The Gridcoin Fireside #19 - New Superblock Developments; WCG

The Gridcoin Fireside #19 - New Superblock Developments; Project brief: World Community Grid

Speaker 1:          00:09          [inaudible]

Speaker 2:          00:09          welcome to the fire side chat today is September 13th, today is Thrusday, Friday. Okay. I am your host Nethlek. Like, uh, with me today is i like chocolate and um, yeah, nobody else is here. Jringo doesn't love us anymore apparently. And Goblin is on his honeymoon,

Speaker 3:          00:28          so just us today, Friday the 13th. Okay. So that seems fitting right. So I was the first thing, all right, the news, but let's jump right into the news. First thing was, um, alright. So the first thing was that, uh, anyone who still has grid coin on Bittrex you need to have to run them really soon. The wallet is going to be removed on October 4th and you're not going to be able to withdraw that anymore after. Um, right. What else? Uh, apparently that the TXbit integration is underway, so will be listed on there pretty soon. Um, that's the next thing. Uh, I'm sorry, I'm kind of rushing through this cause I did a, I don't really know how to elaborate on any of that. Uh, the team requirement removal testing is still underway and we're apparently still looking for more teams to volunteer for team white listing. Uh, anyone wanna add anything to that or? Pardon, Jim?

Speaker 4:          01:21          No, that's true. And by the way, it's, it's, it's the, it's on main net. We're, we're satisfied with the a test net phase of, uh, the, the team white list testing. So, uh, the teams we're looking for are going to be doing main net, uh, level testing. And, and as, as everyone, I think, I think we were talking about the last time, there are now three teams white listed on main net. That's uh, so, so we're going to add more. So

Speaker 3:          01:46          that's interesting. I didn't know that. I thought it was still on testnet, but all right, good to hear that. That's going in underway. Um, right. So the next item is apparently does a Thor Challenge run by the crunchers sans frontiers? Uh, Brenton, could you talk about that for a second? [inaudible]

Speaker 5:          02:02          yes. So the Thor challenges a, a annual competition run by the country Sans frontiers, uh, uh, team. And this year we are participating. Uh, there are 95 teams participating this year and the way it runs is, uh, each, so it's a five week event. Um, each week the top teams advance to the next week. So week one, the top 20 teams advance to the following week. Week two, I'm not entirely sure the amount, but basically you keep running and you keep doing better and keep advancing. So far we are uh, four days in the uh, world community grid. Daily stats update actually just completed about two minutes ago. So I have the most recent data. Um, the data they compiled for yesterday was that over the first three days, 355 years of a runtime was accumulated by all 95 teams. A as of yesterday, grid coin was in second place with 53 years and 201 days of crunch time. As of today's of date, we are in second place with 70 years and 290 days of run time. Uh, so all things continuing as they are, it appears that team grid coin will advance to week two, uh, pretty pretty handily. Um, but yeah, keep, uh, crunching world, community grid week one ends, uh, in three days and then a, we should be entering into week two, but we'll update you on that.

Speaker 6:          03:39          Who was that competition in first place? Right, that's what I was going.

Speaker 5:          03:42          Uh, yes, first places, uh, by 0bite. Uh, people may [inaudible] them by their previous name, which was bite ball. They rebranded as [inaudible] bite. Uh, they do pay out for a world community grid crunchers and uh, yeah, are beating us by quite a bit. Uh, the latest stats has them at 184 years, which is a more than double our runtime. Uh, but

Speaker 3:          04:08          no, because, um, they, uh, the payouts are minimal compared to ours. I mean, it's not like you're making bank with Gridcoin, but, uh, ever since they reduced them, it's really a pittance what you're getting from them.

Speaker 5:          04:19          Yeah. But, uh, so in terms of years, first places by Oh byte with 184 years. Second place is grid coin with 70 years. Third place is team China with 28 years. And from there it goes down. So you can see there's a big discrepancy between first and second and then second and third. And it's much more even from third down in terms of years.

Speaker 3:          04:42          Well, I'm doing my part what effect does the whitelist. I don't think this has anything to do with the white list specifically world community grid.

Speaker 5:          04:50          This has a, a competition that's by team. So the team whitelist would not affect this unless you were someone who was on grid coin and then left team grid coin to go join another team that was white listed. So in the future or the team requirement is removed, you could argue that yes, we would see lower results in these kinds of competitions. But as it currently stands, I don't think any significant amount of computing power has left team grid coin to join the white listed team. Someone could look into that, but I don't think it's really a, not a huge, uh, chunk of our crunching power, let's put it that way.

Speaker 3:          05:28          Right. Sorry, I misunderstood that question. Okay. So, um, does anyone have anything else? Oh yes. Right there. During a thread on Reddit about what is blink to try and find the simplest explanation to make it more easily understandable to people who are not in the know. I thought I had a good answer, but apparently I do not. Uh, anybody want to talk about this a bit?

