Somebody pin this from Sharif.. sometimes the world needs to hear the sweet tones of an Irishman telling you that
- 06.06.2018 11:29:05, Brett Jeffery @ EOS the World [@EOStheWorld]: not criticizing you, just don’t want to get our team locked out just because we’re not posting much.
- 06.06.2018 11:29:36, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Very good call other than that , well done all!!
- 06.06.2018 11:31:23, Craig Murray - EOS BlockSmith [@craigmurray100]: no
- 06.06.2018 11:32:59, Ville Sundell [@SolariusFI]: Good 😆 (Personal opinion only, I think the existing chain is good to take over by the appointed ABP)
- 06.06.2018 11:46:31, Marc-Antoine Ross - EOS Canada [@marcantoineross]: [[Webpage]]All,
- Thanks for another productive call!
- PRIORITY: get that testing done! https://docs.google.com/spreadsheets/d/1dWqr9Izk4b2xJnoS94nMD0aKR6UU3Xssy7iowwtcxBQ
- Find bellow the minutes, if you find anything not accurate, please suggest/comment and we will fix.
- The raw log is included at the bottom of the document.
- https://docs.google.com/document/d/17cwUuPoaH1NAYfSGgvomSvjTkDrMHdVhLaxCS5t-orQ
- Next steps:
- June 7, 0100 UTC: the next go/no go discussion, hopefully with the initial security report in hand
- Meshing and ABP protection
- Ben from shEOS asks: How can we insure that voting portals are handling private keys well?
- More?
- Thanks, good after-mor-night!
- .....06.06.2018 11:46:49, Marc-Antoine Ross - EOS Canada [@marcantoineross]: >>unsupported service message type: messageActionPinMessage<<
- 06.06.2018 12:34:49, Denis | EOS Nation [@DenisCarriere]: 👍 thanks Cypherglass, your last video was very professional and your messaging was accurate to the current events.
- 06.06.2018 12:45:49, Denis | EOS Nation [@DenisCarriere]: Amazing 💪💪 we should continue to use the same lingo P0 & P1 when presenting the security audits findings
- 06.06.2018 12:47:24, Denis | EOS Nation [@DenisCarriere]: 👍 agreed, Jungle has the most experience in launching new Testnets
- 06.06.2018 12:49:28, Denis | EOS Nation [@DenisCarriere]: +1 we should not be putting any labels anymore, we're all in this together
- 06.06.2018 12:53:02, Denis | EOS Nation [@DenisCarriere]: 💪 we fully support your efforts! Everyone is very happy you and the other Chinese BPC'S/Community/Exchanges are pushing for this.
- Looking forward to see these findings (good or bad) We will be more informed on our next steps moving forward after these results surface.
- 06.06.2018 12:54:43, David P - EOS42 [@GunnisonCap]: I think from now on we need to all stop referring at all to "Bios", "Core" and any other labels and just focus on the collective solution we implement.
- 06.06.2018 12:55:11, Sharif Bouktila - EOS Dublin [@sharifbouktila]: +1
- 06.06.2018 12:55:35, Mr.Z [@zlajoni]: Can't agree more!
- 06.06.2018 12:57:12, Ben Sigman [@Bensig]: What will we call it?
- 06.06.2018 12:58:30, Kevin Rose - EOS New York [@eosnewyork]: First few digits of the chain ID seems to make sense.
- 06.06.2018 12:59:00, Craig Murray - EOS BlockSmith [@craigmurray100]: unfortuantely the chain starting in 0d6c has been forked a few times, so it's hard to refer to it just by chain id
- 06.06.2018 12:59:53, Craig Murray - EOS BlockSmith [@craigmurray100]: there is the candidate fork, for which the public endpoints appear to be down. There is the voting fork of it, which has spoofed the contracts so you can vote and test producer schedule, and then there is the fork of it that was accidental that many of the peers are synced to. There may be more, i don't know.
- 06.06.2018 13:00:24, Sharif Bouktila - EOS Dublin [@sharifbouktila]: A-Z , hopefully we don’t need > 26
- 06.06.2018 13:00:30, David P - EOS42 [@GunnisonCap]: Anything but labels that are divisive to us reaching consensus for a productive launch - that's just my view. I think we can implement solutions and select a chain without that.
- 06.06.2018 13:03:17, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Or how about jungle animals - a nod to the super work done there
- 06.06.2018 13:03:22, Craig Murray - EOS BlockSmith [@craigmurray100]: so there are two chains: 579a which has a single line. And there is 0d6c which has multiple forks. For a total of 4 chains I know of... So chains B, C, D, and E?
- 06.06.2018 13:04:00, Craig Murray - EOS BlockSmith [@craigmurray100]: where C D and E have the same chain ID
- 06.06.2018 13:04:23, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Whatever works
- 06.06.2018 13:04:28, Michael Yeates [@michaelyeates]: there are no multiple forks
- 06.06.2018 13:04:42, Michael Yeates [@michaelyeates]: we have the one chain and separate private chains for testing
- 06.06.2018 13:04:54, Michael Yeates [@michaelyeates]: out blocks db is available for download and testing
- 06.06.2018 13:04:55, Craig Murray - EOS BlockSmith [@craigmurray100]: those are forks because they have the same ID
- 06.06.2018 13:05:05, Craig Murray - EOS BlockSmith [@craigmurray100]: that's the definition of a fork
- 06.06.2018 13:05:15, Craig Murray - EOS BlockSmith [@craigmurray100]: it's not derogatory, it's jsut fact
- 06.06.2018 13:05:43, Craig Murray - EOS BlockSmith [@craigmurray100]: i spent sevearl hours yesterday trying to get off the 10k fork and onto the 11k fork. I was finally successful, but those two chains have the same ID, that's a fork
- 06.06.2018 13:06:34, Craig Murray - EOS BlockSmith [@craigmurray100]: we're talking about a way to describe them that doesn't use these silly team names
- 06.06.2018 13:06:52, Craig Murray - EOS BlockSmith [@craigmurray100]: trying to say they are not forks is just confusing to people... they see the chain ID and think they are on the right one
- 06.06.2018 13:07:19, Denis | EOS Nation [@DenisCarriere]: 👍👍 Marc is the best scribe ever!
- 06.06.2018 13:08:55, Patrick B - Aloha EOS: career change? :D
- 06.06.2018 13:11:29, Denis | EOS Nation [@DenisCarriere]: Agreed, I think the wording "print vs. issue" for optics the word "issuing" should be used for the global message.
- CC: Cypherglass & EOS New York
- 06.06.2018 13:12:12, Denis | EOS Nation [@DenisCarriere]: 😂 full-time scribe 📜
- 06.06.2018 13:12:39, Patrick B - Aloha EOS: haha
- 06.06.2018 13:13:14, Jose Toriello [@josetoriello]: +1
- 06.06.2018 13:16:57, Denis | EOS Nation [@DenisCarriere]: 👎 disagree, attempting to reimplement anything that touches token balances would jeopardize the entire validation process (this is why I mentioned +2 days)
- If we change anything that's related to the token balance, then we're back at step 0 with zero confidence on that new chain.
- This type of change would cause a drastic delay (for what reason?)
- 06.06.2018 13:19:10, Craig Murray - EOS BlockSmith [@craigmurray100]: and woudl also degrate public trust that we are making progress
- 06.06.2018 13:19:53, Denis | EOS Nation [@DenisCarriere]: 🤔 restarting a chain for the sake of testing new code that's untested would be a disaster. I personally would not have confidence in any change to core chain injections.
- 06.06.2018 13:19:59, David P - EOS42 [@GunnisonCap]: If a chain is spun up with this solution, fully validated/tested to the same level as others, then we can vote on whether to use this on the next call. That I'm fine with, but really the tests and security need to remain the bigger priorities.
- .....06.06.2018 13:20:40, Alex - EOS SOCAL [@gmory]: >>invited Cesar Diaz EOSoCal [@EOSLA]<<
- 06.06.2018 13:21:06, David P - EOS42 [@GunnisonCap]: And if nobody does it for the next call then this should be parked given we had reached consensus on the balance if not. It's not a pretty solution but explainable.
- 06.06.2018 13:21:23, Craig Murray - EOS BlockSmith [@craigmurray100]: we've got a network with 80+ hours since injection, tested repeatedly, why on earth would we throw that away if no one can prove it invalid? That's bad optics, wastes time, and would have a very hard time reaching consensus. On the call yesterday there was overwhelming consensus that the 579a chain is valid and should be used
- 06.06.2018 13:21:30, Denis | EOS Nation [@DenisCarriere]: If we spin up a new chain we're at day 0 again.
- The advantage of chain 579a is the timestamp, if our timestamp changes for today that means we've only tested this chain for <24 hours
- 06.06.2018 13:22:28, deck - EOS New York [@deckw]: What does that matter? The software has been tested and the validation of chain content takes 15-20 minutes
- 06.06.2018 13:22:48, Timothy Lewis [@trustlayer]: +1
- 06.06.2018 13:22:52, Denis | EOS Nation [@DenisCarriere]: {"initial_timestamp":"2018-06-03T13:04:31","initial_key":"EOS7m4uBgRXBM5DbXFX7FTrNiZyRzxfktmto6qQTzc2z3BhgAGHKM"}
- June 3rd is the timestamp of 579a
- 06.06.2018 13:22:59, Timothy Lewis [@trustlayer]: Why does that matter
- 06.06.2018 13:23:11, Craig Murray - EOS BlockSmith [@craigmurray100]: it takes more than that. We want it tested by multiple teams, with multiple tools, and given plenty of time for validation. You don't rush validation that's why we're still waiting on a chain that's up 80 hours and not processing transactions
- 06.06.2018 13:23:31, Denis | EOS Nation [@DenisCarriere]: The software isn't changing unless we find P0 or P1
- 06.06.2018 13:23:39, Timothy Lewis [@trustlayer]: Why? I’d rather a chain that is correct.
- 06.06.2018 13:23:56, Timothy Lewis [@trustlayer]: “These guys can’t even get token issuance right, lol”
- 06.06.2018 13:24:42, Craig Murray - EOS BlockSmith [@craigmurray100]: we're back to ram again? ugh
- 06.06.2018 13:24:46, Denis | EOS Nation [@DenisCarriere]: 👎 more then running a few scripts.
- Starting a new chain for the sake of it says "we have no confidence in the chain we've booted and validated"
- 06.06.2018 13:25:57, David P - EOS42 [@GunnisonCap]: Exactly. Buddy is right I don't know why you think this takes long Denis, it doesn't.
- 06.06.2018 13:26:01, Denis | EOS Nation [@DenisCarriere]: 👎 so you're against BM's decision and the majority of the BPs?
- 06.06.2018 13:26:13, Timothy Lewis [@trustlayer]: I’m just super frustrated that someone may have spent the time to create a real solution and in everyone’s rush to get things done, the solution was completely disregarded.
- 06.06.2018 13:26:26, Denis | EOS Nation [@DenisCarriere]: It's not a solution
- 06.06.2018 13:26:39, Craig Murray - EOS BlockSmith [@craigmurray100]: what chain uses this solution? Is this solution coded in github? Tested?
- 06.06.2018 13:26:44, Timothy Lewis [@trustlayer]: Look above at the original post.
- 06.06.2018 13:26:54, Craig Murray - EOS BlockSmith [@craigmurray100]: what are the p2p addresses so i can sync to this new "proper" chain?
- 06.06.2018 13:27:10, Timothy Lewis [@trustlayer]: Have a good day guys, I’m back to work.
- 06.06.2018 13:27:55, Jae Chung - HKEOS [@jaechung]: How else would you run an acceptance testing chain without forking?
- 06.06.2018 13:28:09, Jae Chung - HKEOS [@jaechung]: just wondering
- 06.06.2018 13:28:14, David P - EOS42 [@GunnisonCap]: To be fair, nobody had another solution it was a very quick "ah, would this even be acceptable" type check by David Moss. Nobody has since asked for his view - and it's about our consensus. As I said, if somebody wants to have that chain and solution ready by the next call, we can talk in through in 5mins, then vote to use that or not instead. If it's a better solution and there's no time cost to us then it's a win. If there were some delay impacting our overall timeline fair enough.
- 06.06.2018 13:28:14, Craig Murray - EOS BlockSmith [@craigmurray100]: it does actually. One person can run validation in 20 minutes but we have to provide time for BPs all over the world to validate and then get consensus that we are valid. That's 24-48 hours at least
- 06.06.2018 13:28:15, Denis | EOS Nation [@DenisCarriere]: Starting a new chain for the sake of it is not a good decision.
- If you don't agree with the RAM purchase, then join another chain that has a different chain injection process.
- Right now 579a has been validated and nothing is wrong with it
- 06.06.2018 13:28:54, David P - EOS42 [@GunnisonCap]: Fair enough guys, I'm just trying to give everybody a chance here and keep us moving forwards.
- 06.06.2018 13:29:13, Craig Murray - EOS BlockSmith [@craigmurray100]: a new chain is moving backwards unless we find a p0 or p1, we should not start a new chain
- 06.06.2018 13:29:32, Denis | EOS Nation [@DenisCarriere]: 👍 exactly, global planet wide validation will take forever, just because 3-5 BP's say it's good to go doesn't mean it's validated by the community
- 06.06.2018 13:31:30, Denis | EOS Nation [@DenisCarriere]: We are moving forward, wait for security audits from the Chinese community.
- If no P0 / P1 problems arise, we decide after all reading the findings and call a Go/NoGo meeting (whenever that day comes)
- 06.06.2018 13:31:52, Denis | EOS Nation [@DenisCarriere]: 👌
- 06.06.2018 13:35:43, Denis | EOS Nation [@DenisCarriere]: This just indicates that that Chain C is a "testnet" and not a "candidate chain"
- We've only got 1 real candidate chain by your comment
- 06.06.2018 13:36:14, Craig Murray - EOS BlockSmith [@craigmurray100]: I did not suggest you would
- 06.06.2018 13:36:47, Jae Chung - HKEOS [@jaechung]: I wasn't referring to that at all. I was wondering if there's another way to acceptance test an existing chain without forking. My question has been answered.
- 06.06.2018 13:37:38, Craig Murray - EOS BlockSmith [@craigmurray100]: i think there is a chain paused at 11778 on chain id 0d6c, but it's hard to tell because the public endpoints appear to be down for me. If that still exists I think it's a candidate chain but not sure
- 06.06.2018 13:39:18, Michael Yeates [@michaelyeates]: the 1178 chain is in the hands of the community
- 06.06.2018 13:39:25, Jae Chung - HKEOS [@jaechung]: ^
- 06.06.2018 13:39:26, Michael Yeates [@michaelyeates]: paused and distributed
- 06.06.2018 13:39:37, Jae Chung - HKEOS [@jaechung]: we have a public endpoint that we stopped with the fully synced copy at 11778
- 06.06.2018 13:39:37, Craig Murray - EOS BlockSmith [@craigmurray100]: who is running the boot node for it?
- 06.06.2018 13:39:48, Thomas "Disclaimer" Cox (all opinions mine) [@thomasbcox]: Real public keys from the snapshot? Or a blank testnet?
