| 00:29:23 | Luke-Jr: | gmaxwell: guess we're not going after all :< |
| 00:29:38 | gmaxwell: | good since I have nowhere to put you all. :P |
| 00:32:43 | sipa: | there was lots of space when i was there! |
| 00:32:48 | sipa: | oh, that was on a holiday... |
| 00:35:46 | gmaxwell: | Mozilla office is in the process of moving (wrote mining at first…) right now. |
| 00:38:15 | nsh_: | nsh_ is now known as nsh |
| 00:40:42 | sipa: | gmaxwell: what is your primary source of income? mozilla or mining? |
| 00:42:45 | phantomcircuit: | sipa, i would venture 50/50 |
| 00:42:59 | sipa: | (it being mining would explain your almost-typo) |
| 00:42:59 | gmaxwell: | I'm not sure if mining is really a source of income anymore just because its returning coins I spent on hardware now. |
| 00:43:22 | phantomcircuit: | gmaxwell, you know whats fun |
| 00:43:30 | sipa: | clowns |
| 00:43:54 | phantomcircuit: | graphing return(days, daily growth rate) |
| 00:44:35 | phantomcircuit: | and then looking at the negative second derivative since knc |
| 00:44:56 | tacotime_: | it's less fun when you're a hashfast customer |
| 00:45:18 | sipa: | * sipa 's last mining hardware purchase was BFL |
| 00:45:26 | sipa: | i've learnt since then :) |
| 00:45:37 | kinlo: | heh |
| 00:45:42 | kinlo: | I bought an avalon |
| 00:45:51 | phantomcircuit: | tacotime_, interestingly the hashfast equipment actually looks fairly well designed |
| 00:45:55 | kinlo: | good way to throw away monniez |
| 00:46:02 | tacotime_: | i should have really known since i'd spent a lot of time chewing bfl out on the forum, but pulled the trigger when hashfast put up a sale. |
| 00:46:05 | phantomcircuit: | i mean except from using consumer grade ATX psus and the airflow being weird |
| 00:46:07 | gmaxwell: | kinlo: not if you got batch 1 avalons... |
| 00:46:18 | kinlo: | gmaxwell: batch 3, severly delayed |
| 00:46:18 | tacotime_: | phantomcircuit, it mostly works well. |
| 00:46:27 | kinlo: | and well overpaid |
| 00:46:39 | phantomcircuit: | tacotime_, i love how the exhaust is run over the boards before it exits the case |
| 00:46:49 | kinlo: | my roi isn't even 50% of the purchase price, not even including power costs |
| 00:47:24 | tacotime_: | phantomcircuit, the VRMs run pretty hot too, I cabled tied fans onto the front of the boards :P |
| 00:47:29 | kinlo: | so this weekend I'm going to go full throttle on the new shares calculator |
| 00:47:30 | tacotime_: | *cable tied |
| 00:47:35 | kinlo: | eh, wrong channel |
| 00:47:46 | tacotime_: | I really just wish bitcointalk would ban both bfl and hashfast |
| 00:48:15 | gmaxwell: | yea. I think the only profitable devices are Avalon B1 and Bitmain S1s maybe a smattering of bitfury that got good deals and timing.. and I've had reports from someone who got some of the very first KNC units that inist they mined back their coins, but it doesn't look like that for most peoplle. |
| 00:48:34 | gmaxwell: | tacotime_: banning people doesn't really help though. I don't think they're gaining any customers there. |
| 00:48:50 | tacotime_: | Yeah. |
| 00:48:57 | phantomcircuit: | gmaxwell, the very very first batch of knc units had a positive return assuming that you received a fully functional unit |
| 00:49:00 | tacotime_: | My October bitfury almost roi'd |
| 00:49:10 | phantomcircuit: | in reality the average for kncs first batch was about 250 Gh/s |
| 00:49:27 | phantomcircuit: | (220 if you include completely dead boards) |
| 00:49:45 | tacotime_: | Now all my asics are out to pasture on the peercoin network *tear* |
| 00:49:47 | gmaxwell: | I believe the two antminers I got (in dec and jan) have paid for themselves, maybe not including power costs yet. |