Speaker 6:          05:51          Ah, probably you're asking the most technical people a way to dumb it down to the most simplest, which is not a particularly the best thing, but I mean we can do our best and I think this post is great because it'll allow our people to figure out a way that they can tell other people in a way that doesn't turn them off from the conversation.

Speaker 3:          06:11          Yeah. We don't have a test subjects so that we all have family members. We can test it with this and just ask them if they understood it or not.

Speaker 5:          06:19          The one thing I really liked that Jringo did was he posted it all over the place. He posted on project forums, team forums, Reddit, steam it, um, really all over the place just to really get a, a wide ranging amount of opinions. And then he kind of summarized it up. And it's really interesting to see the differences in, you know, people who are deeply involved in projects like responses from the project forums versus responses from Reddit versus, you know, the more general and [inaudible] I would advise people to look at the Reddit thread and his particularly his comment where he summarized there a lot of the responses he received. Cause I think it's pretty telling to see how different people reply and even how differently people who are all doing really the same thing conceptualize what they're doing at a basic level.

Speaker 3:          07:08          Hmm. I was another way of that. I thought it was just a Reddit post but I if I had actually read it I might've known it. So yeah. Anyway, does um, anyone else have any news? No. Alright, moving on. Um, Delta, we'll be doing another project brief now. I think this time it was, world community grid, couple of sub projects. Uh, go, we have Delta.

Speaker 6:          07:30          Yeah. All right. Project brief time. All right, so the project brief is a little show run by me on the gridcoin fire side. Um, almost every week now. Unfortunately, I've been caught up with university and a bunch of programming I've done the most programming have ever done in my life the second semester of uni. So, uh, here I am again. I'm going to talk about a world community grid today. So I'd asked, uh, if everyone leaves their questions to the end and we can have a little bit of a discussion about the project and what it does, uh, after I've finished the segment, just before we get into it, there's been a bit of science news, so hopefully everyone he knows about cyclone Dorian that hit The Bahamas and a little bit of Florida or something like that. Uh, and uh, what happened was, is they wanted to do an, uh, they needed emergency response and they needed quick emergency response.

Speaker 6:          08:19          They need to get supplies down into the ground, but they couldn't find anywhere to land cause it was all debris everywhere. So what happened was, um, I think the national guard or some other organization released a whole bunch of satellite imagery and they gave it to zooniverse.org um, a site for volunteer science. And they said to all the, all the people doing science on there, Hey, find us some helicopter landing sites. And so within hours of posting all those satellite images that completed that entire dataset and found some really good helicopter landing spots that actually helped in the, um, relief efforts for cyclone Dorian, uh, I'll post a Twitter link. Um, right now currently they've topped up the project with a lot more data, but, uh, even as soon as that project came out, within hours, it already got completed, which is phenomenal. Anyway, let's get on to world community grid.

Speaker 6:          09:13          So, um, as was mentioned in the news, we're currently running the Thor Challenge, which includes which involves will community grid. Um, and a world community group was a project started in 2004. And uh, it's currently run and owned by IBM and they do a whole bunch of different projects. This is a project for projects. Pretty much. The great thing about these projects is that they directly apply to, um, our reality. So they try and tackle the practical problems that we're facing. So, for example, currently most of the projects that are running are for uh, biological and, um, viralogy and all sorts of, um, stuff like that trying to cure a disease. And it's, well, community grid is pretty much the go to project if you don't know what to do. And it has a large variety for you to choose from and it has the most practical sort of things.

Speaker 6:          09:59          So it makes, it makes it easy to sort of talk to people and say, Hey, my computer's trying to cure HIV or my computer's trying to cure cancer. Instead of saying, Oh my my computer's doing foster, we had transforms to search for aliens. So let's start with the sub-project one, a mapping, mapping cancer markers. So what a cancer market is, is just sort of like a little segment of like DNA or a little segment of a protein that sort of identifies what cancer cell you're looking at. Now, the important thing about this and why will community grid is doing a project about it is not only can you identify the type of cancer you can predict how it's going to progress throughout the human body and your, and the most importantly is that you can determine what treatment to give the person. And so this project is all about taking those cancer markers and associating them with cancer cells and results from a cancer screenings, results from treatments and all sorts of stuff like that to find the most effective way of understanding its progression and find the most effective treatment for the person.