- 06.06.2018 13:39:49, Michael Yeates [@michaelyeates]: its anonymous
- 06.06.2018 13:40:03, Michael Yeates [@michaelyeates]: we may have just found something
- 06.06.2018 13:41:01, Craig Murray - EOS BlockSmith [@craigmurray100]: what are the public endpoints for this paused chain?
- .....06.06.2018 13:44:57, Peter Bue | EOSAmericas: >>invited James Davis<<
- 06.06.2018 13:48:43, Jae Chung - HKEOS [@jaechung]: Craig, we provided you with 6 public endpoints to peer from and how to fully sync through them yesterday. This conversation happened both in the EOS Core channel and Keybase team channel. Must we keep repeating this?
- 06.06.2018 13:49:17, Jae Chung - HKEOS [@jaechung]: Here's a block database you can sync from, for your convenience: https://transfer.sh/OIyZl/boot-mainnet-post-resign-ramfix.tar.gz
- 06.06.2018 13:49:26, Craig Murray - EOS BlockSmith [@craigmurray100]: as I said I'm having trouble with your endpoints today but if you dont want to hel that's fine
- 06.06.2018 13:49:28, Craig Murray - EOS BlockSmith [@craigmurray100]: dont help
- 06.06.2018 13:51:03, Jae Chung - HKEOS [@jaechung]: "Here's a block database you can sync from, for your convenience" <— is this not "help"? I just provided you with the quickest way to sync up to the chain. Not sure why this may be perceived as not wanting to help you.
- 06.06.2018 13:59:51, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Guys looks like @jaechung was helping, there is plenty of other channels to discuss endpoints etc..
- 06.06.2018 14:01:39, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Lets try and keep in here on point. We can spin up another channel for Which Chain type discussions.
- If I'm off topic, I'd expect someone to nudge me back into line, so please don't take offence.
- 06.06.2018 14:05:22, Craig Murray - EOS BlockSmith [@craigmurray100]: 👍
- 06.06.2018 14:06:18, Denis | EOS Nation [@DenisCarriere]: 👍agreed, we should keep the discussions on topic.
- 06.06.2018 14:07:38, Peter Bue | EOSAmericas: 👍👍
- 06.06.2018 14:20:14, Hugo Yang - EOS Silicon Valley [@hugo_eossv]: Tactical Question for 579a chain: if we decide to promote it as the chain voters / BPCs go, how do we deal with all those “dummy” block after injection? A clean up or leave it there?
- 06.06.2018 14:24:14, Denis | EOS Nation [@DenisCarriere]: That could be a discussion topic for tonight (I won't be able to attend, in a plane during call)
- IMO: Having a large amount of empty blocks proves to the community that ZERO transactions have occurred during this validation process, I believe it's a positive for having 60+ hours of empty blocks.
- The filesize of these blocks is minimal, including them isn't a big deal for "launching"
- My vote: Keep empty blocks until we launch
- 06.06.2018 14:28:34, Ben Sigman [@Bensig]: So Denis - is it that what you are suggesting is for us to agree to launch with the chain that we already have running as it stands right now with no changes?
- 06.06.2018 14:28:34, Timothy Lewis [@trustlayer]: [[Webpage]]Updated: https://gist.github.com/redjade/bdbcb10708e165b29d0301c5defa1d83
- 06.06.2018 14:31:19, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Folks I've changed the invite link and removed some people from this group. ( who I didn't recognise or didn't have a BP name in their name )
- If I removed someone from your team, feel free to contact me directly.
- @marcantoineross and @xebb82 are also admins , and people in here should be able to invite others.
- 06.06.2018 14:36:54, deck - EOS New York [@deckw]: Thanks. Do we also want to be careful with who we share the zoom links?
- 06.06.2018 14:39:59, Sharif Bouktila - EOS Dublin [@sharifbouktila]: If anyone has any suggestions on that, please let us know. It needs to be public enough so everyone can get it, but private enough not to spread unnecessarily for now.
- Can obv put passwords but posting here is kind of the same thing.
- I think the message sent out was probably clear enough that until we have a consensus we don't want calls streamed.
- 06.06.2018 14:45:15, Rob Finch - Cypherglass.com (Block Producer) [@finchify]: Happy to do a video. Please send me any details you’d like conveyed.
- 06.06.2018 14:50:58, Thomas "Disclaimer" Cox (all opinions mine) [@thomasbcox]: I had random tourists watching several of my breakout calls the last few days. I gave them three chances to rename themselves with real names plus their BP, then booted them
- 06.06.2018 14:51:56, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Sounds like a good approach.
- 06.06.2018 14:57:28, Kyle - EOSVibes [@TheAwakenment]: I'm also happy to do a video. Although I think this should be Steemed as an EMLG too.
- 06.06.2018 14:59:22, Sharif Bouktila - EOS Dublin [@sharifbouktila]: The Videos are being recorded and I think should either be released or edited (to remove noise, bad audio etc.. ) and released at a later point.
- Personally I don't think I could watch it back, but I know there is lots of interest.
- 06.06.2018 15:03:29, Kyle - EOSVibes [@TheAwakenment]: I meant a video regarding the eosio issuance of tokens to gift the genesis accounts with free ram - that's what was being raised right? Apologies if not, currently catching up on all the Telegram channels - There's like 20 of them! This is becoming a fulltime job in and of iteself!!!
- 06.06.2018 15:05:36, Sharif Bouktila - EOS Dublin [@sharifbouktila]: I hear you! - EOS FIREHOSE
- Personally I don't think the RAM question worth going out with for now. But others might have a different opinion.
- 06.06.2018 15:10:42, Jose Toriello [@josetoriello]: Yup
- 06.06.2018 15:11:52, Jose Toriello [@josetoriello]: Agree. Not worth it. It will only heighten the controversy.
- 06.06.2018 15:12:01, David Moss [@DavidMoss]: @igorls you here?
- 06.06.2018 15:18:44, Igor Lins e Silva - EOS Rio [@igorls]: yes
- 06.06.2018 15:20:24, Sharif Bouktila - EOS Dublin [@sharifbouktila]: @DavidMoss
- 06.06.2018 15:21:01, Ben Sigman [@Bensig]: I just popped up this site
- 06.06.2018 15:21:15, Ben Sigman [@Bensig]: [[Webpage]]https://status.producer.vote/
- 06.06.2018 15:21:53, Sharif Bouktila - EOS Dublin [@sharifbouktila]: You should work in IT , nice work!
- 06.06.2018 15:22:56, Sukesh Tedla (EOSGreen) [@Sukeshtedla]: This is cool. Thanks for your effort.
- 06.06.2018 15:23:17, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Hopefully it will be obsolete soon , although would be a nice place for BPs to push performance stats etc...
- 06.06.2018 15:24:07, Peter Sveda EOS Germany [@PeterSvedaEOSGermany]: 👍
- 06.06.2018 15:24:45, Kyle - EOSVibes [@TheAwakenment]: Helpful. Thanks Ben 👍
- .....06.06.2018 15:25:12, Peter Bue | EOSAmericas: >>invited Douglas Horn | EOSAmericas [@DAPPerD]<<
- 06.06.2018 15:33:59, Rob Finch - Cypherglass.com (Block Producer) [@finchify]: The most important thing to convey is that the tokens are basically burned because they’re locked in the system account forever.
- If people know that the 12k tokens are a *necessary* step in launching the chain, they will be fine with the decision.
- 06.06.2018 15:35:28, Kyle - EOSVibes [@TheAwakenment]: Then is it even worth conveying?
- 06.06.2018 15:47:59, Timothy Lewis [@trustlayer]: But they aren't being burned.
- 06.06.2018 15:48:05, Timothy Lewis [@trustlayer]: So that's not correct information.
- 06.06.2018 15:48:28, Sam - EOS Dublin [@samnoble]: Looks great! Although isn't it the 0100 UTC meeting that's the go/no go? The 1300 UTC is an interim status meeting/validation catchup
- 06.06.2018 15:49:32, Rob Finch - Cypherglass.com (Block Producer) [@finchify]: Probably not.
- 06.06.2018 15:49:38, Timothy Lewis [@trustlayer]: I think the issue being raised is that the initial supply will be wrong. "If they can't even get the initial supply right, what else is wrong?"
- 06.06.2018 15:50:03, mini - EOS Argentina [@moonxor]: We didn't reach consensus on this yesterday?
- 06.06.2018 15:50:15, Timothy Lewis [@trustlayer]: That's becuase there was no fix.
- 06.06.2018 15:50:53, Timothy Lewis [@trustlayer]: We are testing a fix. If it fixes, it will be brought up. If it doesn't we are comfortable launching incorrect.
- 06.06.2018 15:52:52, Kyle - EOSVibes [@TheAwakenment]: 12k are being locked in the System contract forever though right? & There is no way for these EOS to ever be released right?
- 06.06.2018 15:54:17, Timothy Lewis [@trustlayer]: it's actually more like 2.0xxxx that will be moved to an unavailable account. It will affect initial RAM costs and BP rewards slightly.
- 06.06.2018 15:54:32, Timothy Lewis [@trustlayer]: It's not a "critical" system issue by any means.
- 06.06.2018 15:54:50, Timothy Lewis [@trustlayer]: It is an optical issue, in my opinion.
- 06.06.2018 15:56:43, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Why not manually add inflation (4%) for Worker Proposals from when the chain was created. And use that to pay for Ram?
- Put the rest into the WP fund. (We are going to need it)
- 06.06.2018 15:57:38, Sharif Bouktila - EOS Dublin [@sharifbouktila]: I'm also fine with the create & burn option to be clear.
- 06.06.2018 15:57:47, Kyle - EOSVibes [@TheAwakenment]: I'm pretty sure token holders would rather see an ever-so-slight increase in total token supply than to be charged a RAM fee individually (no matter how small).
- 06.06.2018 15:58:18, Kyle - EOSVibes [@TheAwakenment]: Yep, send them to a null account when done.
- 06.06.2018 15:58:25, Timothy Lewis [@trustlayer]: That was never on the table.
- 06.06.2018 15:59:40, Timothy Lewis [@trustlayer]: Currently the options that I see as fixes are the proposed solution from EOSeoul, the option where wecharge an unclaimed account (with a WP to repay that person... )
- 06.06.2018 16:00:07, Timothy Lewis [@trustlayer]: Actually... Why don't you all just charge one of my accounts and we create a WP to repay me.
- 06.06.2018 16:00:16, Timothy Lewis [@trustlayer]: It's like 20k tokens, right?
- 06.06.2018 16:00:39, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Don't like that , but thanks for the offer. - Please vote for us 🙂
- 06.06.2018 16:00:53, Timothy Lewis [@trustlayer]: I'll do anon
- 06.06.2018 16:01:34, Sharif Bouktila - EOS Dublin [@sharifbouktila]: We'll never get a WP approved for a repay. We don't even have a Worker Proposal process designed yet.
- 06.06.2018 16:04:15, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Why not manually add inflation (4%) for Worker Proposals from when the chain was created. And use that to pay for Ram?
- 06.06.2018 16:04:37, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Its like 100k per day
- 06.06.2018 16:04:49, Kyle - EOSVibes [@TheAwakenment]: Why not just increase the total supply to the required level and once launched - send the excess to a null account - It's as good as burning them.
- 06.06.2018 16:06:57, Timothy Lewis [@trustlayer]: I like this.
- 06.06.2018 16:07:38, Timothy Lewis [@trustlayer]: maybe.
- 06.06.2018 16:07:41, Timothy Lewis [@trustlayer]: let me think.
- 06.06.2018 16:14:40, Kyle - EOSVibes [@TheAwakenment]: Whatever the decision, we should communally Steemit as an EMLG post, maybe along with other info/updates at the time. We can video explain it later if needs be.
- 06.06.2018 16:15:02, Timothy Lewis [@trustlayer]: +1
- 06.06.2018 16:15:56, Timothy Lewis [@trustlayer]: And the current workaround has blessings from above... So it's not a huge issue. To me its a huge annoyance becuase I think it will be an optics issue.
- 06.06.2018 16:16:18, Sharif Bouktila - EOS Dublin [@sharifbouktila]: +1
- 06.06.2018 16:16:58, Sharif Bouktila - EOS Dublin [@sharifbouktila]: I think we are over thinking this one.
- Its not material , lost in the roundings after a couple of days of the chain running.
- 06.06.2018 16:21:46, Kyle - EOSVibes [@TheAwakenment]: @trustlayer Yes
- 06.06.2018 16:22:06, Kyle - EOSVibes [@TheAwakenment]: & Yes
- 06.06.2018 16:30:27, Kevin Rose - EOS New York [@eosnewyork]: Agree. Not against @finchify making a video, he’s very good at it. But we need a “faceless” statement as well.
- 06.06.2018 16:31:13, Timothy Lewis [@trustlayer]: With a mask and a voice distorter.
- 06.06.2018 16:31:14, Sharif Bouktila - EOS Dublin [@sharifbouktila]: I don't think anyone should make a video until an agreed period after official statements.
- 06.06.2018 16:31:24, Timothy Lewis [@trustlayer]: +1
- 06.06.2018 16:31:55, Kevin Rose - EOS New York [@eosnewyork]: I can do a mean Darth Vader.
- 06.06.2018 16:47:57, Kyle - EOSVibes [@TheAwakenment]: +1
- Current videos around the launch should be light in detail and content at the moment imo. EOS Token Holders are literally hanging off our every word right now.
- This EMLG community literally amazes me everyday. EOSVibes would be totally honoured if we were to get voted in and work alongside you guys (if). Mucho Respeto!
- I really hope it stays like this...
- 06.06.2018 16:51:20, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Really is a solid group of people, I'm more and more optimistic each day.
- 06.06.2018 16:52:27, Sharif Bouktila - EOS Dublin [@sharifbouktila]: But lets stay on topic ... How is the test plans coming along 🙂
- 06.06.2018 17:03:46, Ben Sigman [@Bensig]: Yes I can change it up
- 06.06.2018 17:04:11, Shaheen Counts - EOS BlockSmith [@SeanPaul19]: I’d like to see this.
- 06.06.2018 17:04:42, Kevin Rose - EOS New York [@eosnewyork]: I'm better at Golem from LOTR. #tulip
- 06.06.2018 17:04:44, Sam - EOS Dublin [@samnoble]: Thanks, the countdown is really helpful as time is just a blur right now 🙂
- 06.06.2018 17:05:37, Shaheen Counts - EOS BlockSmith [@SeanPaul19]: I do a mean golem myself. We
- Might have to have a golem- off at Tulip.
- 06.06.2018 17:06:18, Ben Sigman [@Bensig]: Proof of Vader requested by 0100
- 06.06.2018 17:06:41, Ben Sigman [@Bensig]: Well, I believe that there will be a call tonight
- 06.06.2018 17:06:45, Ben Sigman [@Bensig]: But not to vote
- 06.06.2018 17:07:00, Branden Espinoza [@espinbrand]: I can do Chewbacca.
- 06.06.2018 17:07:14, Timothy Lewis [@trustlayer]: That would be hilarious.
- 06.06.2018 17:07:23, Timothy Lewis [@trustlayer]: With the electronic mask.
- 06.06.2018 17:07:31, Timothy Lewis [@trustlayer]: You'd be perfect for it.