| 00:50:03 | phantomcircuit: | gmaxwell, the s1 or the usb thingies |
| 00:50:06 | gmaxwell: | S1s. |
| 00:50:24 | gmaxwell: | I have two S1s and a prototype half-S1 that bitmain gave to me for testing. |
| 00:51:36 | gmaxwell: | My CT box is never going to mine back its coins, even with the extra unit they're supposted to send me .. in March.. uh what month is it? |
| 00:51:45 | phantomcircuit: | gmaxwell, do you have them setup with ducting? |
| 00:52:08 | phantomcircuit: | iirc the bitmain things get ridiculous hot if you dont try to control the airflow |
| 00:52:46 | gmaxwell: | they're not that hot and I only have the 2.5 of them. they're just sitting on a shelf in a rack. also, it's cool here.. they actually run at medium fan speed. most of the time. |
| 00:53:16 | gmaxwell: | though have you seen their schemetics? they're running their voltage regulators at like 400% rated current... seem to be holding up okay though. |
| 00:53:39 | gmaxwell: | I think they were originally designed as 100GH/s devices and serverely over-performed. |
| 00:54:34 | tacotime_: | gmaxwell, at the texas conference they said that their "next gen" of S1s is just the same thing with every chip undervolted to hit <1 W/GH/s |
| 00:54:51 | tacotime_: | So maybe you're right and those are the original specifications |
| 00:55:00 | phantomcircuit: | gmaxwell, yeah the chips on those have weird behaviour |
| 00:55:42 | phantomcircuit: | they're not super sensitive to the hilarious amount of rf noise coming from the boards though |
| 00:55:44 | phantomcircuit: | which is neat |
| 00:56:14 | phantomcircuit: | apparently that's the lower bound on the hashfast equpiment |
| 00:56:28 | phantomcircuit: | below some clockrate the rf noise causes weirdness |
| 00:57:55 | just[dead]: | just[dead] is now known as justanotheruser |
| 01:00:03 | kaptah: | kaptah is now known as Guest50846 |
| 01:00:50 | gmaxwell: | so far the antminer stuff has been the best bitcoin hardware I've used. |
| 01:01:04 | gmaxwell: | but there is a lot of room to improve still. |
| 01:01:10 | tacotime_: | my bitfury stuff has been rock solid |
| 01:02:32 | phantomcircuit: | tacotime_, the chips are suspiciously identical |
| 01:02:52 | phantomcircuit: | bitfury has like 20-30 board designs too |
| 01:03:03 | phantomcircuit: | they could actually be the exact same thing with different names |
| 01:03:40 | tacotime_: | has anyone delidded and stripped them? |
| 01:04:03 | gmaxwell: | they have a different on the wire protocol. |
| 01:04:09 | gmaxwell: | so they're not the same chip. |
| 01:04:21 | gmaxwell: | though they are pretty similar in their overall performance. |
| 01:05:13 | phantomcircuit: | gmaxwell, isn't there a microcontroller on the antminer boards? |
| 01:05:26 | tacotime_: | unlike my hashfast miners my bitfury has trimpots to adjust voltage too, so i'll probably be running them longer. :P |
| 01:05:27 | phantomcircuit: | there isn't one on some of the bitfury boards |
| 01:05:50 | Luke-Jr: | tacotime_: software voltage control > trimpots |
| 01:05:57 | phantomcircuit: | was about to say |
| 01:06:07 | tacotime_: | Luke-Jr, ideally, but hashfast has neither so far |
| 01:06:16 | phantomcircuit: | tacotime_, they do it's just not exposed yet |
| 01:06:22 | tacotime_: | oh, thank god |
| 01:06:24 | phantomcircuit: | (or so i was told by a hf engineer) |
| 01:06:34 | phantomcircuit: | although maybe that's only for their v2 board design |
| 01:06:48 | phantomcircuit: | it looks like they completely redid the board to use easier to source components |
| 01:06:56 | tacotime_: | Well, I'll probably voltmod these soon anyway if I can figure out how to |