Speaker 6:          10:56          And so your computer, when running this project go through billions upon billions of calculations just matched and just mapping a cancer markers to cancer cells, mapping cancer markers to treatment effectiveness, mapping those to that. It's, it's a really important project because if we can stay on top of cancer, we can save lots and lots of lives and hopefully we can find a cure in the end. And so it's all about finding patents. It's finding that association between the marker and the cancer. And through this project we might even get to the stage where we can personalize treatment for cancers. And so what would happen is supposedly if you get a diagnosis for cancer, you just send it into, world community grid, we'll send it into a little scientific shop, get the computers to associate it and look up in the database of what that cancer is.

Speaker 6:          11:42          And then a Presto. You come back within a day at the most and you say, Oh, okay, yep. This is your treatment plan. This is how we predict the cancer's going to progress and this is how we're going to get it out of your system. And the importance of this, of course, is staying on top of cancer in the long run. That's what we want to do. Alright, so I'm not going to go through all the projects, but the last sub project we're gonna talk about is fight AIDS at home. Uh, it started all the way back in 2005. And, uh, according to world community grid. It was the first project to run on Android devices for, uh, world community grid projects was the first one that they were trying out for Android devices. Now, according to the statistics in 2004, over 40 million people had HIV worldwide. And this project covered two phases, uh, to try and tackle the rising problem of HIV phase one, um, uh, is where they went through a whole bunch of protein simulations.

Speaker 6:          12:35          So this is, um, sort of like how we were talking about Rosetta at home. You get a big long strand of protein, uh, virtual, your computer will think, think of the protein and then your computer will sort of merge it. You'll, it'll, it'll squeeze it twisted or, uh, do everything at once with it to try and get into the right shape to fit into the particular arrangement of the HIV virus. And so in phase one of fight AIDS at home, over 20 billion protein docking simulations occurred. So 20 billion arrangements of proteins were fit, were calculated to try and either stop or attack the HIV virus. And the purpose of phase one is to identify all the promising candidates so they get rid of all the ones that don't work well, that can almost guarantee won't work. And they only keep the ones that they say, Oh, this looks pretty cool.

Speaker 6:          13:27          We might keep this and we might test it now, thanks to examines these valuable candidates more closely. So these are the ones that look like that they have an effect on the HIV virus or the HIV proteins. And uh, what they do is they just repeat the same process in phase one. Keep doing those simulations to try and nail down and get the really, really accurate results on these particular ones so that they can probably [inaudible] probably wipe out a good couple thousand from them and leave scientists with only about a hundred proteins left to test, which means that the scientists in the labs only have to test about a hundred different proteins, which saves them thousands upon thousands of dollars just testing these things. And that's the importance of this. Um, this fight AIDS at home project from our community grid is because it just saves scientists money. It saves scientists time, it makes the data more accurate and reliable so that we can hopefully find some sort of cure to the HIV virus.

Speaker 6:          14:22          And so in, in retrospect, it's instead of testing 20 billion different proteins in the lab, we're now testing only a couple hundred. And uh, so that pretty much concludes the project brief. World community grid is a white list of project and as I said before, we're currently running a challenge. So if you want to help out with the grid coin team, we could use your hours of computation to ah, fight, fight cancer or fight AIDS. Now, if you don't have a good computer, you can always try zooniverse at zooniverse.org uh, a place where you can use your brain to solve scientific puzzles and help. Um, as I said earlier in the project brief, you can even hope with disaster relief, there's all sorts of stuff ranging from physics to maths to identifying animals. You can even measure giraffes on this thing. Um, it is, it is a big brain project, but I mean it is a big brain project.

Speaker 6:          15:13          I'm also, um, I have a Twitch channel where I do, uh, volunteered science, so you can check me out. twitch.tv forward slash one five one two Delta. Um, I'll be on occasionally, hopefully I'll be on a little bit more than the next week cause I have a little bit of a break. Um, and yeah, that's where I'll be doing volunteer science on zooniverse.org folded and uh, there's a new uh, RNA one that, um, robotic mind actually mentioned to me. So I'm excited to get started on that one anyway. Um, I'll see you next fortnight. I still haven't decided on a project to get going then, but hopefully we might choose the math project I feel like doing in the boot moments anyway. Now time for questions and discussion. So, um, let me just scroll through the chat and see if we can find any other things.

Speaker 3:          15:57          I'd like to, uh, add a little something. Um, when you talked about the cancer marker thing, you said finding a cure for cancer. The thing is cancer is not diseases. It's more like an umbrella term for a number of very different diseases that just behave similarly or sort of more or less arise in some of the circumstances. But yeah, I don't, I don't think we're ever going to find it. The cure for cancer, so to speak. We might, uh, for specific cancers and you know, some point down the road we might find a cure for each and every one of them. But, uh, I don't, it's a bit pedantic, but I just wanted to, uh,

Speaker 6:          16:34          no, no, you're completely correct. And what I mean by a cure to cancer is finding some commonality between all those different types of cancers and attacking that one commonality. That's sort of like what scientists are trying to do. Um, in quote, finding the cure to cancer. They've want to find that one commonality between them. But as you said, it would be a, it is currently a lot more easier to target specific categories of cancer. The important thing about cancer markers is that we can provide patients with the best treatments for their cancers. Any more questions? Uh, I the, you can post one in the chat or you can, uh, just speak one here and I'll see you about answering it.

