Transcript for #bitcoin-dev 2011/04/26

00:06 gribble {"ticker":{"high":1.701,"low":1.5211,"vol":15075,"buy":1.5564,"sell":1.595,"last":1.595}}
00:06 JFK911 ;;bc,mtgox
00:06 JFK911 ;;bc,stats
00:06 gribble Current Blocks: 120210 | Current Difficulty: 92347.59095209 | Next Difficulty At Block: 120959 | Next Difficulty In: 749 blocks | Next Difficulty In About: 4 days, 12 hours, 23 minutes, and 49 seconds | Next Difficulty Estimate: 104784.37005952
00:09 CIA-89 bitcoin: Luke Dashjr * rfbad645b4bd7 gentoo/dev-libs/jansson/ (.git-info Manifest jansson-1.3.ebuild): Import dev-libs/jansson-1.3 from "nikai" overlay http://tinyurl.com/5spvsz3
00:10 luke-jr ;;later tell jgarzik for some reason, pushpool only links if -lmysqlclient is listed on the command line AFTER db-mysql.o
00:10 gribble The operation succeeded.
00:14 EPiSKiNG ;;bc,calc 1036000
00:14 gribble The average time to generate a block at 1036000 Khps, given current difficulty of 92347.59095209 , is 4 days, 10 hours, 20 minutes, and 47 seconds
00:14 EPiSKiNG ;;bc,est 1036000
00:14 gribble Error: "bc,est" is not a valid command.
00:14 EPiSKiNG ;;bc,estimate 1036000
00:14 gribble 104784.37005952
00:14 EPiSKiNG ;;bc,gen 1036000
00:14 gribble The expected generation output, at 1036000 Khps, given current difficulty of 92347.59095209 , is 11.2838699044 BTC per day and 0.470161246016 BTC per hour.
00:15 EPiSKiNG ;;bc,gend 1036000 104784.37005952
00:15 gribble The expected generation output, at 1036000 Khps, given the supplied difficulty of 104784.37005952, is 9.94459576069 BTC per day and 0.414358156696 BTC per hour.
00:16 programe ;;bc,gen 750000
00:16 gribble The expected generation output, at 750000 Khps, given current difficulty of 92347.59095209 , is 8.16882473773 BTC per day and 0.340367697405 BTC per hour.
00:17 programe ;;bc,gen 120000
00:17 gribble The expected generation output, at 120000 Khps, given current difficulty of 92347.59095209 , is 1.30701195804 BTC per day and 0.0544588315849 BTC per hour.
00:19 CIA-89 bitcoin: Luke Dashjr * r8a79b9003b5f gentoo/net-p2p/pushpool/ (Manifest pushpool-9999.ebuild): net-p2p/pushpool: Initial import of live-git http://tinyurl.com/5vp4z9e
00:29 CIA-89 bitcoin: Luke Dashjr * rede72535995f gentoo/net-p2p/pushpool/ (Manifest pushpool-0.3.ebuild): net-p2p/pushpool: 0.3 http://tinyurl.com/4y9lv53
01:01 programe CPU :: Ufasoft's SSE2 miner would work with pushpool?
01:22 programe im getting upstream RPC error
01:22 programe does anyone know what it means?
01:25 B0g4r7 It means about what you'd think.
01:28 programe B0g4r7 what?
01:28 EPiSKiNG anyone else notice an issue with teh new Phoenix miner not getting the correct MH/s?
01:29 EPiSKiNG I ran mine for a like an hour, and the MH/s show better than poclbm, but when I check deepbit's reports, it shows a substantial decrease
01:30 fabianhjr I have been searching for a while and I am 95% sure there was a Pledge for a MyBitcoin FOSS alternative. :/
01:31 dust1 EPiSKiNG: sample size?
01:34 Kiba so, what happens to your magainze, fabianhjr
01:35 fabianhjr Not much, that really happened. After we parted blogs poped up and there was no writing taking place. So now I am back to coding, graphics, and security.
01:35 Kiba don't you have this thing called the bitcoin times when I briefly help run it
01:35 Kiba so you gave up
01:35 fabianhjr Kiba: yes, I still have the files if you want them.
01:36 Kiba nothing interesting I already extracted, plus I do only public domain
01:36 fabianhjr Kiba: I moved to another marketplace. I made my profits in the journalism area.
01:36 Kiba where do you make your bitcoin now?
01:36 fabianhjr Kiba: yes, I know. Though, I offered because maybe you would want a copy with the other articles.
01:36 EPiSKiNG dust1: phoenix -u http://user:[email protected]:8332/;askrate=15 -k poclbm DEVICE=2 VECTORS AGGRESSION=10 -v FASTLOOP BFI_INT
01:37 fabianhjr Kiba: doing a Bitcoin class in PHP then moving to do a MyBitcoin alternative, FOSS.