| 01:07:30 | Luke-Jr: | phantomcircuit: basically, yes |
| 01:07:30 | phantomcircuit: | inb4electricfire |
| 01:07:41 | tacotime_: | phantomcircuit, electric fires are part of it, man. |
| 01:09:23 | tacotime_: | mostly I just want to be able to crank the vcores down |
| 01:09:25 | gmaxwell: | phantomcircuit: I don't think they do, but I haven't really looked, would be funny if they were pirated bitfury masks. |
| 01:09:46 | phantomcircuit: | gmaxwell, im pretty sure they're bitfury chips |
| 01:09:58 | phantomcircuit: | those guys have all kinds of crazy reseller things |
| 01:10:11 | tacotime_: | yeah bitfury was selling reels like crazy |
| 01:10:20 | phantomcircuit: | hell they might have made a chip specifically for antminer that was slightly different to disguise it |
| 01:10:47 | tacotime_: | 55 nm with that performance is pretty impressive as well |
| 01:11:09 | phantomcircuit: | tacotime_, iirc he basically hand drew the entire mask |
| 01:11:26 | tacotime_: | yeah, it's a work of art |
| 01:11:56 | tacotime_: | http://zeptobars.ru/en/read/bitfury-bitcoin-mining-chip |
| 01:14:10 | phantomcircuit: | tacotime_, the power distribution network is hilarious |
| 01:16:36 | tacotime_: | but it works ;) the m board voltage distributions between h cards is really wacky too if i'm remember correctly |
| 01:45:22 | tacotime_: | http://www.blackcoin.co/ this should be amusing. looks like a peercoin fork but with pow mining turned off after block 10,000. |
| 01:54:36 | super3: | tacotime_, isn't mintcoin doing the same thing? |
| 01:54:43 | tacotime_: | I hope not |
| 01:55:12 | tacotime_: | iirc the stake burn through vulnerability described here: https://bitcointalk.org/index.php?topic=131901.0 |
| 01:55:20 | tacotime_: | has a fix that requires pow |
| 01:55:56 | tacotime_: | without pow |
| 01:55:58 | tacotime_: | ??? |
| 01:57:09 | tacotime_: | should be pretty amusing when they switch to pure pos i guess, heh. |
| 01:59:42 | antephialtic: | could SNARKs be used to make a coin with a "useful" PoW? e.g. Miner does some protein folding computation, then publishes a SNARK for the computation that is then verified by other miners? |
| 02:01:39 | sipa: | i guess, but it would be a horribly inefficient protein folding :) |
| 02:17:40 | gmaxwell: | antephialtic: maybe but thats not clear. |
| 02:17:53 | gmaxwell: | antephialtic: it may make the verification fast, but it doesn't prove there is no shortcutting. |
| 02:18:21 | gmaxwell: | (though perhaps constructing a shortcut is hard in practice) |
| 02:18:45 | gmaxwell: | and the enormous (e.g. 10,000x) slowdown in computation probably makes it uninteresting. |
| 02:19:19 | gmaxwell: | tacotime_: turned off? so does it have the "POS simulate a history were you found all the blocks" problem? |
| 02:19:56 | tacotime_: | gmaxwell: i'm guessing, it's a straight fork of novacoin. checking the github now |
| 02:22:00 | tacotime_: | bool ComputeNextStakeModifier(const CBlockIndex* pindexPrev, uint64& nStakeModifier, bool& fGeneratedStakeModifier){ ... } |
| 02:22:02 | tacotime_: | yuh huh |
| 02:22:56 | gmaxwell: | prev hm. sounds fun. |
| 02:25:20 | tacotime_: | >bitcointalk "http://coinmarketcap.com/ Wonderful bug makes BC more exposures !!!" |
| 02:25:21 | tacotime_: | sigh |
| 02:38:30 | tacotime_: | tacotime_ is now known as tt_zzz |
| 05:52:20 | asianbro: | asianbro is now known as shinybro |
| 06:46:36 | justanotheruser: | justanotheruser is now known as just[dead] |
| 07:51:09 | nsh__: | nsh__ is now known as nsh |
| 09:28:21 | gavinandresen_: | gavinandresen_ is now known as gavinandresen |
| 10:16:08 | [-krypto-]: | [-krypto-] has left #bitcoin-wizards |