Speaker 3:          17:13          Any idea how many, um, IBM people work on?

Speaker 6:          17:16          Well, a, I'm not too sure of the exact numbers. All I know is that they do have support and funding for them. So the project is probably not going to go away for a while now. Um, IBM have the sort of innovation division working on this, so they're always finding new projects to bring in. I know, um, in history we had one where it was like a, um, water managing water or something like that. They, they had a project on managing what are you probably thinking, what, what's my computer going to do to like help people manage what up? But you'd be that it can actually be, um, reduced to a computational problem and it's something that you computer can do to help manage water in third world nations.

Speaker 7:          17:57          When was that? When was this water? This is really interesting.

Speaker 6:          18:01          Yep. Uh, let me just, uh, quickly set it up.

Speaker 7:          18:03          That would mean a political blank project.

Speaker 6:          18:07          Ah, okay. So for some specific information, it's called, uh, computing for sustainable water. So, uh, and I'll just read here a little bit. Um, if anyone has any questions on more discussion, feel free to post in the chat. Okay. So yeah. Uh, it looks like a, this project was running some sort of, um, heuristical algorithms, so it was making sort of guesses and trial and error to figure out how humans sort of, um, use water on a large scale and sort of, um, gain as it says, gain a deeper insight into what actions can be done to, um, help sustainability of water in those nations.

Speaker 7:          18:44          Anybody else heard of other projects doing somethings? Human psychology and

Speaker 6:          18:49          um, no, not that I've heard of. And it is interesting because IBM has done a lot of very different projects compared to everyone, uh, compared to all the other products because most of the other BOINC projects select yet, this is definitely something for computers. This is rigorously computational. We've got to search for aliens, we have to do a fast Fourier transforms and go to fold Proteins, we run simulations. But with weld community grid, it seems like they do a lot of really interesting things. So they, as I just said, computing for sustainable water. Um, uh, there was a clean energy project apparently. Um, another one for clean water. Um, and uh, yeah, most of this stuff does specialize in drug research, but they do have some interesting things that you would never have thought would be able to be computed on a computer.

Speaker 7:          19:35          I think it's because they're, they have a lot of connection, different units.

Speaker 6:          19:39          I think so too. And they, they have, they have the funding, they have the connections and they have everything like that. So it is a very, um, fruitful project.

Speaker 3:          19:48          Right. So thank you delta for the project brief. Um,

Speaker 6:          19:52          no worries. Uh, kind of wait for the next time.

Speaker 3:          19:55          Right. Um, let's head on into the main topic, the development process. Jim Barton, when are you to want to take that? I heard there was a new class of super blocks, but I'm not familiar with what changed.

Speaker 8:          20:08          Well, that's, that's a long, yes. So let's, let's talk about where we are in general. Um, we have, um, you know, we, we next mandatory, you know, as the next release and it's a mandatory and uh, we're trying to stuff in all the things that we have to do a, that's going to require a mandatory cause we don't release mandatories very often, but, uh, uh, one of the things that we have been working on, uh, is to improve the way is super for blocks are actually, uh, generated and recorded in the blockchain. So, um, you can think of, uh, the scraper, uh, what JRingo likes to call Oracles. Cause you know, we're, we're going to label it, what does this new term of art that the blockchain communities are using for, um, devices that gather off chain data and put it on chain, but the scrapers, um, we're phase one of that to be able to stabilize and harden our ability to collect statistics from, from the BOINC projects.

Speaker 8:          21:14          That's phase one and got all that stabilized. It's working great. Phase two is to, uh, change the way the scrapers were created. Uh, essentially outputting the, uh, statistics and generating the superblock in what's called legacy mode. So they actually, after they get through with everything that the individual nodes that stake the super block are staking superblock in pretty much the way that it was staked with the old neural neural net. And we're going to change that in, in Fern. And so a few different things about this. Um, number one, the old way that the super block was created and put in the blockchain is very inefficient. It involves extra steps that we don't need, which is this, uh, voting process that was, was instituted. And that was largely put in place because the statistics gathering was so unstable that a, you really couldn't depend upon every note having the statistics.