01:37 dust1 EPiSKing: I mean what sample size are you basing your conclusion on.
01:37 Kiba persistance*
01:37 fabianhjr Kiba: you aren't profiting from your Weekly. You still haven't won the fight. :P
01:38 Kiba but I got more traffics
01:38 fabianhjr Have you surpassed your 1K Bitcoin wish? I remember 3-5 months ago you were bragging about getting your hand to all the bitcoins you could.
01:38 fabianhjr :D
01:38 EPiSKiNG 1 hr
01:38 Kiba fabianhjr: not really
01:39 dust1 Episking: probably not signifigant
01:39 Kiba but my revenues better now
01:39 Kiba .11 BTC a day
01:39 fabianhjr LOL, ok, that was the biggest impression I got from you when I met you.
01:39 fabianhjr s/met/meet/
01:39 Kiba fabianhjr: anyway, I have 445 BTC
01:39 Kiba and I still do scrap for bitcoin
01:40 fabianhjr LOL, at least you didn't follow genjix path of saling himself.
01:42 Kiba apperantly, I made a profit of 320% at witcoin
01:43 fabianhjr Kiba: congratz! You deserve it for your journalism!
01:44 hello_ Hey guys, I have a quick question.
01:44 programe which year it will be 1 block = 25 btc ? right now its 1 block = 50 btc.... when it will happen to be 1 block = 25 btc?
01:44 B0g4r7 At block 210000.
01:44 shazow programe: in 2 years i believe
01:44 programe shazow: 2014 ?
01:45 hello_ I made 5 successive payments of .2 btn to a gambling website about 40 minutes ago but they are all still unconfirmed :(
01:45 hello_ Does anyone know why?
01:45 shazow programe: i think so (don't quite one m on that)
01:45 hello_ All my other transactions are getting confirmations.
01:45 shazow programe: 2013 i think
01:45 programe shazow: and when that happens it will be required to have the double processing power to earn 50 btc in the same amount of time that we do now?
01:45 B0g4r7 Maybe you're unlucky and the block producers decided not to include your transactions.
01:45 B0g4r7 Wait more.
01:46 B0g4r7 Or include a tx fee.
01:46 shazow programe: no, since the rate is scaled by the processing power, if everyone doubles it then it will remain the same
01:46 B0g4r7 Roughly scaled.
01:46 shazow programe: you'll require double relative to everyone else's staying the same
01:46 hello_ B0g4r7, how often does that happen? And about how long should I wait?
01:46 B0g4r7 A block is generated about every 10 minutes.
01:46 programe shazow: so basically if my 6990 card now does 50btc each 6 days, how much it will do in year 2014 or so ?
01:47 programe 25 btc?
01:47 B0g4r7 Every time that happens, the block producer has an opportunity to add your transactions to the chain.
01:47 shazow programe: that all depends on how many people are mining, if you assume the same compute power exists in the total mining pool, then you'll get 25 btc in 6 days
01:47 B0g4r7 Depends on difficulty.
01:47 programe i see
01:48 hello_ I've been waiting for 40 minutes, so it seems like I missed out on 4 opportunities to have my transactions added to the chain :(
01:48 B0g4r7 Did you include a transaction fee?
01:48 B0g4r7 Also make sure you're connected to peers and see the block count advancing.
01:50 hello_ 34 connections. Did not include a fee as I didn't see an option for it when I sent payments to PROBIWON through the default Bitcoin client.
01:50 hello_ Actually the block count isn't advancing, weird...
01:50 hello_ Although I have 35 connections.
01:50 retinal ;;bc,blocks
01:50 gribble 120227
01:51 hello_ Yep, 120227
01:51 retinal are you on the current one?
01:51 hello_ Yep
01:51 retinal it's a waiting game, then
01:52 hello_ 120228! :P
01:52 retinal ;;bc,blocks
01:52 gribble 120228
01:52 retinal :o
01:52 hello_ :o
01:53 B0g4r7 http://blockexplorer.com/b/120228
01:53 B0g4r7 You can see if you see your transaction in the block.
01:53 B0g4r7 I see a couple payments of 2 btc.
01:54 hello_ Confirmed ;) Wow that took forever.
01:55 B0g4r7 Now you get to wait some more, for 6 confirmations, assuming that's the requirement.
01:56 hello_ Just one confirmation :)
01:56 hello_ So how come the block count was going up so rapidly today?
01:56 B0g4r7 1 confirmation per block.
01:56 hello_ And now it's slower.
01:56 B0g4r7 Maybe someone switched off some miners.