| 11:11:32 | airbreather_1: | airbreather_1 is now known as airbreather |
| 13:28:19 | tt_zzz: | tt_zzz is now known as tacotime |
| 13:28:24 | tacotime: | tacotime is now known as tacotime_ |
| 13:38:33 | Guest23940: | Guest23940 is now known as Muis |
| 14:17:58 | rastapopuloto: | rastapopuloto has left #bitcoin-wizards |
| 15:11:16 | OneFixt_: | OneFixt_ is now known as OneFixt |
| 15:17:33 | 17SAATCLT: | 17SAATCLT is now known as HM2 |
| 15:23:19 | nikitab_: | nikitab_ is now known as nikitab |
| 15:44:56 | nikitab_: | nikitab_ is now known as nikitab |
| 16:00:26 | gmaxwell: | I'm very exicted about the ECDSA 2of2. |
| 16:02:09 | tacotime_: | What advantage does that have over 2of2 multisig? |
| 16:02:49 | shesek: | gmaxwell, ECDSA 2of2? |
| 16:02:58 | tacotime_: | I assume this |
| 16:02:58 | tacotime_: | https://bitcointalk.org/index.php?topic=511074.0 |
| 16:03:40 | gmaxwell: | tacotime_: transactions which are indistinguishable from ordinary payments, which increases privacy. They're also smaller and faster to verify. |
| 16:04:04 | gmaxwell: | (because they're indistinguishable from ordinary payments) |
| 16:04:15 | tacotime_: | did you lock the kimono gravity well thread? haha |
| 16:04:26 | gmaxwell: | tacotime_: I did. |
| 16:04:30 | tacotime_: | ah, okay, the latter i figured but not the former. |
| 16:04:38 | tacotime_: | yeah it was kind of silly |
| 16:05:27 | tacotime_: | it took me forever to figure out what the heck kgw was doing in the first place because the documentation made absolutely no sense |
| 16:05:35 | tacotime_: | but then looking at the code it was like oh |
| 16:05:37 | tacotime_: | okay |
| 16:06:38 | gmaxwell: | yea. it's trivial, and they've gooped it up with sciency sounding crap. |
| 16:21:36 | TD: | gmaxwell: is it really so new? i thought this paper did n-of-m ecdsa for a long time already: http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.67.9913&rep=rep1&type=pdf |
| 16:24:50 | Emcy: | "Even in the event that an attacker gains more than 50% of the network's computational power, only transactions sent by the attacker could be reversed or double-spent. The network would not be destroyed." |
| 16:24:59 | Emcy: | is that seriously a purchased advert? |
| 16:25:48 | jtimon: | mhmm I think I heard somewhere that some people were trying to do "an equivalent of multisig without changing anything on bitcoin" concurrently to the work on multisig a while ago |
| 16:26:12 | artifexd: | artifexd is now known as artifexd_afk |
| 16:26:36 | jtimon: | not sure if it's the same thing but sounds like it, it was in one episode of "the bitcoin show" (safebit maybe?) |
| 16:27:32 | jtimon: | trying to remember...maybe stefan thomas, what was his 2-men company in 2011 ? |
| 16:30:23 | Emcy: | so that ecdsa thing is better than normal multisig because the anon set is all of bitcoin instead? |
| 16:31:51 | jtimon: | I think it was in this chapter but, sorry I won't watch it again to confirm http://onlyonetv.com/2011/11/the-bitcoin-show-053-stefan-thomas-and-eli-sklar-introduce-safeb-it/ |
| 16:41:50 | TD: | Emcy: also smaller outputs so lower fees and smaller utxo set |
| 16:42:59 | gmaxwell: | TD: Hm! I wasn't aware of that. That would be fantastic. |
| 16:43:22 | TD: | i encountered it a long time ago (>1 year) but the maths was way over my head and there doesn't seem to be any open source implementation |
| 16:43:49 | Emcy: | has this 2 of 2 mothod been possible since the beginning? Just no one saw it before? |
| 16:43:59 | TD: | it's certainly more complex than the 2-of-2 protocol described in the doc (assuming it works). the latter looks like something even i could implement quite quickly :) though i am uncertain as to some of the assumptions it makes, like you can't reverse a modular inverse |