Speaker 8:          22:14          And so instead of dealing with that problem directly, they, they put a voting process in. So only a few nodes would have to vote to actually record the statistics in the blockchain, but that also introduces vulnerabilities that, you know, uh, it's just not as good security wise. So where we're headed with this is, um, the, uh, now that we've, we've gotten the scrapers in place and every note is getting a reliable copy of statistics information from all the scrapers and they're able to validate that statistics information. We don't need the voting processing more when a node goes to create the superblock and stakes, the superblock, every other node that receives that stupid superblock can verify it directly based on the statistics information that they have. And this is a much more straight forward, much more reliable and in fact much more secure, uh, method of, uh, recording and validating the roadblock.

Speaker 8:          23:10          And in fact it's more distributed because every node, just like a regular block, every node is going to be doing the validation. So that's where we're headed now. In addition to changing the process on how that works, we are, or we are also changing the format of the superblock itself. So it turns out that some less than optimal choices were made in terms of how the data format was actually done for the super block. And it causes the superblock to be very inefficient in terms of processing it. And any efficient in terms of the storage. So one of the things that we've done is we created a new super block class, which also includes new ways to store the data in a compressed way on the blockchain, which reduces storage space. And a, the class is also going to enable us to enable on protocol, on chain, uh, gray listing and some other things that we've been wanting to do for quite a while.

Speaker 8:          24:04          Uh, related to, uh, statistics, uh, We're probably sewer block class itself. There's a few more things that need to do to it. Uh, we've actually been testing super block class, uh, on testnet, uh, using some pretty clever, uh, test algorithms to be able to, uh, accurately simulate what the validates be like on the reading nodes, uh, on receiving the superblock information. So, uh, and, and we've massaged, uh, we believer most of the bugs out of that. So we've made some exciting progress. We've got a long way to end by, by, by the way, the superblock stuff is, is only part of what is Fern. There's a lot more stuff that's going in. If you're interested, you know, you can go on the, you can go on and get on, git hub into the Fern milestone and see the kinds of things that are there is quite more than we've got to do. But look, we're making good progress with, so be exciting and you know, uh, just a comment on what was totally it. It seems like the uh, Elizabeth milestone really went very smooth. We're seeing know we're seeing very few problems from it. So that allows us to concentrating on getting Fern out the door rather than worrying about problems remarkably. I don't know if [inaudible] wants to add any thoughts or update.

Speaker 9:          25:12          Oh, that's nice. I can do or no,

Speaker 10:         25:14          what, what he should add is that we've been saying his day abroad for months,

Speaker 8:          25:17          I think at the CC is cycy

Speaker 10:         25:23          yeah. Everyone's been saying his name wrong since

Speaker 8:          25:25          Trevor. He says he doesn't mind them in a bit. Do you have anything to add to that?

Speaker 10:         25:32          That goes far over my head, so no, I defer to Jim.

Speaker 8:          25:37          So any, any questions on that from folks? It's a, is very technical, but it's a pretty big deal.

Speaker 9:          25:44          So can you somehow, can you quantify how much

Speaker 8:          25:49          yeah, we can barely hear you. I can't,

Speaker 9:          25:51          can you hear me now? Yeah. Um, I was going to say, can you, can you quantify empirically how much better to this as the old system?

Speaker 8:          25:59          Oh, what in what way?

Speaker 9:          26:01          Um, is it, how much faster is that? How much less memory does it take up? Um,

Speaker 8:          26:05          so, so the answer is that the current, it looks like the current super block Oh, serialization is a process of taking an object packaging, writing it out as a store it or send it over the network. Okay. And so when we serialized super block class methods that we're doing what your native, uh, we're super block size of about 15% smaller than the Old binary pack format that for the old, for the old superblock.

Speaker 9:          26:35          So just to be clear, is that 10 or 15% smaller? 10% 10 15% the size of the old super blocks

Speaker 8:          26:41          10 or 15% smaller. So it's about 85% OSI. 85 to 90% the size of the old format. The thing you've got to realize is we're actually more information despite the smaller size. So there, there are, there are extra fields in the Superblocks that are in there to support, um, gray list protocol work that we're doing. Plus, uh, uh, the, uh, other things that we're going to be doing with the, uh, research rewards calculation that, uh, uh, would be, would actually, if you tried to do that with the old format, it would probably make format maybe 10 to 20% larger. So in reality, the Delta, you know, probably more like 30% difference in terms of size. So it's pretty, it's a pretty big, uh, three big advantage. Um, the other, the other thing that's important to note as a serialization because it's using native C++ methods doing that that were implemented in some of the original, essentially the Bitcoin classes for serializations. There it is far faster, much faster than the old, uh, binary packing. So it's going to enable us to process Superblocks in memory much faster than the, than the old way.

Speaker 5:          27:50          And in terms of measuring performance differences, a lot of that will come when it's actually, you know, fully realized and put into testing. A lot of that isn't, you know, you can extrapolate a lot of that on paper, but the real world performance is a lot harder to estimate.