01:56 hello_ Please tell them to turn it back on <3
01:57 B0g4r7 Or if you just started your client and it was behind it had to catch up.
01:57 hello_ I think IBM should get in on this mining.
01:57 hello_ Get those blocks up faster.
01:57 B0g4r7 The network will readjust no matter how muchy compute power is brought to bear.
01:57 B0g4r7 To achieve a constant rate of 1 block per 10 minutes (or so)
01:57 hello_ So the number of blocks always increases at the same rate?
01:57 hello_ Ah, I see.
01:57 B0g4r7 However...
01:57 retinal but ... then the difficulty will adjust accordingly to average one block every ten minu... what B0g4r7 said
01:58 B0g4r7 Difficulty readjustent only occurs once per 2016 blocks.
01:58 B0g4r7 So you could gain a temporary boost.
01:58 hello_ Yeah.
01:59 B0g4r7 If someone wanted to hose the network, they could bring in massive power, wait for the readjustment, and then drop off.
01:59 B0g4r7 Drop the block rate to one per 120 minutes or something.
02:00 B0g4r7 Then it would be forever until the next readjustment.
02:00 B0g4r7 Then they could pull the same thing again.
02:00 B0g4r7 Even worse, bring in massive power, but encode no transactions in any blocks solved.
02:00 hello_ Jeez.
02:00 hello_ What safeproofs are there against that?
02:01 B0g4r7 Lots of grassroots power...?
02:01 programe YEAAAAAAAAAAAAAAAAAAAAAAAAAAH
02:01 programe I GOT PUSHPOOL WORKING
02:01 programe HAHA
02:01 programe IT TAKED ME 4 DAYS
02:01 programe full working days
02:01 B0g4r7 I don't know what that is, but right on.
02:01 hello_ Pushpool?
02:02 programe https://github.com/jgarzik/pushpool
02:02 programe yeah its a polling software
02:02 programe to create pools
02:02 programe polls
02:03 programe mining pools
02:03 B0g4r7 Push notification of a new block?
02:03 programe yes, and share earnings between miners etc
02:40 programe jgarzik
02:41 Kiba I argued with a bitcoin user on twitter who argued that the exchanges are already decentralized
02:41 jrabbit lol.
02:42 Kiba mtgox got a big fat bank account painted on it
02:42 Kiba tell me that's not centralized
02:42 Kiba it's a big fat choke point
02:43 programe jgarzik: pushpool isnt registering the shares for some reason, do you have idea why? i need to wait until the miners solve one whole block for anything to be registered?
02:53 programe weird
02:53 Kiba jgarzik: so, will pastecoin get back?
02:55 programe jgarzik isnt answering :/
02:56 programe weird
02:56 programe when he will be back
02:56 ersi lol
02:56 programe weird
02:56 programe i didnt know
02:57 jgarzik learn some IRC etiquette, people
02:57 programe sorry
02:57 programe jgarzik: can you help me to answer some questions regarding pushpool?
02:58 jgarzik as long as it's not fscking beep beep beep with each message. damn.
02:58 jgarzik highlight I can deal with.
02:58 ersi programe: How about taking a hint
02:58 programe jgarzik: one thing to note is that blkmond crashes every 10 min or so
02:59 programe jgarzik: probably because my bitcoin server client is newer than the protocol it supports?
03:00 programe jgarzik: minerd works okay with the pushpool setup i have made using json-http interface however CPU :: Ufasoft's SSE2 miner doesnt work ?
03:00 programe jgarzik: i see minerd to hash and show MHash/s etc however i dont see any shares beign recordered on the db
03:00 programe jgarzik: those are all the questions :)
03:01 ersi (??????)
03:03 programe jgarzik ?
03:03 ersi I hope he did put you on ignore, 'cause you sure are annoying
03:04 programe i tought this channel is to discuss dev stuff
03:04 programe im joining his project to solve bugs
03:04 programe and review the code
03:04 programe and implement new features
03:04 programe and he is not going to help me to join
03:04 programe ?
03:04 ersi then go and review the code
03:04 programe i tought open source developers were more open
03:04 programe and friendly
03:04 davex__ hmm... wonder how phoenix miner could possibly run the gpu cooler but get better mh/s than other miners
03:05 jgarzik open source means you are free to read the code and figure out problems youself.
03:05 ersi Well, you can't just expect to get support just cause something is open.
03:05 jgarzik Free support is ----> that way
03:05 programe jgarzik: i can pay
03:05 programe jgarzik: if thats what you want
03:06 ersi Oh man, exactly what I'd wish for when on vacation! Work! :)
03:06 jgarzik yeah really
03:06 programe however my idea was to join and help on the project and work togheter in team
03:06 programe i just have those questions and nothing more
03:07 jgarzik might have to cut vacation short to deal with daughter's croup, even :/ We'll see in the morning.