- 06.06.2018 17:07:37, Timothy Lewis [@trustlayer]: (the msg)
- 06.06.2018 17:07:51, Shaheen Counts - EOS BlockSmith [@SeanPaul19]: Haha
- 06.06.2018 17:09:35, Branden Espinoza [@espinbrand]: Oh geeze.
- 06.06.2018 17:16:28, Branden Espinoza [@espinbrand]: I am happy to help however I can. 😁
- 06.06.2018 17:22:00, Kevin Rose - EOS New York [@eosnewyork]: I’ll take “Things No One Wants” for $1,000, Alex.
- 06.06.2018 17:22:17, Shaheen Counts - EOS BlockSmith [@SeanPaul19]: Hahaha
- 06.06.2018 17:24:17, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Thank you @eosnewyork , see you on the other side
- 06.06.2018 17:44:59, David P - EOS42 [@GunnisonCap]: Right approach 👍
- 06.06.2018 17:56:33, Josh Kauffman - EOS Canada [@joshkauff]: Nice to see all the arguing has now been focused on who can do a better Golem impression. I think we need EOS Go to be a neutral judge in this though.
- 06.06.2018 18:06:26, Sharif Bouktila - EOS Dublin [@sharifbouktila]: The natural endpoint of all this is a dance off.
- I trust Kev and Blue Jays to be neutral in judging.
- 06.06.2018 18:08:27, Stuardo Rodríguez 🇬🇹 [@stuardo_str]: In case we find a p0 or p1 issue, and we do need to start a new chain, would it be OK to reconsider how to alocate RAM?
- 06.06.2018 19:58:12, Roman CryptoLions [@romanskaskiw]: zoom call coming up?
- 06.06.2018 19:58:37, Nathan Rempel | GenerEOS Sydney [@nsrempel]: in one hour i think
- 06.06.2018 20:01:10, Kevin Rose - EOS New York [@eosnewyork]: I should start stretching now then.
- 06.06.2018 20:09:33, Marc-Antoine Ross - EOS Canada [@marcantoineross]: In one hour yes
- 06.06.2018 20:16:08, Nathan Rempel | GenerEOS Sydney [@nsrempel]: I see that eos-mainnet is 80 commits behind eosio
- 06.06.2018 20:20:02, Sam - EOS Dublin [@samnoble]: [[Webpage]]@Bensig created a handy tracker for exactly that! https://status.producer.vote
- .....06.06.2018 20:35:15, Marc-Antoine Ross - EOS Canada [@marcantoineross]: >>invited Celu - Blockgenic [@celur]<<
- .....06.06.2018 20:35:53, Celu - Blockgenic [@celur]: >>invited Jk (Blockgenic) [@topcat2001]<<
- 06.06.2018 20:41:34, Peter Bue | EOSAmericas: ^ I'd also like to confirm the call.
- 06.06.2018 20:48:50, Craig Murray - EOS BlockSmith [@craigmurray100]: yes, but I'm not sure it needs to track master. I suspect they will merge based on new tags and certainly new versioned releases
- 06.06.2018 20:49:03, Nathan Rempel | GenerEOS Sydney [@nsrempel]: Fair enough
- 06.06.2018 20:50:27, Andrey Marchuk [@madz7z]: and David confirmed on the morning call that there are no critical fixes to worry about
- 06.06.2018 20:57:14, Marc-Antoine Ross - EOS Canada [@marcantoineross]: [[Webpage]]All,
- Here's the proposed agenda:
- Security report update - Dafeng EOS Asia
- ABP selection update - Kevin EOS New York
- go / no go
- Ben from shEOS asks: How can we insure that voting portals are handling private keys well?
- Create a sub-team to prepare recommendations?
- https://zoom.us/j/885770550
- .....06.06.2018 20:57:22, Marc-Antoine Ross - EOS Canada [@marcantoineross]: >>unsupported service message type: messageActionPinMessage<<
- 06.06.2018 20:57:42, Timothy Lewis [@trustlayer]: Time?
- 06.06.2018 20:57:50, David Moss [@DavidMoss]: 6pm
- 06.06.2018 20:57:53, Marc-Antoine Ross - EOS Canada [@marcantoineross]: Now!
- 06.06.2018 20:58:14, Marc-Antoine Ross - EOS Canada [@marcantoineross]: It's always 0100UTC and 1300UTC until further notice
- 06.06.2018 20:59:29, Craig Murray - EOS BlockSmith [@craigmurray100]: FOREVER!
- 06.06.2018 21:00:13, Timothy Lewis [@trustlayer]: 🙂
- 06.06.2018 21:08:17, Bo - EOS Silicon Valley [@bo_eossv]: [[Document, size 195'930 bytes]]
- 06.06.2018 21:15:47, James Sutherland - Cypherglass.com [@JamesSutherland]: [[Document, size 719'171 bytes]]
- 06.06.2018 21:17:22, Andrey Marchuk [@madz7z]: thanks
- 06.06.2018 21:18:04, James Sutherland - Cypherglass.com [@JamesSutherland]: Google mangled it, but it's somewhat readable
- 06.06.2018 21:18:24, Nathan Rempel | GenerEOS Sydney [@nsrempel]: the github translates pretty well
- 06.06.2018 21:25:00, Devin Lee KOREOS.IO [@indend007]: Already reported by securty firm?
- 06.06.2018 21:30:33, James Sutherland - Cypherglass.com [@JamesSutherland]: Preliminary, just 3 pages. More to come
- 06.06.2018 21:32:09, Devin Lee KOREOS.IO [@indend007]: Thanks.
- 06.06.2018 22:00:48, Craig Murray - EOS BlockSmith [@craigmurray100]: another highly productive and this time extremely effecient meeting, thanks everyone!
- 06.06.2018 22:01:59, Nathan Rempel | GenerEOS Sydney [@nsrempel]: Have you been able to successfully execute these changes using the boot bios? I'm working on a similar fix now (not because i think it's needed, but just as a test) but day job conflicts
- 06.06.2018 22:02:13, Nathan Rempel | GenerEOS Sydney [@nsrempel]: Yes agreed
- 06.06.2018 22:02:55, Nathan Rempel | GenerEOS Sydney [@nsrempel]: @marcantoineross If i can propose an item for next meetings agenda, review the current state of the UAT. Are people supplying proof of tests, adding tests, running tests
- 06.06.2018 22:05:22, Timothy Lewis [@trustlayer]: I'll DM you. May not fully solve the entire issue.
- 06.06.2018 22:05:32, Nathan Rempel | GenerEOS Sydney [@nsrempel]: 👍
- 06.06.2018 22:15:05, Tim Weston - GenerEOS (BP Sydney) [@timweston]: Hey guys, did we get a channel created to discus the voting portal etc ?
- 06.06.2018 22:16:06, Tim Weston - GenerEOS (BP Sydney) [@timweston]: If not I’m happy to throw one up
- 06.06.2018 22:18:44, Marc-Antoine Ross - EOS Canada [@marcantoineross]: [[Webpage]]All,
- Thanks for another productive call!
- Find bellow the minutes, if you find anything not accurate, please suggest/comment and we will fix.
- The raw log is included at the bottom of the document.
- https://docs.google.com/document/d/1HR16WF5nhNXNzkBh963ZKjSflsyBsMb0PsOhHSbQTsI
- Next steps June 7, 1300 UTC:
- Security report update - Dafeng EOS Asia
- ABP selection update - Kevin EOS New York
- Follow-up constitution
- Review the current state of the UAT testing. Are people supplying proof of tests, adding tests, running tests
- More?
- Thanks, good after-mor-night!
- .....06.06.2018 22:18:50, Marc-Antoine Ross - EOS Canada [@marcantoineross]: >>unsupported service message type: messageActionPinMessage<<
- 06.06.2018 22:24:17, Jk (Blockgenic) [@topcat2001]: Let me know if you do.
- 06.06.2018 22:28:43, Zhao Yu@ EOSLaoMao.com [@zhaoyu]: and the group?
- 06.06.2018 22:29:00, Nathan Rempel | GenerEOS Sydney [@nsrempel]: Tim has to up his Telegram game 🤣
- 06.06.2018 22:29:41, Branden Espinoza [@espinbrand]: Tim, you ceated a channel, not a group. That is for 1 way comms only.
- 06.06.2018 22:32:11, Tim Weston - GenerEOS (BP Sydney) [@timweston]: whoops will create now
- 06.06.2018 22:35:24, Nathan Rempel | GenerEOS Sydney [@nsrempel]: https://t.me/joinchat/IbfNrxA6yqKbzBXSXZxteA
- 06.06.2018 22:36:01, Tim Weston - GenerEOS (BP Sydney) [@timweston]: 👆🏻👆🏻 here it is guys!
- 06.06.2018 23:30:26, Nathan Rempel | GenerEOS Sydney [@nsrempel]: Some fun messages from @daniellarimer
- 06.06.2018 23:30:26, Nathan Rempel | GenerEOS Sydney [@nsrempel]: {{FWD: Daniel Larimer [@daniellarimer], 06.06.2018 23:12:38}}
- I like how serious everyone is, but perfection is enemy of good enough
- 06.06.2018 23:30:27, Nathan Rempel | GenerEOS Sydney [@nsrempel]: {{FWD: Daniel Larimer [@daniellarimer], 06.06.2018 23:16:27}}
- No, we wil run and live!
- William Wallace: Yes!, Fight and you may die. Run and you will live at least awhile. And dying in your bed many years from now, would you be willing to trade all the days from this day to that for one chance, just one cahnce, to come back here as young men and tell our enemies that they may take our lives but they will never take our freedom!!!
- 06.06.2018 23:30:27, Nathan Rempel | GenerEOS Sydney [@nsrempel]: {{FWD: Daniel Larimer [@daniellarimer], 06.06.2018 23:19:25}}
- I see people petrified by fear of unknown bugs. Life is messy. Be brave. We can overcome anything
- 06.06.2018 23:30:28, Nathan Rempel | GenerEOS Sydney [@nsrempel]: {{FWD: Daniel Larimer [@daniellarimer], 06.06.2018 23:20:59}}
- Fortune favors the bold. And voters will notice.
- 06.06.2018 23:30:28, Nathan Rempel | GenerEOS Sydney [@nsrempel]: {{FWD: Daniel Larimer [@daniellarimer], 06.06.2018 23:28:17}}
- I cannot promise you there will be no bugs or no losses. But each day people lose money to scams and new risks Mount. The rag tag teams of producers on steem and bitshares can do this. Surely the professionals here can.
- 06.06.2018 23:30:53, Kyunghwan Kim - EOSeoul.io [@resvera]: Thank you! 👍
- 07.06.2018 00:37:29, Ben Sigman [@Bensig]: [[Photo]]
- 07.06.2018 01:20:11, Matthew Darwin - EOS Nation [@mdarwin]: Marc, thanks for the great notes!
- 07.06.2018 01:38:17, Roshan - EOS Authority [@telepound]: Thanks for the notes!
- 07.06.2018 03:27:23, Sandwich [@sandw1ch]: 🔥
- .....07.06.2018 05:04:13, Denis | EOS Nation [@DenisCarriere]: >>invited Yves La Rose - EOS Nation [@YBNorml]<<
- 07.06.2018 05:45:46, Roman CryptoLions [@romanskaskiw]: nice
- 07.06.2018 08:26:08, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Next Validation Call is in 35 minutes
- Security report update - Dafeng EOS Asia
- ABP selection update - Kevin EOS New York
- Follow-up constitution
- Review the current state of the UAT testing. Are people supplying proof of tests, adding tests, running tests
- More?
- Either @marcantoineross or I will post Zoom link 5 minutes before call.
- 07.06.2018 08:46:28, Tim Weston - GenerEOS (BP Sydney) [@timweston]: 👍
- 07.06.2018 08:54:27, Marc-Antoine Ross - EOS Canada [@marcantoineross]: [[Webpage]]All,
- Here's the proposed agenda:
- Security report update - Dafeng EOS Asia
- ABP selection update - Kevin EOS New York
- Follow-up constitution
- Review the current state of the UAT testing. Are people supplying proof of tests, adding tests, running tests
- More?
- https://zoom.us/j/438337388
- .....07.06.2018 08:54:32, Marc-Antoine Ross - EOS Canada [@marcantoineross]: >>unsupported service message type: messageActionPinMessage<<
- 07.06.2018 08:57:03, Thomas "Disclaimer" Cox (all opinions mine) [@thomasbcox]: Reword "Follow-up constitution" to "follow-up on Constitution Article 9 Rollback Proposal" please?
- 07.06.2018 08:57:58, Marc-Antoine Ross - EOS Canada [@marcantoineross]: done in the zoom chat
- .....07.06.2018 09:01:30, Marc-Antoine Ross - EOS Canada [@marcantoineross]: >>invited Miłosz Eos Emerge - Poland [@Om4tic]<<
- 07.06.2018 09:03:16, Brett Jeffery @ EOS the World [@EOStheWorld]: 👍
- .....07.06.2018 09:04:30, Elliot@EOS Pacific [@Elliot521]: >>invited jason wang [@Jason_wang]<<
- 07.06.2018 09:16:49, Nathan Rempel | GenerEOS Sydney [@nsrempel]: [[Webpage]]https://docs.google.com/spreadsheets/d/1gJnn0KS4rgDnTowPlN80gulVI77lRgldPhH17kZOnnI
- 07.06.2018 09:17:12, Nathan Rempel | GenerEOS Sydney [@nsrempel]: https://t.me/joinchat/IbfNrxA6yqKbzBXSXZxteA
- 07.06.2018 09:25:12, Kyunghwan Kim - EOSeoul.io [@resvera]: [[Webpage]]Being late, but finally got tested version of contracts for adjusting 1) token supply, 2) balance of system accounts, and 3) bancor state.
- https://github.com/eoseoul/eos-ramcost
- 07.06.2018 09:31:05, Nathan Rempel | GenerEOS Sydney [@nsrempel]: looks good to me, nice and clean. we''ll need this tested in boot sequence asap (12 hours)
- 07.06.2018 09:49:21, Alexandre Bourget - EOS Canada [@abourget]: There is a problem with the .abi files.. if at any time the ABI files do not contain all the methods from all the contracts that take place in the eosio account.. we will have problems fetching those blocks.
- 07.06.2018 09:50:29, Nathan Rempel | GenerEOS Sydney [@nsrempel]: i looked at the cpps :(
- 07.06.2018 09:50:39, Nathan Rempel | GenerEOS Sydney [@nsrempel]: good catch
- 07.06.2018 09:51:52, Alexandre Bourget - EOS Canada [@abourget]: That was already voted on though..
- 07.06.2018 09:52:08, Nathan Rempel | GenerEOS Sydney [@nsrempel]: ok, i'll shelf it
- 07.06.2018 09:56:08, Thomas "Disclaimer" Cox (all opinions mine) [@thomasbcox]: [[Webpage]]Strongly encourage all BPs who have done testing, to log their tests here:
- https://docs.google.com/spreadsheets/d/1dWqr9Izk4b2xJnoS94nMD0aKR6UU3Xssy7iowwtcxBQ/edit#gid=1501553294
- 07.06.2018 10:09:48, Kyunghwan Kim - EOSeoul.io [@resvera]: All actions in eosio.system.ramcost and eosio.token.ramcost are implemented there. Am I right in thinking that you mean that after replacing to original contracts removed actions cause problems?
