<@coblee> Luke-Jr: what's up?
<Luke-Jr> coblee: is there any chance you would be willing to shutdown litecoin, or request mtgox to not trade it, for the sake of Bitcoin?
<Luke-Jr> (pending further discussion of course)
<@coblee> Luke-Jr: you do know that no one (not even me) can shutdown a decentralized currency, right?
<iddo> Luke-Jr: it's good to test social behavior with litecoin, suppose for example that litecoin and all other alcoins didn't exist, and you invested in bitcoin, you wouldn't know what will happen to your invest when suddenly some new alt coin like litecoin would appear
<Luke-Jr> coblee: yes and no
* C0deMaver1ck (~C0deMaver@unaffiliated/beau/x-8351432) Quit (Ping timeout: 264 seconds)
<Luke-Jr> iddo: the threat isn't litecoin per se, as much as it is MtGox trading litecoin
<iddo> coblee: maybe Luke has this in mind: http://www.reddit.com/r/litecoin/comments/1bf2l6/ask_for_litecoin_on_mtgox_and_on_every_other_btc/c96abou
<@coblee> and if somehow litecoin does die, some other alt coin will take its place
<@coblee> you should talk to MagicalTux instead then
<Luke-Jr> coblee: MagicalTux is not discussing it
* C0deMaver1ck ([email protected]) has joined #litecoin-dev
* C0deMaver1ck ([email protected]) Quit (Changing host)
* C0deMaver1ck (~C0deMaver@unaffiliated/beau/x-8351432) has joined #litecoin-dev
<@coblee> i thought it was just an april fools joke
<iddo> i think that it's just the marketing dept. of mtgox
<iddo> not sure about april fools, two people there said that they got same reply
<Luke-Jr> coblee: currently my only option is to coordinate a FPGA-based 51% shutdown; I'm hoping to find some other more friendly avenues
<[yAK]> https://data.mtgox.com/api/2/LTCUSD/money/ticker
<Graet> he is goxxing ltc, price has tripled since the gox announce - what will happen if he pulls support now? major crash....
<[yAK]> ^ ill just leave that there
<@coblee> might be a mtgox pump and dump. buy a ton of litecoins, announce rumor of supporting it, adnd dump
<iddo> seems more like standard marketing mumbojumbo
<Graet> Luke-Jr, how about not supporting scrypt in your miner - if you really want o do something....
<Luke-Jr> Graet: explain how that would change anything?
<iddo> Luke-Jr: do you have a rough estimate of the cost of that 51% FPGA attack?
<Graet> well seems strange you running around calling ltc a scam, aking coblee to take it down, but you support it in your miner....
<@coblee> Luke-Jr: I don't think you will be able to shut Litecoin down without losing a ton of money. And you'd only be able to do it temporarily. look at bbqcoin, it's back too
<Luke-Jr> coblee: it doesn't need to be shutdown permanently, just enough to make MtGox reconsider
<Luke-Jr> iddo: existing Bitcoin FPGAs are sufficient, just a matter of the right bitstreams
* rdponticelli ([email protected]) Quit (Ping timeout: 248 seconds)
<iddo> Luke-Jr: how many FPGAs could you control?
<Luke-Jr> iddo: 10 FPGAs is enough to get 51%
<iddo> i doubt that figure
<Luke-Jr> admittedly, it's not proven in practice yet, but I doubt it would vary significantly
<@coblee> Luke-Jr: that's funny
<iddo> how much faster is a typical FPGA comapred to say 5970 ?
<Luke-Jr> coblee: ?
<@coblee> are you working on a script fpga?
<iddo> Luke-Jr: don't forget that a lot of GPU power is probably migrating to litecoin now, because of bitcoin ASICs
<Luke-Jr> coblee: I know multiple people who are thinking about it, and who are willing to take action to protect bitcoin
<@coblee> afaik, fpga are equivalent to gpus. maybe a bit more efticient
<iddo> there was forum post on scrypt FPGA, i don't remember if it was in final stage yet
<Graet> funny thing Luke-Jr - there are plenty of gpu miners that would jump over if you attacked....
<Graet> but really Luke-Jr stooping to altchain type wars isnt good fort bitcoin....
<@coblee> right now, you can make 5x the bitcoins mining litecoins: http://dustcoin.com/mining
<iddo> Luke-Jr: multiple people? i'm not aware of anyone else hostile to litecoin as much as you... who?