03:07 nanotube programe: it seems jg has other things on his mind atm... so just save your inquiries for a bit later.
03:07 programe okay
03:07 ersi jgarzik: Croup? :o
03:07 programe no problem
03:07 programe ill research the code myself
03:08 programe however im a bit dissapointed
03:08 ersi jgarzik: Ouch, that does not seem nice. Hope she'll get better man
03:08 nanotube jgarzik: haha i was going to ask if your daughter is a horse... then used the dict to figure out it has alternate meaning :)
03:08 nanotube best wishes on your daughter's health.
03:08 programe jgarzik: best whishes
03:09 programe jgarzik: hopefully we can work togheter in a future
03:09 programe meanwhile ill have to review the code and anser the questions myself :P
03:09 programe *answer
03:09 Androgynous hey...
03:09 Androgynous can anyone here maybe possibly help me with my bitcoin.conf for namecoin?
03:10 Androgynous i'm getting an error saying that i need to set my rpcpassword
03:10 Androgynous but i made the file
03:10 programe whats namecoin
03:10 programe i see distributed name system..
03:10 Androgynous chain seperate from bitcoin
03:11 Androgynous yeah, it involved domains and names
03:11 Androgynous but to run namecoind
03:11 Androgynous i need a bitcoin.conf
03:11 programe Diablo-D3: does diablo miner supports http-json requests?
03:11 Androgynous C:UsersUserAppDataRoamingNamecoin
03:11 Androgynous i have it saves there
03:12 Androgynous where it tells me to save it
03:12 Androgynous as owner-readable-only
03:13 Androgynous perhaps i'm going about setting it as "owner-readable-only"
03:14 Androgynous wrong
03:14 Androgynous ping me if anyone thinks they can help
04:02 bk128 anyone know the name of the site that shows a moving chart of all the transactions by size?
04:18 midnightmagic bitcoin watch?
04:19 midnightmagic nope..
04:20 Diablo-D3 [01:11:45] <programe> Diablo-D3: does diablo miner supports http-json requests?
04:20 Diablo-D3 how else would it work?
04:25 luke-jr does anyone know DataSurfer? is he on IRC?
04:25 luke-jr Diablo-D3: all the other pools have or are moving to push-based mining
04:28 Diablo-D3 luke-jr: yes, which is kind of nuts when they're all going to have to convert to mine.
04:53 chmod755 http://random.witcoin.com/p/1302/OperationTop10
05:31 genjix tcatm: britcoin has been up and functioning but disappeared from bitcoinwatch since yesterday
05:32 genjix appears on bitcoincharts though
05:45 midnightmagic bitcoinmonitor.com is the one..
05:46 genjix bitcoinmonitor isnt for currencies
05:52 ersi Holy fuck, people trading 1 BTC for 1.5 USD o_o
05:56 midnightmagic ;;bc,stats
05:56 gribble Current Blocks: 120263 | Current Difficulty: 92347.59095209 | Next Difficulty At Block: 120959 | Next Difficulty In: 696 blocks | Next Difficulty In About: 4 days, 1 hour, 3 minutes, and 12 seconds | Next Difficulty Estimate: 105916.19630086
05:56 midnightmagic bk128 was asking for moving chart of transactions by size.
05:57 gribble (105,916.19630086 - 92,347.59095209) / 92,347.59095209 = 0.146929716
05:57 midnightmagic ;;calc (105916.19630086-92347.59095209)/92347.59095209
05:57 cosurgi ArtForz: what is the endianness of time in "DATA : 000002800000000000000000000000000000000000000000000000000000000000000000000000000000000080000000_nonce__~target~==time==_merkle_" ?
05:57 midnightmagic 14.7% increase.
05:58 midnightmagic that's.. a bit high.
05:58 cosurgi ArtForz: I am looking at 'time' from different blocks and only the high bits are changing, not low bits.
05:59 lfm cosurgi: you mean like the "data" in getwork?
06:01 lfm cosurgi: the "data" in getwork is 32 bit big endian
06:02 lfm they byteswap it as an optimization for the sha256 routine
06:02 Compgenius hey guys, how would I clear my transaction list in bitcoin?
06:03 cosurgi lfm: yep, I just reversed the bytes, and I got a correct time.
06:03 Compgenius atm my bitcoin "all transactions" is flooded with my bitcoin winnings and sendings
06:04 CFSworks Compgenius: Create a new wallet, send all your BTC there, delete the old one?
06:04 lfm Compgenius: you cant
06:04 ersi Modify the client :)
06:05 CFSworks Moving all your Bitcoins into a new wallet would work... You lose your old addresses though. :(
06:05 Compgenius not necessarily a good idea to lose all my address...