- 07.06.2018 10:23:32, Ben Sigman [@Bensig]: As I understand, there were no viable alternatives. Now that there is, don’t we want to offer it to the community?
- 07.06.2018 10:31:11, Kyunghwan Kim - EOSeoul.io [@resvera]: If there is missing function in eosio.system.ramcost and eosio.token.ramcost with respective abi files, please tell me. It would be quite helpful.
- 07.06.2018 10:35:33, Alexandre Bourget - EOS Canada [@abourget]: This proposition is quite complex (technically for code review, boot_seq adjustements, validation tools adjustments, and socially for getting agreement), seeking to adjust a 1/10 of a day's inflation fluctuation.
- I wouldn't go to war for this.
- 07.06.2018 10:36:21, Nathan Rempel | GenerEOS Sydney [@nsrempel]: Concur at this stage
- 07.06.2018 10:39:06, Ren Kaidi - EOSWTZ [@FoxMr]: so, who can give a summary about what to do tomorrow?
- 07.06.2018 10:39:17, Ren Kaidi - EOSWTZ [@FoxMr]: thanks
- 07.06.2018 10:40:13, Kyunghwan Kim - EOSeoul.io [@resvera]: Okay. eos-ramcost is fully tested in my local environment. That is to explain and present tools that people requested and to get feedbacks. I have no intention to argue.
- 07.06.2018 10:40:25, Nathan Rempel | GenerEOS Sydney [@nsrempel]: instructions will be available pending the result of the go / no meeting
- 07.06.2018 10:40:45, Nathan Rempel | GenerEOS Sydney [@nsrempel]: I think it's fantastic work, the timing is just difficult
- 07.06.2018 10:41:08, Kyunghwan Kim - EOSeoul.io [@resvera]: I have same opinion with you. :)
- 07.06.2018 10:41:23, Nathan Rempel | GenerEOS Sydney [@nsrempel]: I think anyone helping to solve issues for the launch and the community should be commended
- 07.06.2018 10:41:27, Ren Kaidi - EOSWTZ [@FoxMr]: thanks
- 07.06.2018 10:49:20, deck - EOS New York [@deckw]: Everyone. I am treating the 4th criteria from the spreadsheet to mean that you need to send an overview of your security/tech specs to me (please keep it to 3-4 sentences. I do not believe we have time for me or someone else to do a full security audit. Is that sufficient for everyone?
- 07.06.2018 10:49:26, deck - EOS New York [@deckw]: [[Webpage]]https://docs.google.com/spreadsheets/d/1t4p1tWs0fXf28nF2w4m5wwzO1cAknYs5XQGJLAIf1pU/edit#gid=753803902
- 07.06.2018 10:52:02, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Yes
- 07.06.2018 10:52:09, Nathan Rempel | GenerEOS Sydney [@nsrempel]: Fine for us
- 07.06.2018 10:53:01, Nate D - Aloha EOS [@n8dee]: 👍
- .....07.06.2018 11:02:34, Miłosz Eos Emerge - Poland [@Om4tic]: >>invited Krzysztof Manijak - EOSEmerge.io - Poland [@true_manner]<<
- 07.06.2018 11:12:12, Timothy Lewis [@trustlayer]: Awesome. @jemxpat @jaechung
- 07.06.2018 11:13:23, Timothy Lewis [@trustlayer]: This was really well done. I commend you for the effort, regardless of outcome.
- 07.06.2018 11:13:54, David Moss [@DavidMoss]: +1
- 07.06.2018 11:14:14, Marc-Antoine Ross - EOS Canada [@marcantoineross]: [[Webpage]]All,
- Thanks for an amazing call!
- Find bellow the minutes, if you find anything not accurate, please suggest/comment and we will fix.
- The todo for next call is included in the minutes
- The raw log is included at the bottom of the document.
- https://docs.google.com/document/d/1Fpevju50lpXDH50dMH0A4NOUnUioHb7xYy76RQXG1w4
- Next call June 8, 0100 UTC
- Thanks, good after-mor-night!
- .....07.06.2018 11:14:20, Marc-Antoine Ross - EOS Canada [@marcantoineross]: >>unsupported service message type: messageActionPinMessage<<
- 07.06.2018 11:14:53, David Moss [@DavidMoss]: Thanks @marcantoineross and team!
- 07.06.2018 11:20:44, Jose Toriello [@josetoriello]: Someone leaked minutes...
- 07.06.2018 11:20:44, Jose Toriello [@josetoriello]: {{FWD: Eduardo Cuevas, 07.06.2018 11:10:23}}
- [[Webpage]]https://www.trustnodes.com/2018/06/06/eos-launch-delayed-pending-audit-results
- 07.06.2018 11:23:26, Nathan Rempel | GenerEOS Sydney [@nsrempel]: neat we can reverse blocks at will
- 07.06.2018 11:23:32, Nathan Rempel | GenerEOS Sydney [@nsrempel]: apparently
- 07.06.2018 11:27:20, Craig Murray - EOS BlockSmith [@craigmurray100]: @eosnewyork is one of the questions we're asking the abp candidates to provide "what method of injection are you using and is it publicly available" ? It's not yet clear to me how the thing is going to be set up. This is just up to the abp's discretion to make something himself or using something pubicly available?
- 07.06.2018 11:27:57, Craig Murray - EOS BlockSmith [@craigmurray100]: @DavidMoss perhaps you have comment on this as well?
- 07.06.2018 11:28:22, Nathan Rempel | GenerEOS Sydney [@nsrempel]: don't think its possible for @eosnewyork to make a decision based on boot method
- 07.06.2018 11:28:39, Kevin Rose - EOS New York [@eosnewyork]: I have shoddy service. On a plane. Enter any criteria you think is valuable into the google doc. We will go through to create a final list. David Moss will gave input and it will be pushed back out to the community.
- 07.06.2018 11:28:40, Nathan Rempel | GenerEOS Sydney [@nsrempel]: as long as the chain is validated it could be hand bombed i think
- 07.06.2018 11:28:43, Craig Murray - EOS BlockSmith [@craigmurray100]: we've argued about ram for a while now, do we even know what method one of these 16 people will use?
- 07.06.2018 11:29:00, Kevin Rose - EOS New York [@eosnewyork]: I don’t think that is up to the ABP.
- 07.06.2018 11:29:24, Kevin Rose - EOS New York [@eosnewyork]: Have we tested JEMs RAM option that rounds to a perfect 1BB?
- 07.06.2018 11:29:37, Michael Yeates [@michaelyeates]: we have that on our chain now
- 07.06.2018 11:29:46, Craig Murray - EOS BlockSmith [@craigmurray100]: what is the github link?
- 07.06.2018 11:29:52, Michael Yeates [@michaelyeates]: to our chain?
- 07.06.2018 11:30:02, Craig Murray - EOS BlockSmith [@craigmurray100]: to the tool used to inject it, JEM's option
- 07.06.2018 11:30:14, Michael Yeates [@michaelyeates]: ill get it for you
- 07.06.2018 11:30:27, Kevin Rose - EOS New York [@eosnewyork]: We have until 1600 UTC to input ABP required criteria.
- 07.06.2018 11:30:46, Kevin Rose - EOS New York [@eosnewyork]: I don’t want this process to hold up a go/no-go decision at 0100
- 07.06.2018 11:31:02, Jae Chung - HKEOS [@jaechung]: [[Webpage]]https://github.com/HKEOS/Ghostbusters-Testnet/tree/master/chain-validator
- was sent ~3 days ago
- 07.06.2018 11:31:41, Craig Murray - EOS BlockSmith [@craigmurray100]: it's like 90% of what the ABP does... the other 10% staying alive... and there has been heaps of fighting about ram solutions, just leaving it up to the abp chosen seems like a good way to continue combat, and also for the selected abp to have to deal with things while remaining anon
- 07.06.2018 11:32:27, Kevin Rose - EOS New York [@eosnewyork]: Like I said, I think that any ABP that is applying for this position is consenting to the community approved process.
- 07.06.2018 11:32:43, Craig Murray - EOS BlockSmith [@craigmurray100]: then we need a community approved process pretty quick!
- 07.06.2018 11:33:22, Kevin Rose - EOS New York [@eosnewyork]: Put what you want into the criteria list. The best way to solicit consensus is when that list is final and pushed back out to the community at 1600 UTC.
- 07.06.2018 11:37:04, Craig Murray - EOS BlockSmith [@craigmurray100]: 16 candidates, i really like everyone's entheusiasm. Personally I think there is a very short list of people that should be doing this. If the anon ABP does not get a valid chain the first try, it undermines this entire process and sets us back. Do we let them try again, or choose another person?
- 07.06.2018 11:37:36, Sheldon | EOShenzhen [@SheldonXP]: this is really awesome
- 07.06.2018 11:37:40, Craig Murray - EOS BlockSmith [@craigmurray100]: I have added criteria
- 07.06.2018 11:43:05, Kevin Rose - EOS New York [@eosnewyork]: If they fail I propose we go back to the list of applicants. But that’s just my opinion.
- 07.06.2018 11:47:02, Craig Murray - EOS BlockSmith [@craigmurray100]: thx
- 07.06.2018 11:52:32, Alex - EOS SOCAL [@gmory]: Agreed
- 07.06.2018 11:53:29, Craig Murray - EOS BlockSmith [@craigmurray100]: well there is a bit of an elephant in the room, so I'm just trying to shine some light on it. Who does the injection and how it's done has been a massive fighting point
- 07.06.2018 11:53:58, Craig Murray - EOS BlockSmith [@craigmurray100]: i recommend we figure that out before the abp is chosen
- 07.06.2018 11:55:25, Sukesh Tedla (EOSGreen) [@Sukeshtedla]: Where can one find the list of ABP candidates? I cannot find that in the document, should we add it to the doc?
- 07.06.2018 11:56:14, Kevin Rose - EOS New York [@eosnewyork]: It is secret to avoid security risks and merchandizing the responsibility to the token holders.
- .....07.06.2018 11:56:45, Eric_EOSARMY.io [@Eric_EOSARMY]: >>invited phajek eosarmy Patrick [@Phajek]<<
- 07.06.2018 11:56:48, Kevin Rose - EOS New York [@eosnewyork]: The BPCs present on the previous 1300 UTC asked that EOS New York field and manage the list with David Moss as a witness.
- .....07.06.2018 11:57:48, Miłosz Eos Emerge - Poland [@Om4tic]: >>invited Marek EOSEmerge.io - Poland [@mbaclawski]<<
- 07.06.2018 11:58:11, Nate D - Aloha EOS [@n8dee]: I don't think the how matters, as long as the data state matches what everyone expects and tests pass.
- 07.06.2018 11:58:38, Craig Murray - EOS BlockSmith [@craigmurray100]: so we leave it up to the abp to decide then?
- 07.06.2018 11:59:22, Craig Murray - EOS BlockSmith [@craigmurray100]: we have suspended radical transparency and the list is hidden, as will be the name of the anon abp
- 07.06.2018 11:59:24, Sukesh Tedla (EOSGreen) [@Sukeshtedla]: Sorry! I wasn't able to join the meeting. I don't know what was discussed but we wanted to go with the consensus decision.
- 07.06.2018 11:59:50, Nate D - Aloha EOS [@n8dee]: The process they use to boot? Yes, I think so.
- 07.06.2018 12:03:09, Craig Murray - EOS BlockSmith [@craigmurray100]: the tool they used to inject the snapshot, purchase the ram, setup the b1 account, resign the eosio account, validate the steps taken, and im sure there are other things. The method used will affect the result and we then validate that... but we dont have consensus on the method of doing such, so how can we validate? This has been a major point of contention, that's my broader point. But not figuring out what we plan to do, either choose a method, leave it up to the abp, whatever, we need to make sure everyone is happy because it could lead to lack on consensus on validity. Not to mention it would be nice to be testing the various tools these groups might use to make sure they work so the abp isn't stuck being anon with a broken tool
- 07.06.2018 12:04:22, Nate D - Aloha EOS [@n8dee]: We already voted on how to handle the ram purchase and tokens, so I would expect the ABP to stick to that unless we vote again to change it.
- 07.06.2018 12:05:01, Sukesh Tedla (EOSGreen) [@Sukeshtedla]: +1
- 07.06.2018 12:05:04, Craig Murray - EOS BlockSmith [@craigmurray100]: there are 2 tools that I know of, Igor has one that uses eosjs. Alex has one that uses golang.
- 07.06.2018 12:05:23, Craig Murray - EOS BlockSmith [@craigmurray100]: did we? I'm sorry I'm trying to catch up on the details i missed some of the meeting
- 07.06.2018 12:05:41, Craig Murray - EOS BlockSmith [@craigmurray100]: and this method is currently expressed in Igor's repo that was linked earlier?
- 07.06.2018 12:06:48, Nate D - Aloha EOS [@n8dee]: Yes, I'd have to dig up when exactly but it was 2-3 meetings ago.
- 07.06.2018 12:07:29, Craig Murray - EOS BlockSmith [@craigmurray100]: forgive me if i'm missing the obvious, trying to quickly review this code. I don't see the b.1 account setup in that code. Am I missing it?
- 07.06.2018 12:08:13, Craig Murray - EOS BlockSmith [@craigmurray100]: I'm pretty sure we did not have consensus on a ram buying method 2-3 meetings ago. I haven't missed a meeting yet until this morning and people have been debating it more recently
- 07.06.2018 12:08:25, Craig Murray - EOS BlockSmith [@craigmurray100]: but please if you have minutes that would be great, i would be thrilled if this was actually already resolved
- 07.06.2018 12:09:18, Nate D - Aloha EOS [@n8dee]: Yeah I'm looking...
- 07.06.2018 12:10:12, Nate D - Aloha EOS [@n8dee]: [[Webpage]]https://docs.google.com/document/d/1nHEA3rfWLh4uwKmFnY9ypuRnWbOjuSy0LdKUlTapS2c/edit
- 07.06.2018 12:12:47, Kevin Rose - EOS New York [@eosnewyork]: Can someone share the testing doc with with Dan Larimer?
- 07.06.2018 12:14:02, Craig Murray - EOS BlockSmith [@craigmurray100]: [[Webpage]]https://docs.google.com/spreadsheets/d/1t4p1tWs0fXf28nF2w4m5wwzO1cAknYs5XQGJLAIf1pU/edit#gid=668547189
- 07.06.2018 12:14:10, Craig Murray - EOS BlockSmith [@craigmurray100]: i've sent him that one, i think there is also another one
- 07.06.2018 12:14:17, Kevin Rose - EOS New York [@eosnewyork]: Thank you
- 07.06.2018 12:14:39, Craig Murray - EOS BlockSmith [@craigmurray100]: there is also Thomas's one, but i honestly haven't had time to look at it, i dont have a link handy
- 07.06.2018 12:15:18, Josh Kauffman - EOS Canada [@joshkauff]: [[Webpage]]https://docs.google.com/spreadsheets/d/1dWqr9Izk4b2xJnoS94nMD0aKR6UU3Xssy7iowwtcxBQ/edit#gid=1501553294
- 07.06.2018 12:15:26, Josh Kauffman - EOS Canada [@joshkauff]: ^^ Daniel
- 07.06.2018 12:16:09, Josh Kauffman - EOS Canada [@joshkauff]: On a previous call, there was a vote for this and it was chosen that creating the extra EOS made the most sense and the ABP should go with that.