Speaker 8:          28:07          That's right. We are, we are running the court of devs are a couple of crippled nodes on main net, which don't, can't stake, but they can actually do test forms of the super block that are just local node. So we're actually able to see what their performance of packing is for the full for the full set of CPID on main net. And the improvement is dramatic. Uh, but as Barton said, it's to put it, you know, we won't know exactly how this, to get all the pieces put together. Uh, but you know, we have a good idea, a good, um, one of the biggest issues we've had in the past of the, uh, you know, one of the challenges we have is don't stay very often, say a month or two months or three months between stakes. A person's magnitude can change a lot during that time.

Speaker 8:          28:50          And so it could imagine you've gone three months, three months of Superblocks, three months worth of magnitude changes that, that, uh, has gone during that interval time. Well, the current situation, you know, the superblock every day, so you can imagine it roughly a stuff. There's three months worth between someone's taking once and again, you're going to have, you would have to go through, if you wanted to do a very precise calculate and you would have to go through and look up the magnitude of that person and each and every superblock and then put it, put together their research rewards on a day by day of using the magnitude to superblock. Someone may have some folks who additions to development work may have seen that before. It's called the super block West. Work at Ramon was working on, um, the issue is we were never aware of, actually we'll put that into place because when we tried to do it, trying to pull in that many super blocks was so slow that you'd never actually have the computation and drag everything to a halt.

Speaker 8:          29:42          So there was no way we can do that and to prove the way the superblocks were stored and a serialized or your wise but a, so what we've had, you know, historically we've had that it's either a 2 point average or 3 point average. We're just a two end points magnitude. The average lifetime average just kind of like an average of the magnitude in the middle were used as just an average reward if they haven't staked it a long time, that that is really not do that. And we really want to words the super block windows approach to a research rewards, which is far more precise in terms of paying people that haven't staked a long time. Uh, so you know, this all fits into that. The ability to prove the, the reward algorithm. People that you don't know stake very often. It's pretty, pretty exciting stuff.

Speaker 6:          30:27          Um, I have a question. So in terms of just the general user, what the general users are going to see, is it just the increase in performance or is there any like inherent sort of increased chance of staking sort of thing?

Speaker 8:          30:41          Depression, a lot of the, the sad part about this is a lot of this is going to be largely is largely machinery and the engine, um, but may note is they get rewards that, that are more accurate, that feel better or right to them. If they don't stay very often and their magnitude goes up and down a lot, they're probably going to notice that their rewards are more accurate, would match better what they expect. Then the way the wallet currently works, people that say only stake once every 30 days or 60 days, I would say, for people who stake a lot. Very often, no difference. [inaudible] and we will see a little bit lower CDs that she, uh, I mean the CPU usage of the wallet today. So, I mean, we're talking about going below, but there will be, uh, a reduction CPU use is just simply because the process of superblock, but you know, you're talking about the wallet uses probably from an average of less than hello. You guys there?

Speaker 6:          31:32          Yup. Yup. your connection has not been very stable.

Speaker 8:          31:36          So the, the wallet is pretty much to began with probably less than 1% for most people. So you know, you're talking about actually we'll see a reduction.

Speaker 6:          31:44          Okay. Yeah. And I've also, um, in, in the past have also heard of people having CPU spikes and stuff like that. So hopefully that will no longer be an issue.

Speaker 8:          31:52          I think we'll see less of that. And we may actually see an improvement in sync, uh, for the [inaudible].

Speaker 6:          31:57          All right. It would affect sync speed. So when you, when you start up the wallet, it'll, um, get and get gone much faster. Right?

Speaker 8:          32:04          Well, probably not bad. We can't, can't change the format of the Superblocks that are in the past. What will happen in new part of the chain and record the new super black version. That part of the change should sink faster. So we'll start to see a lift as, as to more and more, uh, blocks cause it will definitely be faster. Uh, although, you know, look, the super block only happens every, I think it's averaging once every 25.5 hour like that. So, you know, it's roughly once every thousand blocks. So you know, it's going to be, is only going to be mildly, but, uh, look, if it helps, um, you know, it's pretty interesting now, um, you know, this and in concert with beacon improvements that we're doing are going to allow us the full, you know, no team requirement that we're really tall stepping stones.

Speaker 3:          32:48          Are you going to have a release party? Get some champagne going? Yeah, maybe that's a good idea cause Ferns, but it's a, you know, a really good release out the door. It would be nice if we could properly coordinate it, you know, with the announcements and everything.

Speaker 8:          33:06          Well, I'm not sure what that means. I mean, I think that a pretty good job with them.

Speaker 7:          33:09          Was there any going to be any, um, I guess documentation release? Are there any like for example, the command changes or um, I guess the commands didn't really change but I don't know other things maybe a users should be aware of.

