00:04:16sipa:HM: about what?
00:41:32zooko:Possibly of interest to you wizards: https://leastauthority.com/blog/BLAKE2-harder-better-faster-stronger-than-MD5.htlm
00:41:37zooko:Um I think I messed up that URL.
00:57:32Emcy:at the time of this writing (March 22, 2014), the Bitcoin network is performing enough computation to generate SHA-1 collisions every 131 minutes!
00:58:47sipa:that's ridiculous
00:58:59sipa:it does zero sha1 computations
00:59:26sipa:and most of its hardware is not even capable of doing any
00:59:35Emcy:i suppose it means if all those asics had sha1 engines on them instead
00:59:45zooko:sipa: I thought I stated it accurately.
01:00:13Luke-Jr:Emcy: they don't. there's no comparison
01:00:54zooko:There is a comparison! "computations"
01:01:01Emcy:that article makes sha3 sound pointless
01:01:03zooko:A very broadly defined unit of computation that cryptographers use.
01:01:08Luke-Jr:zooko: being un-like SHA-2, is still a good idea for SHA-3
01:01:46sipa:zooko: not commenting on the article (which i haven't read); just on the quote
01:01:49Luke-Jr:it allows one to do *both* hashes and be more reasonably sure they won't both break
01:02:08zooko:sipa: oh.
01:02:23zooko:Luke-Jr: yeah, I like that sort of idea.
01:05:17zooko:Luke-Jr: I'm afraid my article is a bit ambiguous about something, which is BLAKE being "like" SHA-2.
01:05:49zooko:It is actually unlike SHA-2 in many/most ways, but it is like in that the core operations are adds, xors, and rotates.
01:19:47mike4:mike4 is now known as c--O-O
01:53:03HM:sorry, sipa: still there?
01:53:16sipa:no, sorry
01:53:25sipa:i'm a few thousand kilometers further now
01:53:32HM:I'll catch you when you're back then
01:53:57HM:I just wanted to ask quickly whether you looked at LLVM IR when you were implementing your secp256k1 primitives
01:54:12sipa:i know nothing about llvm
01:54:16HM:I've been playing with it and it has some nice vectorised intrinsics
01:55:22HM:it's about as complex as asm and can be compiled to asm or directly to an object file for a dozen platforms
02:08:57Luke-Jr:HM: but only with LLVM?
02:14:49HM:Luke-Jr, sure
02:14:59HM:but .asm is only for x86 and one assembler :P
02:15:09HM:*x86 .asm
02:15:33sipa:the asm for libsecp256k1 is x86_64 only now
02:17:01HM:lower half of the page has some trivial examples
02:17:17HM:not sure how well it'd cope with something more complex like your group and field element ops
02:17:40sipa:well feel free to convert the 5x52 multiply code to that and compare :)
02:18:12sipa:all the asm is used for is multiplications
02:18:34HM:I'll add it to my list of broken promises.
02:19:25HM:familiarity with the llvm toolchain has been on my todo list for ages
02:30:44rdponticelli_:rdponticelli_ has left #bitcoin-wizards
02:48:49tromp:hi zooko
02:51:50tromp:does that 131 mins take into account that bitcoin uses SHA256^2 rather than plain SHA256?
02:54:50gmaxwell:it's really not compariable in any case.
02:55:08gmaxwell:as hardware for sha1 can easily be much faster than the same die area and process for sha1.
02:55:18gmaxwell:because sha1's pipeline is much tidyer.
02:58:59Emcy:maybe he meant per unit energy with current tech
02:59:04Emcy:you could estimate it
03:04:29gmaxwell:well I don't know where that number came from since log2(2^32*4250217919/10*131.) = 67.696 ... I actually think there is collission attack theorized against sha-1 with closer to 2^60 work.
03:05:04gmaxwell:(I think the earlier attack was 2^69, which is roughly what you get if you add an extra bit (e.g. counting bitcoin as 2x sha256) to the above.
03:13:30austinhill:Luke-Jr: Sorry to have missed you in San Mateo, email me @ austin@isin.to when you have a chance so I can settle up with you on flight & circle back on discussions - hope you had fun with the gang
03:14:03Luke-Jr:austinhill: yeah, what was with that⁈ :P
03:14:37austinhill:Had an emergency with a friend that had to take precedence & had to fly to Denver - much regretted, but unavoidable
03:15:59austinhill:All the more reason to organize another bitcoin house :) Maybe Toronto around the Expo time?