- 07.06.2018 12:17:16, Craig Murray - EOS BlockSmith [@craigmurray100]: Ok so @jemxpat is ok with this? @jaechung and @michaelyeates and @igorls and @xebb82 and @trustlayer are all ok with this?
- 07.06.2018 12:17:45, Craig Murray - EOS BlockSmith [@craigmurray100]: to be clear i do not care at all what method is used I'm just trying to avoid conflict after an abp is chosen
- 07.06.2018 12:26:08, Michael Yeates [@michaelyeates]: please can we put this on the agenda for the next meeting
- 07.06.2018 12:31:14, Marc-Antoine Ross - EOS Canada [@marcantoineross]: [[Webpage]]How the chain is booted does not matter as much as validation of its content.
- We have consensus on the content of the chain, as per the minutes. Look for "Question (21:50:00)" and the notes following.
- https://docs.google.com/document/d/1nHEA3rfWLh4uwKmFnY9ypuRnWbOjuSy0LdKUlTapS2c
- 07.06.2018 12:36:49, Craig Murray - EOS BlockSmith [@craigmurray100]: ok. To be clear, my understanding is now that we have consensus that eos.io should issue an extra approx 19k tokens and they be used to fund the accounts. This means a chain which does not do so is invalid?
- 07.06.2018 12:43:18, Marc-Antoine Ross - EOS Canada [@marcantoineross]: Correct
- 07.06.2018 12:44:14, Yves La Rose - EOS Nation [@YBNorml]: This was the consensus yesterday, yes.
- 07.06.2018 12:45:27, Timothy Lewis [@trustlayer]: I like this.
- 07.06.2018 12:50:15, Timothy Lewis [@trustlayer]: If a solution is found and presented, it may be considered. I do not know if we could reach consensus on the change, but @resvera has created a working, tested solution. It needs to be further audited, but it appears to be ready to go.
- 07.06.2018 12:53:06, Craig Murray - EOS BlockSmith [@craigmurray100]: Per the previous information it is not considered a problem, and any solution would result in an invalid chain. But I understand this is not a unanimous opinion which is why I'm trying to draw out this debate now, before an abp is selected. I do not want someone spinning up a chain without 19k extra eos and having that be an invalid chain because it does not respect the rule imposed by the vote outlined previously
- 07.06.2018 12:58:36, mini - EOS Argentina [@moonxor]: We can't keep delaying things if we already reach consensus. Yes, it would be nice to not send extra tokens to an account with a invalid key to "burn " them. but we reach consensus :)
- We should delay thing only for critical things. Otherwise we are in an infinite loop reaching for perfection
- 07.06.2018 13:01:45, Sharif Bouktila - EOS Dublin [@sharifbouktila]: We know how to validate the chain using a variety of tools.
- Let @eosnewyork and @DavidMoss figure it out.
- If that chain gets to 15% we are in business. If not we start again.
- 07.06.2018 13:02:30, Denis | EOS Nation [@DenisCarriere]: {{FWD: Daniel Larimer [@daniellarimer], 07.06.2018 12:11:53}}
- We are preparing a 1.0.2 update that will fix things with the optional whitelist/blacklist for contracts
- 07.06.2018 13:02:42, Craig Murray - EOS BlockSmith [@craigmurray100]: we are about to validate another chain, we cannot rely on NY and David to figure that out. We must have overwhelming consensus that we are valid and right now I do not know if we can achieve that based on disputes about RAM purchasing methods
- 07.06.2018 13:03:18, Craig Murray - EOS BlockSmith [@craigmurray100]: yup, and then there's that to add to the mix 😄
- 07.06.2018 13:03:50, Craig Murray - EOS BlockSmith [@craigmurray100]: this is why i'm pressing this issue because it needs resolved quickly so that it does not disrupt our existing timelines. more delays are not in our best interest
- 07.06.2018 13:07:27, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Ok - We have a process , I suggest we put it on the agenda at 01:00 UTC
- @jemxpat @michaelyeates @trustlayer @abourget @marcantoineross
- I propose result of next vote is set in stone.
- Agreed?
- 07.06.2018 13:08:09, Craig Murray - EOS BlockSmith [@craigmurray100]: i think the last one is set in stone... im just trying to get those who do not agree that it is in stone to speak up
- 07.06.2018 13:08:30, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Reason why , is the first time it was voted it was the first call we did that type of decision making, on something this decisive I think it’s ok to revisit
- 07.06.2018 13:08:44, Jae Chung - HKEOS [@jaechung]: @abourget you were explaining the math for minimum difference in head block and LIB with 21 producers earlier. Could you forward the message from Bart or copy&paste that here again?
- 07.06.2018 13:09:43, Sharif Bouktila - EOS Dublin [@sharifbouktila]: We are starting a new chain anyway aren’t we?
- 07.06.2018 13:11:29, Jae Chung - HKEOS [@jaechung]: this is something important for all of us to know for mainnet— it means minimum catch up time for LIB is ~3 minutes
- 07.06.2018 13:12:01, Denis | EOS Nation [@DenisCarriere]: Since we are already doing a reset on the chain, it might be something to consider.
- I have no preference, 1.0.1 or 1.0.2 is fine with me (preference to 1.0.2 if we have concensus)
- 07.06.2018 13:13:20, Jae Chung - HKEOS [@jaechung]: so you do have a preference ;)
- 07.06.2018 13:13:29, Denis | EOS Nation [@DenisCarriere]: Yes, new chain is required because of constitution changes (for legal reasons)
- 07.06.2018 13:13:42, Denis | EOS Nation [@DenisCarriere]: I vote 👍 for both
- 07.06.2018 13:13:57, mini - EOS Argentina [@moonxor]: 21*12*2/3
- 07.06.2018 13:14:26, mini - EOS Argentina [@moonxor]: @jaechung
- 07.06.2018 13:14:45, Denis | EOS Nation [@DenisCarriere]: It's a patch update, no feature changes, so it shouldn't be an issue if we do or do not use it
- 07.06.2018 13:15:16, Sharif Bouktila - EOS Dublin [@sharifbouktila]: I agree we should, but don’t quite understand that as we will be able to change constitution later.
- 07.06.2018 13:15:26, Denis | EOS Nation [@DenisCarriere]: Jae, do you have time to look into it? I'll support your judgement
- 07.06.2018 13:16:02, Jae Chung - HKEOS [@jaechung]: yep, we're looking into it now but at the moment I agree with you— preference for 1.0.2
- 07.06.2018 13:16:15, Craig Murray - EOS BlockSmith [@craigmurray100]: only before 15%, and even then it's a little sketchy... we're validating the chain, everything. Once that's done it can't change until 15% and voter ref
- 07.06.2018 13:16:31, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Thanks
- 07.06.2018 13:16:36, Denis | EOS Nation [@DenisCarriere]: We are recompiling the eosio.system ABI, this needs to be loaded in the initial boot sequence.
- 07.06.2018 13:16:38, Craig Murray - EOS BlockSmith [@craigmurray100]: if we're doing a new chain, 1.0.2 yes 👍
- 07.06.2018 13:17:23, Denis | EOS Nation [@DenisCarriere]: 👍 perfect
- 07.06.2018 13:18:08, Jae Chung - HKEOS [@jaechung]: Thanks. On the acceptance testing chain the LIB is ~325 blocks behind head block num— so I'm assuming this is what we'll have on mainnet 😕
- 07.06.2018 13:19:33, mini - EOS Argentina [@moonxor]: Yes shouldn't be 325, don't know if has anything to do with the boot process.
- 07.06.2018 13:19:56, mini - EOS Argentina [@moonxor]: Or a bug :/
- 07.06.2018 13:20:53, Craig Murray - EOS BlockSmith [@craigmurray100]: yes Dan confirmed normal operation was that, or higher
- 07.06.2018 13:21:10, Thomas "Disclaimer" Cox (all opinions mine) [@thomasbcox]: FYI, I'm still finishing off the article expressing the case for reverting Article IX. Team is reviewing now.
- 07.06.2018 13:21:23, Craig Murray - EOS BlockSmith [@craigmurray100]: all the eos-bios launches stage 1 to 20 ran that way as well 300-400 lib behind head
- 07.06.2018 13:24:03, Zhao Yu@ EOSLaoMao.com [@zhaoyu]: Just want to mention that the new chain will honor unregistered users right?
- 07.06.2018 13:24:31, Jae Chung - HKEOS [@jaechung]: The LIB updates in logs every time you go to the next producer, it looks something like this:
- ncy: 3989 ms]
- 1385989ms thread-0 producer_plugin.cpp:290 on_incoming_block ] Received block 910ac250b05781d0... #319206 @ 2018-06-07T17:23:02.000 signed by ccccccccccc1 [trxs: 0, lib: 318877, conf: 0, latency: 3989 ms]
- 1386492ms thread-0 producer_plugin.cpp:290 on_incoming_block ] Received block 9e4af3985fc8b100... #319207 @ 2018-06-07T17:23:02.500 signed by ccccccccccc1 [trxs: 0, lib: 318877, conf: 0, latency: 3992 ms]
- 1386988ms thread-0 producer_plugin.cpp:290 on_incoming_block ] Received block dacc179065da6452... #319208 @ 2018-06-07T17:23:03.000 signed by ccccccccccc1 [trxs: 0, lib: 318877, conf: 0, latency: 3988 ms]
- 1387488ms thread-0 producer_plugin.cpp:290 on_incoming_block ] Received block 4423d0ae362579e6... #319209 @ 2018-06-07T17:23:03.500 signed by ccccccccccc1 [trxs: 0, lib: 318877, conf: 0, latency: 3988 ms]
- 1387989ms thread-0 producer_plugin.cpp:290 on_incoming_block ] Received block 41237899f203295e... #319210 @ 2018-06-07T17:23:04.000 signed by ccccccccccc1 [trxs: 0, lib: 318877, conf: 0, latency: 3989 ms]
- 1388488ms thread-0 producer_plugin.cpp:290 on_incoming_block ] Received block 71e0227b5344b4e0... #319211 @ 2018-06-07T17:23:04.500 signed by ccccccccccc1 [trxs: 0, lib: 318877, conf: 0, late
- ncy: 3989 ms]
- 1388989ms thread-0 producer_plugin.cpp:290 on_incoming_block ] Received block 1a8978b3f7a8a272... #319212 @ 2018-06-07T17:23:05.000 signed by ccccccccccc1 [trxs: 0, lib: 318877, conf: 0, latency: 3989 ms]
- 1389490ms thread-0 producer_plugin.cpp:290 on_incoming_block ] Received block 859105995a15d10c... #319213 @ 2018-06-07T17:23:05.500 signed by ccccccccccc1 [trxs: 0, lib: 318877, conf: 0, latency: 3990 ms]
- 1389989ms thread-0 producer_plugin.cpp:290 on_incoming_block ] Received block 9cfd15544a461c51... #319214 @ 2018-06-07T17:23:06.000 signed by ddddddddddd1 [trxs: 0, lib: 318889, conf: 216, latency: 3989 ms]
- 1390488ms thread-0 producer_plugin.cpp:290 on_incoming_block ] Received block 455f1cb2bceed455... #319215 @ 2018-06-07T17:23:06.500 signed by ddddddddddd1 [trxs: 0, lib: 318889, conf: 0, latency: 3988 ms]
- 1390988ms thread-0 producer_plugin.cpp:290 on_incoming_block ] Received block 0dba201cd93c56b1... #319216 @ 2018-06-07T17:23:07.000 signed by ddddddddddd1 [trxs: 0, lib: 318889, conf: 0, latency: 3988 ms]
- 1391488ms thread-0 producer_plugin.cpp:290 on_incoming_block ] Received block ff111dc7f4185cb4... #319217 @ 2018-06-07T17:23:07.500 signed by ddddddddddd1 [trxs: 0, lib: 318889, conf: 0, latency: 3988 ms]
- 07.06.2018 13:24:44, Jae Chung - HKEOS [@jaechung]: I wouldn't be so fast to call it a problem in the boot process or a bug in the software @moonxor ;)
- 07.06.2018 13:26:01, Craig Murray - EOS BlockSmith [@craigmurray100]: yes lib moves at the end of the producers turn, but still follows behind getting it's 15/21 signatures to be lib
- 07.06.2018 13:26:32, mini - EOS Argentina [@moonxor]: I said I didn't know
- 07.06.2018 13:26:55, Marc-Antoine Ross - EOS Canada [@marcantoineross]: Yes
- 07.06.2018 13:28:05, Zhao Yu@ EOSLaoMao.com [@zhaoyu]: Got some holders who forget to register asking, so just want to make sure
- 07.06.2018 13:28:36, Craig Murray - EOS BlockSmith [@craigmurray100]: yes 100%
- 07.06.2018 13:47:36, Nate D - Aloha EOS [@n8dee]: What about unregistered holders that had erc20 EOS in an ETH wallet where that wallet never sent any ETH or EOS? Was the issue of those wallets not being in the snapshot resolved?
- 07.06.2018 14:11:02, Craig Murray - EOS BlockSmith [@craigmurray100]: i do not believe there is any problem with that. The addresses are recorded, and a mechanism can be built to retrive them given the user has the private key and can sign an eth tx to prove it. This is my understanding of the current state
- 07.06.2018 14:21:59, Thomas "Disclaimer" Cox (all opinions mine) [@thomasbcox]: [[Webpage]]It is the shared belief of the Arbitration Team and of me, Thomas, that the last-minute change to Article IX be reverted. This community is ready to govern itself. It should face that self governing future with confidence.
- https://forums.eosgo.io/discussion/1189/the-case-for-reverting-article-ix-9/p1?new=1
- 07.06.2018 14:22:30, Thomas "Disclaimer" Cox (all opinions mine) [@thomasbcox]: This is the write-up I promised on the 0100 call
- .....07.06.2018 14:46:51, Ben Sigman [@Bensig]: >>invited Kartik Mandaville | SpringRole [@kar2905]<<
- 07.06.2018 14:48:19, Ben Sigman [@Bensig]: Where is this list?
- 07.06.2018 14:48:36, Josh Kauffman - EOS Canada [@joshkauff]: If you check the notes from today's call, there should be links in there
- 07.06.2018 14:49:01, Josh Kauffman - EOS Canada [@joshkauff]: [[Webpage]]@Bensig https://docs.google.com/document/d/1nHEA3rfWLh4uwKmFnY9ypuRnWbOjuSy0LdKUlTapS2c
- 07.06.2018 14:50:12, Ben Sigman [@Bensig]: Thank you
- 07.06.2018 14:51:45, Ben Sigman [@Bensig]: Actually, I do not see any links in there for documents.
- 07.06.2018 14:52:36, Josh Kauffman - EOS Canada [@joshkauff]: Shoot that's my bad, I grabbed the one from yesterday's call. One sec, gonna get the right link.