06:06 Compgenius I'll just leave them for now
06:06 ersi So keep your old wallet around
06:07 Compgenius just to show you... http://cg999.ath.cx/imgs/lolcoin.png damn thats going to be annoying...
06:08 retinal pshaw, that's not even 100 transactions
06:09 CFSworks I have 113 and it's not so bad as long as I don't stare at the scrollbar.
06:09 ersi That's why one usually has a 'internal balance' on sites and then either deposit or withdraw
06:09 CFSworks It's like fear of heights: Don't look at how bad it is and you won't get sick to your stomach.
06:10 Compgenius counted em, 63.
06:10 CFSworks That's odd. The lower-right corner says 67.
06:10 Compgenius meh, might have skipped a few numbers
06:11 CFSworks With so many similar transactions I'd get lose count too.
06:11 Compgenius sometimes i accidently skip from say 45 to 50 >_>
06:11 Compgenius btw CFSworks, those transactions were all done within the space of a day
06:12 CFSworks Wow. Are all of them small trades too?
06:12 Compgenius yeah
06:12 Compgenius 0.01-0.02
06:13 Compgenius there's only one transaction that wasn't on the 25th
06:13 CFSworks I can see what you mean. That would give me a headache.
06:13 ersi sounds like a lottery or some other small-change game
06:13 Compgenius and that was on the 14th, when i got my 0.05 from the faucet
06:13 Compgenius CFSworks, bitcoin darts
06:13 Compgenius really needs some sort-of on site wallet
06:13 Compgenius >_>
06:13 CFSworks Hmm...
06:13 ersi I agree :)
06:13 Compgenius new address to send to every time you bet
06:14 CFSworks Hmm...
06:14 ersi Seems like you've done +- 0 :)
06:14 Compgenius ersi, it's more or less -0.10
06:15 CFSworks What Bitcoin needs is a "hide old transactions" option that hides anything older than X transactions.
06:15 Compgenius since i got 0.10 off of a friend to bring it to 0.15
06:17 CFSworks To any core Bitcoin developers online right now: What do you think about a "Show only last [ ] transactions in the main window." textbox to be added to the GUI preferences?
06:18 Compgenius CFSworks, i'm surprised there's not a remove transaction option
06:18 Compgenius to get rid of a transaction, or atleast a deleted transaction tab
06:18 Compgenius >_>
06:19 CFSworks "Removing" a transaction isn't in the design of Bitcoin, since it's stored permanently in the block chain.
06:19 CFSworks "Hide" would be a little less misleading.
06:21 retinal request: filter by address
06:22 retinal actually, scratch that; it'd be impractical after a couple months of intense usage
06:22 retinal (??????)
06:22 CFSworks Compgenius, did you develop the darts game?
06:28 Compgenius CFSworks, nope
06:29 Compgenius retinal, it really would be useless, since a lot of sites generate a random address for every transaction you make to them
06:30 ersi filter by label, then.
06:33 Compgenius actually yeah.. that makes more sense ersi
06:40 genjix thing i dont get is why so many boost::asio examples use threads when they suck and a reactor pattern means you don't have to use them.
06:52 doublec probably to take advantage of multiple cores
07:20 RenaKunisaki https://verydemotivational.files.wordpress.com/2011/02/58349704-6eab-4b4f-bf4e-358a7235b75c.jpg there's a Bitcoin joke in here somewhere...
07:31 sipa ;;bc,blocks
07:31 gribble 120278
07:31 gribble 0.729372504234
07:31 sipa ;;bc,prob 1200000 5d
07:41 gribble Error: float division
07:41 RenaKunisaki ;;bc,gen 0
07:41 lfm ;;bc,gen 0.001
07:41 gribble The expected generation output, at 0.001 Khps, given current difficulty of 92347.59095209 , is 1.0891766317e-08 BTC per day and 4.53823596541e-10 BTC per hour.
07:59 CIA-89 bitcoin: genjix * rcc5ed7a95ba9 intersango/util.php: added ignore status. http://tinyurl.com/6k5feth
08:11 omglolbbq guys is there a windows desktop gadget to monitor mtgox?