| 16:59:54 | maaku: | someone have a copy of that n-of-m DSA paper? |
| 17:00:08 | maaku: | citeseer has blocked this house for downloading too many papers... |
| 17:00:41 | zooko: | Ha. Really? |
| 17:01:09 | TD: | maaku: https://drive.google.com/file/d/0B4t9VJLm_PWheUtlck1YMGVJWEk/edit?usp=sharing |
| 17:01:18 | maaku: | thanks! |
| 17:01:40 | TD: | it apparently dates from 2003 |
| 17:02:28 | TD: | maaku: sorry, try now |
| 17:02:43 | maaku: | got it |
| 17:04:36 | zooko: | maaku: https://zooko.com/uri/URI%3ADIR2-MDMF-RO%3Afvmtmqxjwrid7di5lrt6es3sky%3A45iu4yzprgna3eod3zt46jgk2yq5vc4ltlxaa57ifpyzhft3fd7a/Latest/sec/Ibrahim-2003-%E2%80%9CA%20robust%20threshold%20elliptic%20curve%20digital%20signature%20providing%20a%20new%20verifiable%20secret%20sharing%20scheme%E2%80%9D.pdf |
| 17:04:39 | zooko: | Oh, you've got it. |
| 17:12:50 | Emcy: | is the p2sh m of n stuff still worth anything now then |
| 17:12:58 | Emcy: | such a big fuss about it 2 years ago |
| 17:14:07 | maaku: | as far as I can tell this N-of-M stuff requires interactive secret sharing, yes? |
| 17:15:17 | maaku: | Emcy: so no, they are not equivalent. with standard script multisig you can pay to any boolean combination of pubkeys without prior setup |
| 17:15:27 | maaku: | (there's two differences in that sentence) |
| 17:32:51 | gmaxwell: | maaku: yes, thats also the case for n-of-m schnorr. |
| 17:33:27 | gmaxwell: | Emcy: P2SH is also a lot more than just n-of-m. It allows compact specification of arbritary scripts. |
| 17:49:41 | justusranvier: | maaku: How specialized is the utxo set commitment data structure? Could it work for other projects with similar needs to Bitcoin? |
| 18:02:56 | maaku: | justusranvier: there's nothing specific to bitcoin |
| 18:03:37 | maaku: | it's a generic key/value index |
| 18:03:52 | maaku: | the keys and values are bitcoin-specific, but only in the way its being used |
| 18:04:29 | maaku: | it's documented here : https://github.com/maaku/bips/blob/master/drafts/auth-trie.mediawiki |
| 18:07:02 | justusranvier: | Thanks |
| 20:05:39 | artifexd_afk: | artifexd_afk is now known as artifexd |
| 21:46:50 | jgarzik: | hum, getting a miner from a new company gratis (well, in exchange for an honest review) |
| 21:47:14 | jgarzik: | * jgarzik ponders the level of suspicion and DD that should be generated... |
| 21:51:11 | zooko: | :-) |
| 21:52:21 | gmaxwell: | jgarzik: I have a seperate lan for that stuff. |
| 21:52:55 | gmaxwell: | had the same question about the antminer s1 prottype I was sent. "Hm.. mips host, from china, to plug into my network" |
| 22:09:57 | Ademan: | it'd be a great attack vector, and "respected community members" probably have a fair amount of BTC |
| 22:11:28 | Ademan: | "[ANN] New USB Miner, generates no heat, requires no configuration. Don't be alarmed that it looks like a regular USB thumb drive. Be sure to roll back windows updates prior to use." |
| 22:23:31 | sipa: | Ademan: is that a quote? |
| 22:24:35 | tacotime_: | I think it's from josh@bfl |
| 22:25:19 | sipa: | does it also cure cancer? |
| 22:26:08 | tacotime_: | X) |
| 22:42:22 | adam3us: | anyone have a contact for someone at bitstamp? |
| 22:45:24 | nsh_: | i barely have contact with reality |
| 22:49:06 | Ademan: | sipa: naw, bad joke, sorry |
| 22:49:28 | Ademan: | probably unintentionally based on a true story though |
| 23:25:29 | ssdt: | howdy! has anyone heard of such a thing as merging two blockchains? any ideas as to whether it even possible? |
| 23:31:48 | kanzure: | you could pre-mine according to the previous distributions but why would you want to? |