- 07.06.2018 14:53:29, Josh Kauffman - EOS Canada [@joshkauff]: [[Webpage]]https://docs.google.com/spreadsheets/d/1t4p1tWs0fXf28nF2w4m5wwzO1cAknYs5XQGJLAIf1pU/edit#gid=753803902
- 07.06.2018 14:53:37, Josh Kauffman - EOS Canada [@joshkauff]: that's the one that has the ABP criteria
- 07.06.2018 14:55:25, Roshan - EOS Authority [@telepound]: {{FWD: Daniel Larimer [@daniellarimer], 07.06.2018 14:46:10}}
- Just a reminder we will be releasing 1.0.2 tomorrow with enhancements / bug fixes for the whitelist / blacklist feature. We will be doing weakly updates to fix rough edges, but nothing hard forking.
- 07.06.2018 15:01:46, Sharif Bouktila - EOS Dublin [@sharifbouktila]: For what it’s worth from everything we’ve seen so far with ECAF, we fully support the latest change and are confident ECAF will be able to handle any and all cases needed from launch.
- One thing they will need is a close relationship and open lines of communication with our BPs , there will be a lot of manual processes involved for quite a while.
- 07.06.2018 15:33:35, Timothy Lewis [@trustlayer]: Excited.
- 07.06.2018 15:39:05, Jae Chung - HKEOS [@jaechung]: [[Webpage]]https://github.com/EOSIO/eos/issues/3917
- 07.06.2018 15:39:14, Jae Chung - HKEOS [@jaechung]: [[Webpage]]https://github.com/EOSIO/eos/issues/3918
- 07.06.2018 15:39:36, Jae Chung - HKEOS [@jaechung]: found some things while testing in the last couple of hours which we've filed
- 07.06.2018 15:40:22, Jae Chung - HKEOS [@jaechung]: seems like they should be addressed soon, #3918 could be critical
- 07.06.2018 15:51:20, Denis | EOS Nation [@DenisCarriere]: 👍Thanks for sharing Jae, great finds, at least using the -j flag can workaround this issue (for now until resolved)
- 07.06.2018 15:56:53, Jae Chung - HKEOS [@jaechung]: yep, and I think Dan will solve #3917 with tomorrow's 1.0.2 release
- 07.06.2018 15:57:30, Kyle - EOSVibes [@TheAwakenment]: Boom!💥 Good on yea Tom cat! ✅
- I’m very happy to see this (apologies for the over-enthusiastic response) 😊
- 07.06.2018 15:57:47, Jae Chung - HKEOS [@jaechung]: but #3918 needs attention, it could be quite dangerous
- 07.06.2018 16:09:53, Denis | EOS Nation [@DenisCarriere]: 🤔I agree with Dan's comment, not too sure if that's a bug in cleos or network forks/local node environment.
- Dan tagged it as "support" (might not be a bug)
- 07.06.2018 16:17:36, Alexandre Bourget - EOS Canada [@abourget]: Yeah I remember experiencing that when some nodes were not fully sync'd ...
- 07.06.2018 16:17:44, Alexandre Bourget - EOS Canada [@abourget]: it's quite surprising, though !
- 07.06.2018 16:18:39, Alexandre Bourget - EOS Canada [@abourget]: We also found that, depending on network topology, setting p2p-max-nodes-per-host helps..
- 07.06.2018 16:19:06, Alexandre Bourget - EOS Canada [@abourget]: we were often trying to connect many times to a node from the same office IP.. so some wouldn't be sync'd..
- 07.06.2018 16:19:10, Alexandre Bourget - EOS Canada [@abourget]: but it wasn't immediately obvious
- .....07.06.2018 16:37:13, Jason Wang - EOS Pacific [@jason_wang_4595]: >>unsupported service message type: messageActionChatJoinedByLink<<
- .....07.06.2018 16:47:06, Dr Ye EOS WashingtonDC - TChain [@doc309]: >>unsupported service message type: messageActionChatJoinedByLink<<
- 07.06.2018 16:47:59, Dr Ye EOS WashingtonDC - TChain [@doc309]: kicking .. just prevent from being kicked out again :-)
- 07.06.2018 16:53:33, Dr Ye EOS WashingtonDC - TChain [@doc309]: Is 597a still the only mainnet candidate as we voted two days ago? sorry missed the latest status
- 07.06.2018 17:02:47, Marc-Antoine Ross - EOS Canada [@marcantoineross]: No, there will be a new chain launched by the anonymous ABP
- 07.06.2018 17:03:47, Dr Ye EOS WashingtonDC - TChain [@doc309]: ok. just let us know the new genesis and peer list
- 07.06.2018 17:04:22, Jason Wang - EOS Pacific [@jason_wang_4595]: do you knwo if we have anonymous ABP selected already?
- 07.06.2018 17:32:03, Timothy Lewis [@trustlayer]: Not yet
- 07.06.2018 17:46:51, David P - EOS42 [@GunnisonCap]: Can ask Kevin Rose shortly he’ll be arriving at Tulip in about half an hour
- 07.06.2018 17:47:16, David P - EOS42 [@GunnisonCap]: But yes not yet we’ll get an update on the call in 3hrs
- 07.06.2018 20:48:53, Douglas Horn | EOSAmericas [@DAPPerD]: For the upcoming call, we are bringing in our Korean-speaking team member to translate summaries in Korean during the call (in the Chat section). He can also translate any questions back into English. We want to test this out and if it works well, we will add more languages in the next few calls. Hopefully Chinese first. We think it will help the call go more efficiently. This morning there were some times when the discussion went on for a while without translation. Also, the current system puts a lot of responsibility on EOS Asia.
- 07.06.2018 20:50:03, Douglas Horn | EOSAmericas [@DAPPerD]: Please let us know if you think it's a model worth exploring and we will try to get more of our multilingual team deployed on future calls.
- 07.06.2018 20:54:35, Dr Ye EOS WashingtonDC - TChain [@doc309]: zoom link for the call?
- 07.06.2018 20:54:49, Marc-Antoine Ross - EOS Canada [@marcantoineross]: [[Webpage]]All,
- For internal use only for the EOS Validation (EMLG) Telegram Channel. Please do not share outside of this channel.
- Here's the proposed agenda:
- Re-ratify on next call the 2/3+1 for Go/NoGo
- Constitution change vote + update on changes
- ABP selection update
- No P0 P1 reported?
- Voting Portal update
- Go/NoGO vote
- 48 hour update + coordination of the announcements
- https://zoom.us/j/326472166
- .....07.06.2018 20:54:54, Marc-Antoine Ross - EOS Canada [@marcantoineross]: >>unsupported service message type: messageActionPinMessage<<
- 07.06.2018 20:57:01, deck - EOS New York [@deckw]: Hey Doug, we usually use the chat window for votes so it is hard to get messages to stay on screen but I say give it a go
- 07.06.2018 20:58:50, Rick Schlesinger - EOS New York [@unlockrs]: please dont forward the zoom link
- 07.06.2018 21:00:21, Douglas Horn | EOSAmericas [@DAPPerD]: Yes, I've been on the last several calls. We've been working up ideas for helping out the process where possible. I've asked him to just do summaries. I'm thinking Korean and Chinese might be enough.
- 07.06.2018 21:01:05, Douglas Horn | EOSAmericas [@DAPPerD]: We're not forwarding it. It's to our team member.
- 07.06.2018 21:01:19, Rohan - EOSauthority [@RohanWillAnswer]: Ok
- 07.06.2018 21:01:42, Rick Schlesinger - EOS New York [@unlockrs]: im saying dont forward it to public channels. you can provide to your team members of course.
- 07.06.2018 21:02:00, Rohan - EOSauthority [@RohanWillAnswer]: Is the link out yet?
- 07.06.2018 21:02:06, Douglas Horn | EOSAmericas [@DAPPerD]: 👍
- 07.06.2018 21:02:09, Rohan - EOSauthority [@RohanWillAnswer]: Can’t find it anywhere?
- 07.06.2018 21:02:16, Rick Schlesinger - EOS New York [@unlockrs]: scroll up about 7 lines
- 07.06.2018 21:21:09, Rohan - EOSauthority [@RohanWillAnswer]: I think it’s my 2AM brain. Thanks.
- 07.06.2018 22:00:03, Alexandre Bourget - EOS Canada [@abourget]: @daniellarimer hey Dan, what would be the impact of transfering all EOS out the eosio.ram and eosio.ramfee accounts, before eosio resignation ?
- 07.06.2018 22:00:38, Daniel Larimer [@daniellarimer]: Ram is used by market maker
- 07.06.2018 22:01:08, Daniel Larimer [@daniellarimer]: Ramfee is at community disposal
- 07.06.2018 22:01:13, Roman CryptoLions [@romanskaskiw]: @jemxpat
- 07.06.2018 22:01:59, jem [@jemxpat]: No user "gifted" with ram would be able to sell until some other user buys new ram post 15% activation. Tested.
- 07.06.2018 22:05:46, Daniel Larimer [@daniellarimer]: It violates bancor algorithm invariant
- 07.06.2018 22:06:01, Kyunghwan Kim - EOSeoul.io [@resvera]: transfer of token doesn’t impact price of ram. price of ram is determined in bancor state in rammarket db only through buy/sellram for now.
- and transfer of token doesn’t impact the supply of token, also.
- 07.06.2018 22:06:30, jem [@jemxpat]: KHKim is correct. There is no violation of Bancor with this method.
- 07.06.2018 22:06:52, Daniel Larimer [@daniellarimer]: It prevents market maker from filling orders
- 07.06.2018 22:07:11, jem [@jemxpat]: Only until new user buys ram.
- 07.06.2018 22:09:03, David Moss [@DavidMoss]: [[Webpage]]Dan, we need a comment on https://github.com/EOSIO/eos/issues/3918
- .....07.06.2018 22:11:25, Marshall - SaltBlock [@marshallslc]: >>invited Robrigo - EOS Detroit [@robrigo]<<
- 07.06.2018 22:17:12, Robrigo - EOS Detroit [@robrigo]: Thanks for the add @marshallslc.
- 07.06.2018 22:17:27, Marshall - SaltBlock [@marshallslc]: You're welcome
- 07.06.2018 22:26:28, Andrey Marchuk [@madz7z]: Tulip conf cannot join back because of 200 participants limits
- 07.06.2018 22:29:15, Thomas "Disclaimer" Cox (all opinions mine) [@thomasbcox]: TRY AGAIN in a few minutes
- 07.06.2018 22:46:14, Ben Sigman [@Bensig]: [[Webpage]]https://docs.google.com/spreadsheets/d/1gJnn0KS4rgDnTowPlN80gulVI77lRgldPhH17kZOnnI/edit#gid=744644035
- 07.06.2018 22:57:26, Alexandre Bourget - EOS Canada [@abourget]: @daniellarimer do you think it makes sense to create an eosio.burned account, where we can put things in ?
- 07.06.2018 22:57:35, Alexandre Bourget - EOS Canada [@abourget]: If you're planning a recall action, maybe it's useless..
- 07.06.2018 23:04:21, Chitty 🅑 - EOS Argentina [@JChitty]: Is phill from eos switzerland here?.. I am just wondering if they are a bp since I havent seen them before
- 07.06.2018 23:07:09, Jose Toriello [@josetoriello]: Good question
- 07.06.2018 23:12:18, Heejae - EOSYS.io [@heejaejoo]: Me too.
- 07.06.2018 23:13:44, Chitty 🅑 - EOS Argentina [@JChitty]: I think there are a couple of fake bps in the chat
- 07.06.2018 23:15:28, Igor Lins e Silva - EOS Rio [@igorls]: 👍
- 07.06.2018 23:15:47, Rick Schlesinger - EOS New York [@unlockrs]: do the vote here?
- 07.06.2018 23:16:10, J.T. - EOS the World [@HartParr]: Yes.
- 07.06.2018 23:16:21, J.T. - EOS the World [@HartParr]: Or post a new straw poll here.
- .....07.06.2018 23:17:20, An Byung Chan - EOSYS [@Anbyungchan]: >>invited Lievin Kim - EOSYS [@lievin_kim]<<
- 07.06.2018 23:18:11, David P - EOS42 [@GunnisonCap]: I think Phil from EOS Switzerland is a fake. Never once heard of them on any channel.
- .....07.06.2018 23:18:15, Ricardo Marlière - EOS Brazil [@rbmarliere]: >>invited EOS - Antarctica [@jsw513]<<
- 07.06.2018 23:18:32, Ross (EOSphere) [@rossco99]: Was thinking the same
- 07.06.2018 23:18:34, mini - EOS Argentina [@moonxor]: Me neither
- 07.06.2018 23:18:43, Chitty 🅑 - EOS Argentina [@JChitty]: he is fake, and looks like a troll
- 07.06.2018 23:18:44, J.T. - EOS the World [@HartParr]: agreed.
- 07.06.2018 23:18:50, Pascal Van Wy | EOS Network by Advanced Blockchain AG [@vanwy]: agreed
- 07.06.2018 23:18:53, Patrick B - Aloha EOS: He's definitely trolling
- 07.06.2018 23:19:19, Rohan - EOSauthority [@RohanWillAnswer]: Good find..
- 07.06.2018 23:19:20, Rick Schlesinger - EOS New York [@unlockrs]: +1
- 07.06.2018 23:19:38, Alex - EOS SOCAL [@gmory]: +1
- 07.06.2018 23:20:09, Hubert Chang-EOSeoul.io [@hjhc00]: Can we kick him out from the call?
- 07.06.2018 23:20:11, Jacky - JEDA [@newbdez33]: +1
- 07.06.2018 23:20:11, Jacky - JEDA [@newbdez33]: +1
- 07.06.2018 23:20:11, Jacky - JEDA [@newbdez33]: +1
- 07.06.2018 23:20:14, David P - EOS42 [@GunnisonCap]: Yes please
- 07.06.2018 23:20:28, David P - EOS42 [@GunnisonCap]: And various others who are clear no’s
- 07.06.2018 23:20:37, Khosi Morafo eosAfrica [@khosimorafo]: +1
- 07.06.2018 23:20:37, Denis | EOS Nation [@DenisCarriere]: +1 for ELMG voting
- 07.06.2018 23:20:47, Andrey Marchuk [@madz7z]: +1
- 07.06.2018 23:20:57, Kevin Rose - EOS New York [@eosnewyork]: +1
- 07.06.2018 23:21:01, Craig Murray - EOS BlockSmith [@craigmurray100]: +1
- .....07.06.2018 23:21:03, Ben Sigman [@Bensig]: >>invited Kurt Braget | White Rabbit ICOs [@kurtybot]<<
- 07.06.2018 23:21:05, Bean@EOS Cannon [@Luckybean]: +1
- 07.06.2018 23:21:14, Pascal Van Wy | EOS Network by Advanced Blockchain AG [@vanwy]: +1
- .....07.06.2018 23:21:23, jem [@jemxpat]: >>invited Nicholas Mullesch EOS.Fish [@BitcoinRootUser]<<
- 07.06.2018 23:21:39, Cryptosufi - 威杰 - BitSpace [@KryptoSufi]: +1
- 07.06.2018 23:21:41, Eric - sw/eden [@xebb82]: Can we postpone the vote until 13:00 and have some tool ready?