03:16:27austinhill:maaku I think is in town for that one….
03:37:54just[dead]:just[dead] is now known as justanotheruser
04:53:14justanotheruser:justanotheruser is now known as just[dead]
04:56:23just[dead]:just[dead] is now known as justanotheruser
04:59:47Guest38922:Guest38922 is now known as ageis
05:06:34justanotheruser:justanotheruser is now known as just[dead]
05:07:02just[dead]:just[dead] is now known as justanotheruser
05:46:57c0rw|afk:c0rw|afk is now known as c0rw1n
06:10:30Luke-Jr:sigh, Counterparty is getting ridiculous, denying exact quotes
06:11:11justanotheruser:Luke-Jr: do you dislike anything about counterparty other than them putting non tx data in the blockchain?
06:11:33Luke-Jr:justanotheruser: afaik it's all tx data anyway; the problem is that they're doing it wrong (abusing multisig)
06:12:05justanotheruser:Luke-Jr: well using tx data is even worse, right? They could be using OP_RETURN
06:12:26Luke-Jr:justanotheruser: eh, I think you're using different terminology
06:12:38Luke-Jr:by "tx data", I mean "represents a financial transaction"
06:12:55Luke-Jr:so, using OP_RETURN would still be tx data
06:13:03Luke-Jr:and is the obvious short-term path for them
06:13:30justanotheruser:Luke-Jr: well are they adding to the UTXO?
06:13:39Luke-Jr:right now they are
06:14:20justanotheruser:yeah, everything is a tx, I meant bitcoin value transferring tx
06:15:04Luke-Jr:they want to extend Bitcoin to do asset tracking. that's fine, but they're doing it wrong.
06:15:15Luke-Jr:the correct way to do it is what maaku and jtimon are doing
06:15:25Luke-Jr:even for a short-term "we want it now!"
06:15:32Luke-Jr:even for a short-term "we want it now!" solution, they should be using OP_RETURN
06:15:47justanotheruser:Luke-Jr: you're referring to freicoin?
06:15:53Luke-Jr:Freimarkets specifically
06:16:20justanotheruser:Is that in the freicoin whitepaper?
06:16:27Luke-Jr:no idea
06:16:41justanotheruser:Is it implemented?
06:17:08Luke-Jr:probably has months, if not years, of development ahead
06:17:20justanotheruser:I see
06:17:37justanotheruser:I've got to read their whitepaper. It's just sitting in a tab in my browser
06:17:47Luke-Jr:that's why their "we want it now!" stuff is semi-understandable
06:26:51nOgAnOo:May the Lord Jesus Christ bless you all.
06:28:07Luke-Jr:nOgAnOo: good morning
06:36:17nOgAnOo:I was thinking of selling the blockchain on USB.. my friend has a bad connection with 3 Bitcoins on an old wallet and cannot access blockchain
06:36:57Luke-Jr:nOgAnOo: off-topic
06:37:11nOgAnOo:I hear that every time I speak here
06:37:21Luke-Jr:perhaps this channel is not a good match for you
06:37:25nOgAnOo:But when I speak of future ideas i am mocked
06:37:55nOgAnOo:The cryptocoin community is snobby, scammy and evil
06:38:07nOgAnOo:Jgarzik, Gmaxwell
06:38:17nOgAnOo:All of the userbase..
06:38:24nOgAnOo:Mean, cruel, rude.
06:38:36nOgAnOo:Do you guys believe in God?
06:38:43nOgAnOo:Or that you will be judged when you die?
06:38:51nOgAnOo:Do you plan to take bitcoins to the afterlife with you?
06:38:57nOgAnOo:You can not buy your way out of hell.
06:39:15nOgAnOo:You will have nothing down there.. just some rotton demons torturing you eternally.
06:39:34nOgAnOo:For failing to accept Christ as Savior and a sacrifice for your sins..
06:40:25justanotheruser:justanotheruser is now known as just[dead]
06:40:52nOgAnOo:Where will you go when you die?