<@coblee> i'm sure a lot of gpu miners are switching over
<Luke-Jr> iddo: they'd prefer I take the backlash of course
<iddo> ok
<Luke-Jr> coblee: anyhow, back to the topic: are YOU willing to do anything to protect bitcoin?
<iddo> but you'll probably need 100s or 1000s of FPGAs, not 10
<iddo> Luke-Jr: i don't think that most people agree with your reasoning, the litecoin hurts bitcoin
<iddo> Luke-Jr: could you please explain why you disagree with the reason that i gave, about testing human behavior towards altcoins in order to get real price discovery for the value of bitcoin?
<iddo> s/the litecoin/that litecoin
<@coblee> Luke-Jr: I don't think Litecoin hurts Bitcoin
<Luke-Jr> iddo: again, the problem isn't litecoin as much as it is MtGox trading litecoin
<@coblee> if it did, i don't think btc will be at $100 today
<Graet> well fix gox, dont break litecoin
<Luke-Jr> litecoin is clearly a pump and dump scheme, and when it fails, the fact that *the* Bitcoin exchange was promoting it is going to reflect terrible on bitcoin
<Luke-Jr> people are going to just write off all cryptocurrency as pump & dump
* GordonG3kko (~GordonG3k@gateway/tor-sasl/gordong3kko) has joined #litecoin-dev
<iddo> ok now at least i understand your reasoning
<@coblee> Luke-Jr: no need to attack litecoin. quoting you: "the existing network effect of Bitcoin, combined with the lack of meaningful differentiation between Litecoin and Bitcoin and Litecoin's adoption of a "designed to fail" proof-of-work algorithm; that Litecoin is bound to fail in the end."
<Luke-Jr> even if you disagree with it being a scam, are you willing to stake Bitcoin on your opinion being right?
<@coblee> so in time, it will fail by itself. unless you don't believe your own lies on the wiki :)
<Luke-Jr> coblee: yes, it will. but with MtGox promoting it, it will kill ALL cryptocurrency at the same time
<iddo> you describe what possible scenario, you seem to be convinced that it's inevitable scenario, but most people (at least here) disagree
<iddo> s/what/one
<@coblee> Luke-Jr: i don't have a god complex such that I believe I have the power to kill or save bitcoin
<Luke-Jr> iddo: are you willing to bet Bitcoin's existence on that?
<Luke-Jr> s/existence/success/
<@coblee> mtgox promoting litecoin and litecoin failing will not kill bitcoin
<Luke-Jr> coblee: think about it, I gotta run to the airport and back
<iddo> Luke-Jr: i disagree with that dichotomy because i think that having litecoin is helpful for bitcoin
<@coblee> 1 exchange + 1 alt currency should not be able to kill bitcoin. if it does, then bitcoin had no chance from the start
<Luke-Jr> coblee: from the perspective of a normal person, who sees MtGox as the face of Bitcoin
<Graet> i agree with coblee , if bitcoin is that fragile its doomed to failure anyway
<@coblee> you're too short sighted. maybe it will hurt bitcoin a bit. like the mtgox hack caused it to drop to $2. but in the long term, bitcoin (and litecoin) will survive
<@coblee> anyways, i got to go. we can talk later if you want
<Luke-Jr> coblee: you're being (unrealistically) optimistic there..
<Luke-Jr> how can you be 100% certain litecoin will survive?
<@coblee> who's 100% certain of anything?
<iddo> Luke-Jr: maybe you should think about it from another perspective: if bitcoin cannot survive the litecoin (supposed) crash, then bitcoin is too weak to be allowed to exist?
<@coblee> iddo: exactly
<Luke-Jr> iddo: it is too weak today, yes
<Luke-Jr> bbiab
<@coblee> ok, later
<iddo> Luke-Jr: ok we can continue this discussion later:)
<OneMiner> Litecoin has the same options as bitcoin in the event of a disaster. Checkpoint, hardfork to new algo. There's more similarity than difference again.
<iddo> OneMiner: about your earlier question where you have scrypt ASIC miner, read this https://bitcointalk.org/index.php?topic=146191.msg1557247#msg1557247
<OneMiner> k
<iddo> OneMiner: if only you have the ASIC miner then it's trivial to hardfork litecoin so that your ASIC is useless and everyone except you will prefer the fork
<iddo> but if the majority of the miners have ASICs then it's less clear what will happen after the hardfork, what gmaxwell wrore there is relevant
<OneMiner> iddo I'm aware of that. What I wonder is how prepared we are for that. If I attack NOW (I won't) how long until we can start to hardfork?