Speaker 8:          33:24          Well, for this for, so yes, for the other parts of Fern were changes needed to the procedure to advertise your beacon? Yeah, we're going to have to, we're going to publish a document document. The super block chains are largely, you know, other than technical documentation, that's not really going to be a, I will put a plug out. I mean our document only badly out of date across the, and you know, core devs actually need help from other People with the documentation. Really don't have the time to do all that ourselves. So we need volunteers to actually know stuff that actually could go in there and fix it. I'm talking about the wiki needs to be updated. I mean there's a lot of stuff that's really this at point quite out of hand and we're looking for volunteer can help. Not, not everything have the, you know, a developer's level of knowledge of code to be able to,

Speaker 7:          34:06          yeah, I think I would like to do that. But, um, I guess haven't been paying too much attention to the actual code so it's a little bit tough. But um, yeah, I know, I noticed also like the documents are all over the place as well and uh, yeah, I think people should really try to do some work there.

Speaker 8:          34:24          Part part of it is like, you know, really I would like, I'd like us to a place where it's either the main website or the git hub, which is where all should go for all their information really that that's where we need to concentrate the updates. Right. And I know not, not that I don't approve sources that have helpful information. You know, the challenge is to, more of the information is spread out among the sources. The more places you have to update and then you end up with information that doesn't agree. And

Speaker 3:          34:51          it varies because

Speaker 8:          34:53          our resources on gain website insurance, it's up to snuff and then our get hub page in our Wiki pages to make sure they're ready.

Speaker 3:          35:01          There's a lot, there's a lot of stuff there just to be updates. Yeah, I agree. Any more questions? Yeah, I can't, I can't run driving chip Jeff scrolls or any other questions. I can't believe it. Nobody has any other, I have a question. Can you repeat all of that really fast? Cause I have another 15 minutes to cover.

Speaker 8:          35:21          Uh,

Speaker 7:          35:21          I've got a question. How about how about how about dependencies or, um, other releases going to go up? Uh, I think that the, I think we are pretty okay with, um, how updated we are up like with the upstream

Speaker 10:         35:38          I guess. Um, there isn't too much issue there. Um, yeah.

Speaker 8:          35:42          So, uh, w what do you mean with respect to the latest? The latest would, could the latest Bitcoin codebase, are you talking about the dependence, the package dependencies that we need? So

Speaker 10:         35:51          yeah, the package dependencies, I guess the basic standard C C stuff, right?

Speaker 8:          35:56          Yeah. So the answer is really going to be no change. Well, well we are probably, uh, our goal is in jail. A SSL in the uh, depends version. Do 1.1.1 probably Charlie was the latest. Um, from 1.01K just what's currently packaged. It depends version. If you're running the wallet, that's two year operating system and you're running Linux upon version of Linux, like you know, eighten oh, four, 19 0 four with Ubuntu, you're already running 1.1 point and it runs fine. So it's just a matter of us getting that updated. But in general, the compiler verge required and dependency requirements between Fern and Elizabeth. Exactly. So if you can compile Elizabeth 4.0 0.5 points, you'll be able to compile. Fern was 4.1. Uh, we don't tend on doing anything that's gonna increase the, uh, dependency. I will say, I will tell you that I spent a while at this four point, well it was when the scraper with 4.0 0.3 0.0, um, be a a G less than five, we'll compile the loaf that distributions that are using, uh, don't have a, where you can install GC fiber already not able to compile the wallet.

Speaker 8:          37:04          So that's, uh, that's, that's probably this chain in the last that we had in terms. And that just has to do with the fact that, uh, we're really use a full C++ 11 standard. And if you're using GCC less than 5 does not, uh, C++ 11, so it won't work.

Speaker 10:         37:21          It's worth noting that you really are, would be hard pressed to find a system that can't upgrade to GCC five in today's world. Right.

Speaker 8:          37:30          Did it have to be really old? Right. I think we're talking about for four right now, but to light, right brightburn?

Speaker 10:         37:35          Yeah. Yeah. Probably so, like many years back. Yeah. Yeah. Cause I think even 16. Oh, install five, yes. You can install it. Yeah. As, yeah. It didn't ship with it, but it could install it. Yeah.

Speaker 8:          37:47          [inaudible] an optional. So, you know, look, we're, we're careful to not try and onerous upgrade requirement or trying to support the broadest range of platforms, both from a [inaudible] perspective and also from a complimentive. I mean, I think we, we really pretty flexible in terms of the, uh, of the, uh, platform support for both, either if you could compile yourself, we're going to run it, you know, a prepackaged executable. And I'll just remind everybody, you know, stored 32 bit windows. That's a good thing too. But you know, the goal is to allow the wallet to run on as wide a platform range as possible.