08:26 toffoo omglolbbq: i think so .. but looks like he's trying to sell it for 1BTC: http://www.reddit.com/r/BitMarket/comments/g94bq/bitcoin_widget_for_windows_for_sale_1_bitcoin/
08:33 xelister hi, when starting new bitcoin node
08:33 xelister how to copy chain files, which ones exactly are needed to not redownload
08:42 omglolbbq tnx toffoo
08:43 eps ;;bc,stats
08:43 gribble Current Blocks: 120288 | Current Difficulty: 92347.59095209 | Next Difficulty At Block: 120959 | Next Difficulty In: 671 blocks | Next Difficulty In About: 3 days, 21 hours, 11 minutes, and 40 seconds | Next Difficulty Estimate: 106298.92485892
08:43 eps woah quite a jump there
08:44 eps it seems like the value rallys are pushing up the difficulty, which seems backwards to me
08:48 Diablo-D3 eps: people are mining harder
08:48 Diablo-D3 to get the money
08:49 eps yeah, i guess it makes sense
08:50 eps i think what bitcoin really needs to take off though is some stability
08:51 eps even if it is stable at low dollar price
08:51 eps cos at least then people can plan more effectively
09:05 manveru eps: it won't get stable if people aren't using it
09:06 manveru mtgox finally seems to be picking up in volume
09:07 manveru but it's still just a few people compared to common exchanges...
09:10 Compgenius eps, probably due to all the pools
09:46 xelister eps: yea
09:46 xelister fuck 110k diff
09:47 eps manveru: if it is never stable then it is unlikely to take off in a big way
09:47 eps and by stable i don't mean static
09:48 eps but an increase or drop of more than 20% (compared to the dollar or whatever) within a few days will make bitcoins difficult to use for legitimate business
10:08 TD eps: yes indeed but there's not much anyone can do about that except get more real traders and merchants into the economy
10:08 TD right now there's lots of speculation and the economy is very small, so the price is driven heavily by press attention
10:10 sipa ;;bc,calc 92000 1
10:10 gribble Error: invalid syntax (<string>, line 1)
10:10 gribble The average time to generate a block at 92000 Khps, given the supplied difficulty of 1, is 46 seconds
10:10 sipa ;;bc,calcd 92000 1
10:10 eps yeah, it is a tricky problem to solve
10:13 anarchyx eps: more volatility will drive in more speculators and increase the amount of traders, market will regulate itself
10:17 TD hmm, faucet is getting low
10:24 mele shares are distributed when a block is solved in a pool?
10:24 BlueMatt mele: depends on the pool, but yes
10:25 mele BlueMat: ok
10:25 sipa mele: you can use [Tab] to complete the nickname of someobe
10:25 sipa someone
10:38 BurtyB just plugged another 5870 in - my UPS isn't too happy lol
11:07 xelister BurtyB: does you UPS actually hold computer with even one 5980
11:07 xelister 5870
12:14 luke-jr hmm, another bubble?
12:16 ArtForz looks more like a single 2kbtc buy
12:16 Kiba Tyler Cowen is very critical of bitcoin
12:16 ArtForz err, 4kbtc
12:16 Kiba hopefully we will succeed and make him look like a fool
12:16 ArtForz Tyler who?
12:17 Kiba sweet revenge
12:17 Kiba dude at Marginal Revolution
12:17 ArtForz still not ringing a bell.. *crawls back under rock*
12:19 CIA-89 bitcoin: Luke Dashjr * r0b4cd41b6a5f gentoo/net-p2p/pushpool/ (Manifest pushpool-0.3.ebuild): net-p2p/pushpool-0.3: latest git eclass is picky http://tinyurl.com/3vlq8zv
12:21 BurtyB xelister yeah, it was doing 2xserver (one with 2x5870) a desktop/2xrouter/switch
12:29 CIA-89 bitcoin: Luke Dashjr * r051b33f5c79d gentoo/net-p2p/pushpool/ (Manifest pushpool-0.3.ebuild): net-p2p/pushpool-0.3: apparently I must not have tested :/ http://tinyurl.com/656hpon
12:29 CIA-89 bitcoin: Luke Dashjr * r844d965662df gentoo/net-p2p/pushpool/ (Manifest pushpool-0.3.ebuild pushpool-9999.ebuild): net-p2p/pushpool: missing DEPEND: dev-libs/libmemcached http://tinyurl.com/6zge35c
12:45 BlueMatt sipa: cpu required?
12:45 sipa i'll benchmark it soon, but i don't expect it to be much more expensive (per key tried) than verification
12:46 BlueMatt sipa: so it would require x2 cpu time (assuming which key works is specified) to verify a tx?
12:47 sipa no, you don't need to verify it anymore
12:47 BlueMatt ah, ok cool then
12:47 sipa if you recover a key, and its hash matches an address, it is certainly a valid signature
12:47 BlueMatt ah, good point that makes sense
12:48 BlueMatt is it possible to specify which key out of the options will be the correct one (ie only one key recovery required)
12:49 BlueMatt I havent really been paying attention
12:49 sipa yes
12:49 sipa there are at most 4 possible keys
12:49 sipa and you can easily specify which one to try
12:50 BlueMatt Id assume we dont want to try any keys other than the one specified
12:50 sipa exactly
12:50 BlueMatt there would be no use
12:51 sipa so i will propose to use a bitcoin-specific 65-byte signature encoding, which contains exactly the same information as the 72-byte signature + 65-byte pubkey now
12:52 BlueMatt sipa: cool, Im assuming we dont have to kill backward compatibility if we use reserved opcodes?