- 07.06.2018 23:21:43, Nicholas Mullesch EOS.Fish [@BitcoinRootUser]: Thank you Jem!
- 07.06.2018 23:22:09, Denis | EOS Nation [@DenisCarriere]: 👎
- 07.06.2018 23:22:09, Issay@OracleChain [@issayTseng]: +1
- 07.06.2018 23:23:17, Douglas Horn | EOSAmericas [@DAPPerD]: 👎
- 07.06.2018 23:23:39, Nicholas Mullesch EOS.Fish [@BitcoinRootUser]: +1
- 07.06.2018 23:23:40, Chitty 🅑 - EOS Argentina [@JChitty]: we have most bps online now, lets vote
- 07.06.2018 23:23:47, Cryptosufi - 威杰 - BitSpace [@KryptoSufi]: We could start cleos voting, until audited tools arrive.
- 07.06.2018 23:23:57, Eric - sw/eden [@xebb82]: I'm up for whatever, this just feels like a farce...
- 07.06.2018 23:24:03, Rohan - EOSauthority [@RohanWillAnswer]: Please stop posting messages.. wait for Sharif’s question and then vote..
- 07.06.2018 23:24:21, Josh Kauffman - EOS Canada [@joshkauff]: -----------------------------------------------------------------------------------------------------------------------------
- PLEASE DO NOT WRITE ANYTHING BELOW THIS UNTIL THOMAS HAS POSTED THE QUESTION IN THIS CHANNEL
- -----------------------------------------------------------------------------------------------------------------------------
- 07.06.2018 23:24:37, Thomas "Disclaimer" Cox (all opinions mine) [@thomasbcox]: [Sharif will post the question]
- 07.06.2018 23:24:59, Thomas "Disclaimer" Cox (all opinions mine) [@thomasbcox]: You need permission
- This form can only be viewed by users in the owner's organization.
- 07.06.2018 23:25:13, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Please confirm using GO/ NO GO to launch at 13:00 UTC 8th June 2018 using current release.
- 07.06.2018 23:25:16, Denis | EOS Nation [@DenisCarriere]: No need for form
- 07.06.2018 23:25:20, Cryptosufi - 威杰 - BitSpace [@KryptoSufi]: Go
- 07.06.2018 23:25:26, Sebastian Castro - EOS Argentina 🇦🇷 [@sebicas]: Go
- 07.06.2018 23:25:30, Ralf Weinand | GenerEOS (BP Sydney) [@RalfWei]: Go
- 07.06.2018 23:25:30, James Sutherland - Cypherglass.com [@JamesSutherland]: Go
- 07.06.2018 23:25:31, Rohan - EOSauthority [@RohanWillAnswer]: Go - EOS Authority
- 07.06.2018 23:25:31, Alex - EOS SOCAL [@gmory]: Go
- 07.06.2018 23:25:32, Nate D - Aloha EOS [@n8dee]: GO
- 07.06.2018 23:25:33, Marek EOSEmerge.io - Poland [@mbaclawski]: GO
- 07.06.2018 23:25:34, Domi - EOS Rio [@domiscd]: the abp wasn't even selected, poor him if it's a go
- 07.06.2018 23:25:34, EOS - Antarctica [@jsw513]: No go.
- 07.06.2018 23:25:34, zhangzc | Onechain [@zhangzc]: GO
- 07.06.2018 23:25:34, Jose Toriello [@josetoriello]: Go
- 07.06.2018 23:25:38, Craig Murray - EOS BlockSmith [@craigmurray100]: go
- 07.06.2018 23:25:40, Andrey Marchuk [@madz7z]: Go - BlockEOS.io
- 07.06.2018 23:25:40, Jorge Ferrando - EOSMetal [@jorfermo]: Go
- 07.06.2018 23:25:40, Denis | EOS Nation [@DenisCarriere]: GO
- 07.06.2018 23:25:40, Ross (EOSphere) [@rossco99]: Go
- 07.06.2018 23:25:40, [email protected] [@crazybits]: Go Go Go
- 07.06.2018 23:25:40, Marshall - SaltBlock [@marshallslc]: Go
- 07.06.2018 23:25:42, deck - EOS New York [@deckw]: go
- 07.06.2018 23:25:43, Kedar Iyer | LibertyBlock [@k26dr]: GO
- 07.06.2018 23:25:43, Nicholas Mullesch EOS.Fish [@BitcoinRootUser]: Go
- 07.06.2018 23:25:43, Joshua Qiu @EOSeco [@Joshuaqiu]: Go
- 07.06.2018 23:25:44, Robert DeWilder - Worbli [@BrownDonahue]: go
- 07.06.2018 23:25:45, Miłosz Eos Emerge - Poland [@Om4tic]: Go
- 07.06.2018 23:25:49, Sharif Bouktila - EOS Dublin [@sharifbouktila]: No Go
- 07.06.2018 23:25:54, Eric - sw/eden [@xebb82]: No Go
- 07.06.2018 23:25:56, Jason Wang - EOS Pacific [@jason_wang_4595]: Go
- 07.06.2018 23:25:57, Jialin Li - EOS WTZ [@Jialinli]: Go
- 07.06.2018 23:25:58, jem [@jemxpat]: jem proxy BP2 EOSPH hkeos eoskh Shenzhen Acro Tribes eosviet
- No Go
- 07.06.2018 23:26:02, Hugo Yang - EOS Silicon Valley [@hugo_eossv]: No go
- 07.06.2018 23:26:03, Timothy Lewis [@trustlayer]: No go
- 07.06.2018 23:26:04, Issay@OracleChain [@issayTseng]: NO GO
- 07.06.2018 23:26:07, Igor Lins e Silva - EOS Rio [@igorls]: No go
- 07.06.2018 23:26:09, Basit Raza - EOSVibes [@boltxyz]: GO
- 07.06.2018 23:26:12, Marc-Antoine Ross - EOS Canada [@marcantoineross]: Go
- 07.06.2018 23:26:12, ndoy |EOS Cafe: Calgary | [@n_doy]: GO
- 07.06.2018 23:26:13, Ian(Yahuang Wu) - MEET.ONE [@wuyahuang]: NO GO.
- 07.06.2018 23:26:13, Pete BlockMatrix [@blockmatrixpete]: no go
- 07.06.2018 23:26:14, Zhao Yu@ EOSLaoMao.com [@zhaoyu]: GO
- 07.06.2018 23:26:17, Ricardo Marlière - EOS Brazil [@rbmarliere]: no go... why the rush? 1.0.2 is not even out
- 07.06.2018 23:26:18, Jacky - JEDA [@newbdez33]: GO
- 07.06.2018 23:26:18, Celu - Blockgenic [@celur]: Go
- 07.06.2018 23:26:20, zhangzc | Onechain [@zhangzc]: GO
- 07.06.2018 23:26:22, J.T. - EOS the World [@HartParr]: No Go.
- 07.06.2018 23:26:22, Ravi (Blockpro.one) [@ravincal]: No Go
- 07.06.2018 23:26:27, Gordo | HelloEOS [@gordoor]: go
- 07.06.2018 23:26:29, [email protected] [@qianxiaofeng]: no go
- 07.06.2018 23:26:35, Robin [email protected] [@robinwen]: Go
- 07.06.2018 23:26:37, Dafeng Guo - EOS Asia [@dftle]: Go
- 07.06.2018 23:26:43, WinLin-EOSBIXIN [@hiwinlin]: GO
- 07.06.2018 23:26:44, Ben Sigman [@Bensig]: No Go
- 07.06.2018 23:26:46, David P - EOS42 [@GunnisonCap]: No go
- 07.06.2018 23:26:49, Emily Bush - shEOS [@emmybush]: No go
- 07.06.2018 23:26:49, Pascal Van Wy | EOS Network by Advanced Blockchain AG [@vanwy]: No Go
- 07.06.2018 23:27:00, Heejae - EOSYS.io [@heejaejoo]: GO
- 07.06.2018 23:27:09, Guillaume @eostitan.com [@gyom_eos_titan]: No go
- 07.06.2018 23:27:12, heipacker eosunion [@heipacker]: go
- 07.06.2018 23:27:16, Hubert Chang-EOSeoul.io [@hjhc00]: No go
- 07.06.2018 23:27:22, Bohdan CryptoLions 🦁 [@BohdanCryptoLions]: no go
- 07.06.2018 23:27:22, Luis Fernando G : EOSMESOdotio 🇬🇹 [@Maski_LF]: No go
- 07.06.2018 23:27:28, Yvonne (EOShenzhen) [@yvonnekama]: no go
- 07.06.2018 23:27:45, Zhao Yu@ EOSLaoMao.com [@zhaoyu]: duplicate
- 07.06.2018 23:27:58, Robrigo - EOS Detroit [@robrigo]: no go
- 07.06.2018 23:28:04, Douglas Horn | EOSAmericas [@DAPPerD]: No go
- 07.06.2018 23:29:20, phajek eosarmy Patrick [@Phajek]: no go
- .....07.06.2018 23:30:13, Robrigo - EOS Detroit [@robrigo]: >>invited jesta — Greymass [@jestagram]<<
- 07.06.2018 23:30:44, jesta — Greymass [@jestagram]: go
- 07.06.2018 23:31:26, Thomas "Disclaimer" Cox (all opinions mine) [@thomasbcox]: END OF VOTING END OF VOTING END OF VOTING
- 07.06.2018 23:31:32, Jose Toriello [@josetoriello]: (Eosmeso 1 vote is no go. Sorry for confusion)
- 07.06.2018 23:32:11, Mathieu Boulianne - EOS Canada [@mathieubo]: On another note, on the Print / Work Out vote, I got (after cleaning the data):
- -55 Print
- -31 work out
- 07.06.2018 23:33:20, jem [@jemxpat]: Appoint ABP, let hiim get ready.
- 07.06.2018 23:33:25, Denis | EOS Nation [@DenisCarriere]: I believe the "no go" is about the RAM disagreement
- 07.06.2018 23:33:25, Timothy Lewis [@trustlayer]: There is a lot of deleting going on in this channel
- 07.06.2018 23:33:36, Jose Toriello [@josetoriello]: Reason was no clarity on print/ram issue
- 07.06.2018 23:33:53, Craig Murray - EOS BlockSmith [@craigmurray100]: dan was very clear we should not take the RAM balance
- 07.06.2018 23:34:01, Craig Murray - EOS BlockSmith [@craigmurray100]: Dan said there is only 1 good option, print
- 07.06.2018 23:34:02, Sharif Bouktila - EOS Dublin [@sharifbouktila]: I've deleted some comments in the voting window
- 07.06.2018 23:34:11, Sebastian Castro - EOS Argentina 🇦🇷 [@sebicas]: +1
- 07.06.2018 23:34:16, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Only Me , Marc and Eric are admins
- 07.06.2018 23:34:17, Robrigo - EOS Detroit [@robrigo]: My specific reason is waiting to launch with v1.0.2 tag. Would vote GO after tomorrow.
- 07.06.2018 23:34:25, Timothy Lewis [@trustlayer]: Same
- 07.06.2018 23:34:26, Jose Toriello [@josetoriello]: I know. But the vote was not counted
- 07.06.2018 23:34:32, Craig Murray - EOS BlockSmith [@craigmurray100]: why not wait to 1.0.3? or 1.1? or 2.0?
- 07.06.2018 23:34:42, Craig Murray - EOS BlockSmith [@craigmurray100]: no major issues fixed in 1.0.2
- 07.06.2018 23:34:52, Robrigo - EOS Detroit [@robrigo]: Because 1.0.3 is 8 days away, not 1 day away
- 07.06.2018 23:35:03, jesta — Greymass [@jestagram]: You wouldn't even have to upgrade to 1.0.2 immediately.
- 07.06.2018 23:35:13, Robrigo - EOS Detroit [@robrigo]: You asked, I answered :)
- 07.06.2018 23:35:17, Timothy Lewis [@trustlayer]: We also will likely get an opinion from Dan on the lost transaction issue.
- 07.06.2018 23:35:41, Sherman EOSREAL [@ShermanXie]: go
- 07.06.2018 23:35:42, Timothy Lewis [@trustlayer]: Get ABP selected
- 07.06.2018 23:35:46, Issay@OracleChain [@issayTseng]: we thought the 3918 issue might be P1, so wait for 1.0.2
- 07.06.2018 23:35:48, Timothy Lewis [@trustlayer]: Have them prepare
- 07.06.2018 23:36:18, EOS - Antarctica [@jsw513]: The reasoning for someone’s vote is irrelevant. We took a vote. Use the numbers.
- 07.06.2018 23:36:23, Thomas "Disclaimer" Cox (all opinions mine) [@thomasbcox]: I'm counting 39 GO to 25 NO GO in the telegram channel
- 07.06.2018 23:37:00, Thomas "Disclaimer" Cox (all opinions mine) [@thomasbcox]: I can't promise that the data is completely clean
- 07.06.2018 23:37:13, Ravi (Blockpro.one) [@ravincal]: Many BP missing here because not everyone here
- 07.06.2018 23:37:15, EOS - Antarctica [@jsw513]: One does not get to the chance to try and sway opinion after the election.
- .....07.06.2018 23:37:24, Ben Sigman [@Bensig]: >>invited Tal Muskal | LiquidEOS [@tmuskal]<<
- 07.06.2018 23:37:36, Denis | EOS Nation [@DenisCarriere]: 👎 1.0.2 is not a show stopper
- 07.06.2018 23:37:43, Jose Toriello [@josetoriello]: +1
- 07.06.2018 23:37:46, Issay@OracleChain [@issayTseng]: but another reason is not all BPs is present here
- 07.06.2018 23:37:53, Alexandre Bourget - EOS Canada [@abourget]: https://abourget.keybase.pub/votetallysum.txt
- 07.06.2018 23:37:55, Jacky - JEDA [@newbdez33]: is 1.0.2 fixes any P0 or P1 issue?
- 07.06.2018 23:37:57, Alexandre Bourget - EOS Canada [@abourget]: That's my data ^^
- 07.06.2018 23:38:23, Nate D - Aloha EOS [@n8dee]: No
- 07.06.2018 23:38:27, Domi - EOS Rio [@domiscd]: I'm just concerned with the APB, it wasn't even selected yet, we don't know what his timezone will be, let him prepare
- 07.06.2018 23:38:27, Craig Murray - EOS BlockSmith [@craigmurray100]: no
- 07.06.2018 23:38:30, James Sutherland - Cypherglass.com [@JamesSutherland]: 👍
- 07.06.2018 23:39:42, Thomas "Disclaimer" Cox (all opinions mine) [@thomasbcox]: The first vote was 52 to 21 // a 71% YES
- 07.06.2018 23:39:45, heipacker eosunion [@heipacker]: go
- 07.06.2018 23:40:02, Jacky - JEDA [@newbdez33]: then there is no reason we have to wait for 1.0.2 for my understanding.
- 07.06.2018 23:40:18, EOS - Antarctica [@jsw513]: Why are we taking a vote, if we're not going to put the results into place?
- 07.06.2018 23:40:25, Alex - EOS SOCAL [@gmory]: Agree with this. Should've been sorted earlier than this call. If we wind up with No Go, the APB needs to announce endpoints asap fo syncing.