Speaker 10:         38:20          So what can we expect the Atari 2,600 release? What I said, when can we expect the Atari 2,600 release? Not going to happen. Yeah. According to Wikipedia, the Atari 2,600 had 128 bytes of Ram. Unfortunately, we're a little, a little short on the Ram requirements there. Yeah. Yeah. You guys will figure it out. It's not kilobytes, not megabytes. That's 128 bytes of right now.

Speaker 8:          38:46          Yeah. I don't even know how, I don't even know how they made some space.

Speaker 10:         38:49          You should have better luck on your calculator. Run it on your calculator. There you go. That's right. What happens with, um, people with the old wallet one step upgrade happens. Do they?

Speaker 8:          39:02          Okay. So, so yeah, so this is the way a mandatory is not like the leisure is exactly that, right? You don't have to install it. And, uh, you know, the idea is that, you know, between mandatory is when you do a leisure upgrade, someone can upgrade it at your, if they want to is perfectly fine. What that means, we're saying, in other words, that set the leisure wallet firms are blocking protocol compatible, right? And so, you know, a wallet, you're running a four Oh three, Oh, a wallet should run fine on the network, you know, without any trouble. You may have some loss of functionality, but it's gonna it's gonna sync up, but it's going to run when a mandatory goes in place, a mandatory is accompanied by essentially what is, what is a, a, uh, encoded fork and what that really means. There's, you know, the developers when they decide, Hey, we're going to demand auditory.

Speaker 8:          39:49          They will encode in the release a archive who set a protocol to effect whatever the changes you want. If, uh, what happens is this, if you have a node and you have it upgraded, by the time that block hit your node, it's stuck and it will remain that way. Until you decide to upgrade. And, uh, that's the way it works. Um, conversely, network could rebel and decide they don't want to do an entry upgrade and they don't have to do it, you know, but, but in general, you know, if it's done right and the communities will, and we're doing things as a community that almost everybody up them. So in practice, what happens in the network is the nodes that failed to upgrade by the time the fork point hits, they just get stuck and I'll, they'll remain stuck and fork off or decide to upgrade. Does that makes sense? Um, so you know, and by the, by the way, we have to get this, why we have to give exchanges plenty of notice cause they have to have time to compile up, correct. Test for the testing, whatever they want to do and then get the new wallet in place for that blocktime. And so there has to be, I mean, I think the general rule is Barton remind me of sports. I've noticed right?

Speaker 10:         40:48          Two weeks. But we strive for, we strive to do three or four if we can, but it's always minimum two weeks. Yeah.

Speaker 8:          40:54          Right. Just to make sure that, you know, we don't, we don't fork off a wallet exchange. It's not a good, so, um, that's the way it works. And Oh, by the way, we don't, we try to keep to like at the most it would be two a year. We really, we really try and minimize the mandatory is right. So you really have to think very carefully about what goes into mandatory cause you really don't want to be doing these,

Speaker 10:         41:13          but we don't need to I guess, um, remove the database or anything like that. Just

Speaker 8:          41:20          Oh no, no, no. You don't touch the data by ideas. You don't touch it. You don't want to [inaudible]

Speaker 10:         41:24          that's scary. No, you don't touch the database.

Speaker 8:          41:28          If someone, if somehow, uh, mandatory went bad and then, you know, you blow up the, you know, we don't want to do that. So, you know, the idea around the mandatory to set the block height in such a place that you know, people have a chance to upgrade, but for the block height, so the, the, the switchover is essentially transparent, uh, and that, and then basically it's, it should be easy just you know, you'll see a luggage, new protocols go on. So is there if a everything goes right, but you know, look, that's, that's why we have a testnet. We do. I have to say, you know, we have a really well run testnet that uh, you know, our test has got quite, quite a good to now and it allows us with a lot of testing. They quite frankly in the old days was very difficult. And I'll tell you what, dark good test as a secret. It's part of the secret to why our code has been so much better. I mean, it's not just coding to good job and writing high quality code and eliminating bugs. I think it's also that we're going to got good job testing and I think it's showing and that's why we're seeing very few problems, which is I think a huge break past. Good.

Speaker 3:          42:28          Yeah. Good work. Excellent question. All right. Anything else? Okay, let's start. If we can wrap this up now. We have almost an [inaudible] I think [inaudible] will be happy that this time you won't have to work through two to three hours worth of stuff. All right. Thank you everyone.

Speaker 2:          42:44          And um, I just want to say one last time that the Bittrex warning, if you still have clients on there, get them out of their SAS. There wallet will be removed on October 4th. You won't be able to recover your coins

Speaker 1:          42:58          once that happened. All right, then. have a good night everyone.