12:53 BlueMatt ie old nodes just reject all txes
12:54 sipa http://www.bitcoin.org/smf/index.php?topic=6430.msg94738#msg94738
12:54 sipa that proposal means we switch (at some point in the past) to new scriptPubKeys which require new scriptSigs
12:54 sipa *future
12:56 sipa i believe it would be possible (using yet another opcode...) to use a "compact signature" for an old scriptPubKey as well
12:57 BlueMatt is the destination address specified not a hash of the current scriptPubKey...
12:58 sipa no
12:58 sipa an address is a hash of a pubkey
12:58 BlueMatt sorry, yes of course
12:58 sipa the scriptPubKey (which typically contains a pubkey or an address) is the thing in a txout which determines who is allowed to spend
12:59 sipa and scriptSig (which typically contains a signature and sometimes a pubkey) is the thing in a txin which proves you had the right to spend
13:00 BlueMatt if we change the format of scriptPubKeys, Im assuming we sign with old format to spend old coins and require new format for spends thereafter?
13:00 sipa there are two ways, i believe
13:01 sipa either add a new opcode to be used in scriptPubKey, which only requires a sig + 2-bit code in scriptSig - so old coins will be signed using the old way, and new coins with the new way
13:02 sipa or add a compatibility opcode to be used in scriptSig, which takes a scriptSig + 2-bit code, and converts it into a scriptSig + pubkey, to be fed to an old scriptPubKey
13:05 BlueMatt Either way old nodes will flat out reject all new signs
13:05 BlueMatt I prefer #1 as IMHO its nicer
13:08 BlueMatt though the results would effectively be the same either way
13:09 sipa #1 allows simpler scriptPubKeys as well, but the effect is very small
13:09 sipa #2 allows small sigs also for old coins
13:10 jgarzik Has bitcoin ever introduced a new feature that made it impossible for older clients to verify a block?
13:10 BlueMatt I dont see sigs for old coins to be really any advantage
13:10 jgarzik sendmany, for example, could be verified by ancient clients, if a sendmany TX makes it into a block.
13:11 BlueMatt IMHO the feature should be added, but the clients shouldnt sign that way by default until it is strictly necessary due to network load
13:11 BlueMatt we cant never add new opcodes just because it breaks compatibility
13:11 BlueMatt though we shouldnt use them by default
13:12 ISA_ Hello
13:14 BlueMatt ISA_: hi
13:16 sipa for a 2-txin-2-txout transaction, it would allow a more than 33% reduction in tx size
13:16 devrandom hi BlueMatt
13:16 sipa 438 -> 290 bytes
13:16 BlueMatt hi
13:17 sipa somehow that seems worth it to me, but obviously not immediately
13:17 BlueMatt sipa: I agree, but IMHO it shouldnt be turned on by default for quite a while
13:17 sipa something like "if (nBestheight > 200000) fUseCompactSigs=true;" :)
13:18 BlueMatt sipa: I disagree
13:18 devrandom so nightlies are built in a VM... I wonder if there's a solution to that
13:18 prax_ anyone got a suggestion for domain name registrar?
13:18 BlueMatt It should be off until absolutely required
13:19 devrandom what kind of VM are you using?
13:19 BlueMatt devrandom: I dont know, depends on the virtualization method
13:19 BlueMatt devrandom: sadly, the proc its on doesnt support hardward virt so its a vmware server
13:19 BlueMatt kvm wont run there
13:19 BlueMatt virtualbox might but vmware definitely wont
13:20 devrandom I wonder if non-kvm qemu would run and if it would be too slow...
13:20 BlueMatt does qemu do soft virt?
13:21 BlueMatt looks light it might
13:21 devrandom yeah
13:21 BlueMatt still dont know if it will boot inside of vmware, but I suppose its worth a shot
13:24 devrandom http://forums.virtualbox.org/viewtopic.php?f=1&t=20589&start=15
13:24 devrandom someone got qemu running under vbox
13:24 BlueMatt worth a shot then
13:25 BurtyB prax_ depending on tld I do registrations
13:25 BlueMatt probably still needlessly slow...