- 07.06.2018 23:43:42, Alex - EOS SOCAL [@gmory]: But also regarding that, the schedule was announced long ago. The abp knew this 12 hour slot was a potential go time.
- 07.06.2018 23:43:47, Alex - EOS SOCAL [@gmory]: Regardless of timezone
- 07.06.2018 23:44:29, Issay@OracleChain [@issayTseng]: [[Photo]]
- .....07.06.2018 23:45:05, Thomas "Disclaimer" Cox (all opinions mine) [@thomasbcox]: >>invited Li Peng | ChainPool [@EOSMalaysia]<<
- 07.06.2018 23:45:48, Li Peng | ChainPool [@EOSMalaysia](you): Thanks
- 07.06.2018 23:46:17, ndoy |EOS Cafe: Calgary | [@n_doy]: this might be fake:
- From deadlock - EOSIO.SG to Everyone: 09:44 PM
- why celeb has EOSIO.SG website under the name
- 07.06.2018 23:46:41, ndoy |EOS Cafe: Calgary | [@n_doy]: deadlock doesn't know him
- 07.06.2018 23:46:49, Li Peng | ChainPool [@EOSMalaysia](you): Sorry, there's been some confusion
- 07.06.2018 23:46:51, Li Peng | ChainPool [@EOSMalaysia](you): One second
- 07.06.2018 23:48:37, Denis | EOS Nation [@DenisCarriere]: No go's should provide reasons (not blank no go's)
- 07.06.2018 23:48:54, EOS - Antarctica [@jsw513]: Why? A vote is a vote
- 07.06.2018 23:49:07, Lee - EOSTEA [@strahe]: go
- 07.06.2018 23:49:55, Jose Toriello [@josetoriello]: Some no gos were due to uncertainty on the vote count about RAM. If that is cleared and whichever option has been taken, so the ABP doesnt have to guess.
- Then we are changing from no go to GO
- .....07.06.2018 23:50:11, An Byung Chan - EOSYS [@Anbyungchan]: >>invited Ashley Gan<<
- 07.06.2018 23:50:54, Denis | EOS Nation [@DenisCarriere]: Can't update to 1.0.2, code is not released (maybe released)
- 07.06.2018 23:50:59, Marek EOSEmerge.io - Poland [@mbaclawski]: too bad that it didn't happen today
- 07.06.2018 23:51:17, Marek EOSEmerge.io - Poland [@mbaclawski]: Its a real pitty that they can't make a good reason for no go
- 07.06.2018 23:51:50, zhangzc | Onechain [@zhangzc]: +1
- 07.06.2018 23:53:20, Denis | EOS Nation [@DenisCarriere]: How many hours for validation + meshes (12 hours?)
- 07.06.2018 23:53:44, jem [@jemxpat]: Yes, when you don't like the vote, be sure to challenge the motivation and reason of the opposition. What a great strategy!
- 07.06.2018 23:54:39, Daniel Larimer [@daniellarimer]: conclusion?
- 07.06.2018 23:55:02, jem [@jemxpat]: No conclusion yet, still herding cats.
- 07.06.2018 23:55:14, Craig Murray - EOS BlockSmith [@craigmurray100]: too many people are no-go, so it's no go..
- 07.06.2018 23:55:20, Daniel Larimer [@daniellarimer]: would these be crypto-kitties
- 07.06.2018 23:55:25, Rick Schlesinger - EOS New York [@unlockrs]: its a no-go
- 07.06.2018 23:55:30, jem [@jemxpat]: Exactly!, Herding crypto kitties!
- 07.06.2018 23:55:35, Roman CryptoLions [@romanskaskiw]: It'll happen next vote.
- 07.06.2018 23:55:46, David P - EOS42 [@GunnisonCap]: Yes, I think so next vote
- .....07.06.2018 23:55:47, Pascal Van Wy | EOS Network by Advanced Blockchain AG [@vanwy]: >>invited Pascal | EOS Network by Advanced Blockchain AG [@maplefrog]<<
- 07.06.2018 23:55:59, Domi - EOS Rio [@domiscd]: yes, next will be go for sure ☺
- 07.06.2018 23:56:05, Rick Schlesinger - EOS New York [@unlockrs]: +1
- 07.06.2018 23:56:11, jem [@jemxpat]: I think so.
- 07.06.2018 23:56:22, Roman CryptoLions [@romanskaskiw]: See you all in 9 hrs.
- 07.06.2018 23:56:27, Roman CryptoLions [@romanskaskiw]: 7 am here. 😊
- 07.06.2018 23:56:27, Pete BlockMatrix [@blockmatrixpete]: 👍
- 07.06.2018 23:56:32, Lars Ihler - BitSpace [@LarsIhler]: 06 am
- 07.06.2018 23:56:47, Igor Lins e Silva - EOS Rio [@igorls]: why not use the top_eos_bps_test team on keybase?, the one that Thomas set up for disaster recovery
- 07.06.2018 23:57:21, Igor Lins e Silva - EOS Rio [@igorls]: it far more verifiable than telegram
- 07.06.2018 23:57:52, jem [@jemxpat]: 👍
- 07.06.2018 23:58:01, Kyunghwan Kim - EOSeoul.io [@resvera]: +1
- 07.06.2018 23:58:09, Craig Murray - EOS BlockSmith [@craigmurray100]: that would take a lot of time to onboard all those people, especially asian groups
- 07.06.2018 23:58:24, Igor Lins e Silva - EOS Rio [@igorls]: most of the bps already have keybase setup
- .....07.06.2018 23:58:31, Dafeng Guo - EOS Asia [@dftle]: >>invited Edward Lew - EOS Land [@ecslew]<<
- 07.06.2018 23:58:34, Craig Murray - EOS BlockSmith [@craigmurray100]: "most"? based on what metric?
- .....07.06.2018 23:58:44, Dafeng Guo - EOS Asia [@dftle]: >>invited Goh MEET.ONE<<
- 07.06.2018 23:59:00, Ricardo Marlière - EOS Brazil [@rbmarliere]: keybase +1
- 07.06.2018 23:59:13, Igor Lins e Silva - EOS Rio [@igorls]: based on 207 members we validated in the ghostbusters team
- 07.06.2018 23:59:52, Igor Lins e Silva - EOS Rio [@igorls]: we can select one vote per BP official website with dns proof on keybase
- 08.06.2018 00:00:17, Craig Murray - EOS BlockSmith [@craigmurray100]: if that's what everyone wants and you can get consensus
- 08.06.2018 00:00:26, Craig Murray - EOS BlockSmith [@craigmurray100]: thomas is right now telling us how he thinks we are going to do it next meeting
- 08.06.2018 00:00:36, ndoy |EOS Cafe: Calgary | [@n_doy]: yes for keybase, no need to reinvent the wheel. We have validated BPs there, just to make sure no duplicate e.g. between me and Syed
- 08.06.2018 00:00:41, Igor Lins e Silva - EOS Rio [@igorls]: it was just a suggestion ;)
- 08.06.2018 00:00:51, Craig Murray - EOS BlockSmith [@craigmurray100]: 👍
- 08.06.2018 00:01:09, Douglas Horn | EOSAmericas [@DAPPerD]: Keybase is a good suggestion.
- 08.06.2018 00:01:58, [email protected] [@crazybits]: [[Webpage]]https://eos.live/detail/2500 looks like this is EOS crypto-kitties called Pandafun 😁
- 08.06.2018 00:02:02, Igor Lins e Silva - EOS Rio [@igorls]: and is super easy for people to cast the vote on their public folders
- 08.06.2018 00:02:15, ndoy |EOS Cafe: Calgary | [@n_doy]: David says we'll use zoom poll
- 08.06.2018 00:02:39, Igor Lins e Silva - EOS Rio [@igorls]: but can that prevent multiple votes per bp?
- 08.06.2018 00:03:01, ndoy |EOS Cafe: Calgary | [@n_doy]: I never use zoom poll, so I'm not sure how it gets counted
- 08.06.2018 00:03:21, ndoy |EOS Cafe: Calgary | [@n_doy]: unless with invite meeting maybe, using our email
- 08.06.2018 00:03:50, Thomas "Disclaimer" Cox (all opinions mine) [@thomasbcox]: I want to see only one person attending PER BP
- 08.06.2018 00:03:51, Douglas Horn | EOSAmericas [@DAPPerD]: Short of any unknown issue, EOSAmericas will be changing our vote to “Go” in the next vote. We will examine issue #3918 in the interim.
- .....08.06.2018 00:04:52, Sherman EOSREAL [@ShermanXie]: >>invited Felicia EOSREAL<<
- 08.06.2018 00:05:43, Domi - EOS Rio [@domiscd]: @thomasbcox we can use keybase, just a suggestion :)
- 08.06.2018 00:06:20, ndoy |EOS Cafe: Calgary | [@n_doy]: lots of people in the zoom chat I never talk to them in any testnets channel. I'm not saying that they are not BP candidates but we do need to verify who casts vote
- 08.06.2018 00:08:12, Ben Sigman [@Bensig]: [[Webpage]]https://status.producer.vote/
- 08.06.2018 00:09:20, Marek EOSEmerge.io - Poland [@mbaclawski]: Then keybase is a good option
- 08.06.2018 00:10:15, Denis | EOS Nation [@DenisCarriere]: If "print" is decided, we still have "no go" votes for that reason (we need to include arguments for "no" votes to filter out some votes)
- 08.06.2018 00:10:47, Edward Lew - EOS Land [@ecslew]: Awesome! Thanks all of your hard working
- 08.06.2018 00:11:36, James Sutherland - Cypherglass.com [@JamesSutherland]: @thomasbcox @marcantoineross @sharifbouktila — Please can I add to the agenda for the next meeting; Permission to stream these meetings publicly.
- 08.06.2018 00:13:07, Nathan Rempel | GenerEOS Sydney [@nsrempel]: lol
- 08.06.2018 00:13:52, James Sutherland - Cypherglass.com [@JamesSutherland]: IMHO the best way to deal with the over crowded meeting is to stream them. HoweverI WILL NOT do so unless the meeting agree to it.
- 08.06.2018 00:15:01, David P - EOS42 [@GunnisonCap]: No thank you
- 08.06.2018 00:15:23, Sharif Bouktila - EOS Dublin [@sharifbouktila]: I would ask if you are not a BP taking part in the Launch , or have multiple team members from a bp in this channel please leave to make our job a little cleaner.
- 08.06.2018 00:16:37, Rick Schlesinger - EOS New York [@unlockrs]: +1
- 08.06.2018 00:16:39, James Sutherland - Cypherglass.com [@JamesSutherland]: Yeah, perhaps I am in the minority. But I feel like I owe it to the comunity to bring it up. Go read the Core and BIOS Boot channels. People are quite upset about the percieved lack of communication.
- .....08.06.2018 00:19:34, zhangzc | Onechain [@zhangzc]: >>invited shiqi @eosstore [@shiqizhang]<<
- 08.06.2018 00:20:09, Douglas Horn | EOSAmericas [@DAPPerD]: If you think people are upset about not getting much information, think how upset they would be if they saw this whole Zoom chat.
- 08.06.2018 00:20:14, Lucas @EOS Beijing [@LucasEosBeijing]: welcome @shiqizhang
- 08.06.2018 00:20:56, David P - EOS42 [@GunnisonCap]: People will be just fine in 24hrs when we announce a launch. Everybody just needs to relax and keep perspective here.
- .....08.06.2018 00:22:16, zhangzc | Onechain [@zhangzc]: >>invited ligh-eoshcc<<
- 08.06.2018 00:22:49, David Moss [@DavidMoss]: Well said
- 08.06.2018 00:23:15, David Moss [@DavidMoss]: I’ve been working on this project for almost a year. It can wait another day.
- 08.06.2018 00:23:35, Craig Murray - EOS BlockSmith [@craigmurray100]: Having multiple members of the same team is a problem? If we only vote once per BP group it should not matter?
- 08.06.2018 00:24:10, Sam - EOS Dublin [@samnoble]: I’ll leave you to it from here on out!
- 08.06.2018 00:24:12, Ren Kaidi - EOSWTZ [@FoxMr]: really sorry about that, it's my misatke 😭
- 08.06.2018 00:24:32, Craig Murray - EOS BlockSmith [@craigmurray100]: for example do we kick our Marc or Alex? Both Canada, both very important, both should be here. But Canada has one vote
- 08.06.2018 00:24:36, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Only one per vote obv. I don't mind if established teams want to have more than on in here until then.
- 08.06.2018 00:24:50, Craig Murray - EOS BlockSmith [@craigmurray100]: ok cool
- 08.06.2018 00:24:53, Craig Murray - EOS BlockSmith [@craigmurray100]: thank you
- 08.06.2018 00:24:55, Nathan Rempel | GenerEOS Sydney [@nsrempel]: clarity: established hah
- 08.06.2018 00:24:57, Sam - EOS Dublin [@samnoble]: Ok, then I’ll stay :)
- 08.06.2018 00:25:03, Sharif Bouktila - EOS Dublin [@sharifbouktila]: I need your Real Name and BP team on your Profile
- 08.06.2018 00:25:26, Ren Kaidi - EOSWTZ [@FoxMr]: ok , i will rename it
- 08.06.2018 00:25:39, Nicholas Mullesch EOS.Fish [@BitcoinRootUser]: I'll add mine as well
- 08.06.2018 00:26:58, Sharif Bouktila - EOS Dublin [@sharifbouktila]: I'm not accepting direct requests from people I don't know.
- This is the process I am following
- If you are not in the Validation Channel and Think you need to be
- Join this Channel https://t.me/joinchat/HbgyfBIDVpsD-2J0g9FGvQ and have someone who either @marcantoineross @xebb82 or I know and get them to recommend you in that channel for inclusion.
- We will come up with a way to distribute the invite to trusted BPs tomorrow. It is late here.
- .....08.06.2018 00:26:58, zhangzc | Onechain [@zhangzc]: >>invited ben [@cw_ben]<<
- 08.06.2018 00:28:30, Emily Bush - shEOS [@emmybush]: Thanks for you work!
- 08.06.2018 00:29:44, Stuardo Rodríguez 🇬🇹 [@stuardo_str]: I'm part of a BP team, and we would like to stay. The main reason, is because we have shifts to be 24/7 to have all the information so we don't miss anything. We do understand only one of the team members is allowed to vote.
- 08.06.2018 00:30:26, Craig Murray - EOS BlockSmith [@craigmurray100]: because your team can only run 1 producer node... therefore you get 1 vote in this
- .....08.06.2018 00:30:51, Lucas @EOS Beijing [@LucasEosBeijing]: >>invited Lee@EOS Beijing [@shiva0187]<<
- 08.06.2018 00:31:38, Stuardo Rodríguez 🇬🇹 [@stuardo_str]: Yes, I undersand the voting, we only publish one single vote per BP, and we agree 100% with it. But please do reconsider allowing more than a single member of the BP to hear/read the discussions.
- 08.06.2018 00:33:52, jem [@jemxpat]: 👍
- 08.06.2018 00:35:32, Sharif Bouktila - EOS Dublin [@sharifbouktila]: Ok noted, please add your BP name to your Telegram profile. Going to get some sleep now 🙂
Add Comment
Please, Sign In to add comment