06:43:12Luke-Jr:Luke-Jr has kicked nOgAnOo from #bitcoin-wizards
06:55:15jcorgan:i think he regularly gets kicked from #bitcoin, too
07:00:36Luke-Jr:jcorgan: it's frustrating because he's not wrong and it's important, just it's off-topic :/
08:39:04Luke-Jr:hum, interesting how merely phrasing the same thing differently can get a better reaction
10:08:29maaku:maaku is now known as Guest21459
10:52:45gmaxwell_:gmaxwell_ is now known as Guest10382
10:57:52bobke_:bobke_ is now known as bobke
11:19:38fanquake:fanquake has left #bitcoin-wizards
11:29:51Aesthetic:Aesthetic is now known as Logicwax
11:32:47pigeons:pigeons is now known as Guest28482
11:46:40Graet:Graet is now known as Guest59244
11:53:19Guest59244:Guest59244 is now known as Graet
12:46:04Guest10382:Guest10382 is now known as gmaxwell
12:48:57rdymac_:rdymac_ is now known as rdymac
15:47:38zooko`:zooko` is now known as zooko
16:21:00Guest21459:Guest21459 is now known as maaku
17:54:40wallet42:wallet42 is now known as Guest76110
17:54:40wallet421:wallet421 is now known as wallet42
18:06:20OneFixt_:OneFixt_ is now known as OneFixt
18:13:23Guest28482:Guest28482 is now known as pigeons
20:05:37gmaxwell:Quiet lately!
20:15:50tacotime:I've been mostly working on technical specifications for my PoS stuff and coding, so I haven't had a lot of time for theory. Such is implementation, I guess.
20:17:21tacotime:I'll be at the Toronto expo checking out everyone's talks in a couple of weeks, though.
21:01:16zooko:TOO quiet
21:11:28andytoshi:i've been working [reading] with a potential supervisor since last week, and before that rosemary was in town. no time for bitcoin work :(
21:25:38pigeons:Can someone point me somewhere to help understand the point-of-view of people building applications like counterparty/mastercoin/etc on top of the bitcoin blockchain who feel very strongly that the blockchain itself is the best place to store actual data instead of a hash of the data?
21:26:02pigeons:I'm trying to see what are the potential advantages they think this gives. Non-reliance on external dependencies I suppose, or some perceived or real censorship resistance and data persistance?
21:26:49nsh:you forgot "magical thinking" :)
21:27:18gmaxwell:pigeons: I saw some mastercoin stuff that was some crazy vbscript stuff.
21:28:16pigeons:i'm seriously trying to understand that perspective
21:28:19gmaxwell:So it seems to me that part of it is simply escaping actually building any low level infrastructure, and instead externalizing those costs on other people.
21:28:20tacotime:pigeons, it saves them from having to implement a proper P2P protocol to retrieve the data the metadata itself refers to for one.
21:28:40pigeons:yeah true
21:28:56tacotime:At the expense of the Bitcoin blockchain. :P
21:29:28gmaxwell:and perhaps I'm being a snob but I have a strong impression that anyone who'd consider writing vbscript probably doesn't currently have the required intellectual toolset to actually do systems work.
21:30:19gmaxwell:(and I don't mean to imply these people are less good; bitcoin needs a lot of higher layer applications development, sadly these efforts aren't really efforts for bitcoin but generally competative parasitic systems.)
21:31:07tacotime:Yeah. It seems weird to me that they just didn't release it on a separate fork, or spam the namecoin chain that no one uses anyway except onename.
21:32:19HM:gmaxwell, i guess you haven't heard about NodeVB then
21:32:28HM:VB is making a revival
21:33:39gmaxwell:while I'm well aware of the universality of computation, I've never met a competent systems person who wanted anything to do with a toolset like that, and perhaps this is a bit of snobbery on my part, but I'm skeptical...
21:41:15Luke-Jr:gmaxwell: perhaps competent systems people just move past it quickly? :p
21:44:02tacotime:In some ways I think MasterCoin may have been the "Producers" plot gmaxwell jokingly mentioned, wherein the MSC foundation sells their premine near launch for massively inflated prices and then MSC itself is doomed to obsolescence by fees competition in the Bitcoin main chain. Then the MSC people walk away with their moneybags shrugging.
21:44:14tacotime:But that's my conspiracy theory, and is someone OT anyway. :P
21:57:29gmaxwell:I wish the world were as simple as theater plots.
22:25:31wallet42:wallet42 is now known as Guest89320
22:25:31wallet421:wallet421 is now known as wallet42