<OneMiner> Because with my imaginary ASIC, I'd presume that the cost to attack would be trivial.
<OneMiner> ~couple hundred watts.
<iddo> it's easy to change some scrypt param and tell the pools to use the new version of the client
<iddo> anyway, scrypt ASIC is not in the cards
<OneMiner> Oh, I think it is.
<OneMiner> It's a matter of time as it was for bitcoin.
<iddo> OneMiner: you should also read the PoA thread, there the ASIC miners will be useless anyway
* redeeman ([email protected]) has joined #litecoin-dev
<OneMiner> I hate the fourm so much.... Could you link me please?
<iddo> OneMiner: did you see page 19 of the scrypt slides ?
<OneMiner> No. lol?
<iddo> page 19 explains why scrypt ASIC isn't in the cards
<OneMiner> I'm not even sure what you are refrencing. Scrypt slides?
* bernard75 (~bernard75@unaffiliated/bernard75) has joined #litecoin-dev
<iddo> OneMiner: about PoA, start either from the first post or from here: https://bitcointalk.org/index.php?topic=102355.msg1437768#msg1437768
<OneMiner> But you can replace ASIC with FPGA then. Assuming Luke is onto something.
<iddo> OneMiner: scrypt slides are at http://www.tarsnap.com/scrypt/scrypt-slides.pdf
<lupine> it's your average time-space tradeoff
<lupine> it's perfectly possible to do scrypt in low-memory, it's just more computationally expensive
<lupine> if your asics are ridiculously more powerful than everything else around, then maybe that's not an issue and you can get away without the ram. I don't know the numbers
* area (~area@unaffiliated/area) has joined #litecoin-dev
* pjorrit_ ([email protected]) has joined #litecoin-dev
<iddo> lupine: you need 128.5k of memory per hash attempt, if you have low memory then you cannot do many attempts in parallels
<OneMiner> I disagree in general that scrypt is ASIC proof at all. You simply make a mask and create chips. The complexity of the design is a non point because you just make the mask for it and boom, you're in buisness.
<lupine> the memory is to store intermediates that you can, in theory, generate from scratch each time. they are used lots of times to try to make that unattractive
<[yAK]> OneMiner: the reason why it is is because how easy it is to change the params of it
<[yAK]> asics cost a lot to initially create
<iddo> OneMiner: the point is that producing that ASIC will be a lot more expensive than SHA256 ASIC
<[yAK]> one small change makes them worthless
<[yAK]> iddo: is it the consensus of the litecoin development that asics type devices are unwanted?
<helo> that part is the same as sha asics, so it's not relevant
<iddo> lupine: no idea what you meant, each scrypt hash attempt must fill 128.5k of memory with pseudorandom sequence, you cannot get around that
<OneMiner> How much tolerance would we have for hardforks though? A single ASIC could probably be produced for under $100k (pulling that from my butt) and it could potentially dominate the mining.... landscape, whatever.
<OneMiner> You don't have to do a full run is what I'm saying.
<OneMiner> Some university student could potentially create one at very low cost.
<OneMiner> You just cram cache on the chip like a mofo. Problem solved for a custom design.
<iddo> [yAK]: initially scrypt was chosen because GPUs were unwanted, but the chosen scrypt params weren't intensive enough to exclude GPUs, anyway i'd say that the answer is yes, ASIC is unwanted, and isn't cost-efficent to produce anyway
* space_cadet (~space_cad@unaffiliated/space-cadet/x-5908614) has joined #litecoin-dev
<lupine> iddo, in the canonical implementation, you generate the sequence, and then take many subsets of that sequence by indexing to it
<[yAK]> what if in theory litecoin was worth what btc is worth right now? would you think that it would be cost effective then?
<lupine> if you remember the seed (4 bytes), then for each byte you need to index, you can regenerate the sequence from the PRNG, into a small buffer, for each access
<iddo> lupine: why? each hash attempt has completely different sequence?
<lupine> this is while trying to generate one hash
<lupine> I'm not saying that it is actually a good idea
<lupine> I don't have the numbers for that
<OneMiner> [yAK] For sure! Someone would gamble on producing a chip. There's the process, what's it called.... MOSIS? I'll look.
<lupine> if RAM really is as expensive as they say for ASICs, it may be the better option though
<OneMiner> http://www.mosis.com/pages/about/whatis
<lupine> http://en.wikipedia.org/wiki/Scrypt goes over it if you're unsure
<OneMiner> Not that I know what I'm talking about but I think you'd shoot for epic amounts of low latency cashe on the chip.