13:25 BlueMatt IMHO its not really worth signing nightlies
13:26 prax_ hmm well why should I use you? (probably just a .com for now though)
13:26 BlueMatt devrandom: signing is important for distributions, but for nightlies...well, its not that important really
13:27 devrandom it would simplify your life if you didn't have to do the releases differently
13:27 BurtyB prax_ icann accredited registrar (not a reseller) for one :)
13:28 BlueMatt devrandom: true, but Ive already got the nigtlies set up (really not hard) and building releases isnt bad with your scripts
13:28 devrandom ok
13:28 prax_ was reading some stuff about how godaddy is a pain to transfer
13:28 prax_ you got info on a website?
13:28 prax_ I'm looking for private registration too
13:28 BlueMatt devrandom: In any case, I think this kind of trust is important for releases
13:29 devrandom BlueMatt - so sounds like the two of us can build releases and that's all we need? gavinandresen?
13:29 B0g4r7 godaddy huh.
13:29 BurtyB prax_ http://www.othellonames.net/ no restrictions (other than the tld operators) on transfering out you can just request an EPP via the control panel
13:29 BlueMatt devrandom: it would be nice to get as many as possible
13:29 gavinandresen Y'all going to take over building releases? Wahoo!
13:29 BlueMatt devrandom: jgarzik, gavinandresen etc
13:29 B0g4r7 I've used them for my domains for many yrs now.
13:29 BlueMatt gavinandresen: I would, but Id rather let people trust you/jgarzik
13:29 B0g4r7 Never had any real problems, other than a kind of "spammy feel" from them.
13:30 devrandom I think gavin builds on ec2?
13:30 BlueMatt gavinandresen: mac building probably not possible though (unless you know of a xcompiler from linux-> mac)
13:30 BlueMatt devrandom: yes he does
13:30 gavinandresen yup.
13:31 devrandom I don't yet have a solution for that, although I can envision using the API
13:31 prax_ BurtyB I'll check it out and maybe talk to you later, still confused what I am really doing too =)
13:32 BlueMatt devrandom: I suppose we can go build 0.3.21 now and get some sigs out if gavin wants to distribute that instead
13:32 BlueMatt get the system started now
13:32 BlueMatt its a good opportunity
13:32 BurtyB prax_ np, let me know if you have any Qs
13:32 gavinandresen How about for the 0.4 release we plan on having the builds done using gitian? I can get a cheap linux netbook to run VMs to build...
13:32 BlueMatt gavinandresen: its up to you
13:32 BlueMatt then the next version is 0.4?
13:33 gavinandresen I'd like the next version to be 0.4 with jgarzik's encrypt-private-keys incorporated
13:33 BlueMatt and autotools, Im assuming?
13:33 gavinandresen (next version after stamping 0.3.21 final today or tomorrow)
13:33 CrazyThinker Is the From and Message data encrypted?
13:33 gavinandresen Yeah, and autotools. And some source re-org, I think
13:33 BlueMatt nice
13:33 jgarzik jaromil's autotools work is quickly overtaking my branch, as the preferred autotools junk station
13:34 jgarzik (with my blessing)
13:34 BlueMatt would like to see bitcoin: uris as well
13:34 jgarzik I like it when people do my work for me :)
13:34 BlueMatt and his should support xcompile soon as well :)
13:34 sipa CrazyThinker: bitcoin does not use any encryption at all
13:34 CrazyThinker sipa, so the messages are visible to everyone on the network?
13:34 sipa all transactions are public, as everyone must be able to verify their integrity
13:34 devrandom we can always re-release the 0.3.21 binary if we are impatient
13:35 devrandom hm... there's this: http://www.phenona.com/blog/using-lxc-linux-containers-in-amazon-ec2/
13:35 sipa CrazyThinker: that does however not mean that everyone knows which addresses belong to who
13:35 BlueMatt devrandom: We can release it of course, and probably should
13:35 gavinandresen What exactly does the gitian build/verify? Just the executables, yes?
13:35 CrazyThinker yeah, but are those visible to everyone?
13:35 sipa jgarzik: seen my comment in your encrypt-private-keys thread?
13:35 sipa some time ago already
13:35 BlueMatt devrandom: that way we can get it out there and tested
13:35 BlueMatt devrandom: until it actually gets used
13:36 jgarzik sipa: I was hoping someone would answer your question :) http://www.bitcoin.org/smf/index.php?topic=4983.msg85445#msg85445
13:36 jgarzik sipa: the main remaining issue seems to be IV
13:36 devrandom gavinandresen - it can have arbitrary output. Right now I think I copy the source to the output and include that in the signature.
13:37 noagendamarket http://twitter.com/#!/MemoryDealers is now accepting bitcoins
13:37 sipa CrazyThinker: there is no identity info in the bitcoin network
13:37 devrandom BlueMatt - ok
13:38 jgarzik anyway...
13:38 sipa CrazyThinker: and bitcoin continuously creates new addresses to send change transactions to, in order to hide which is change and which is payment
13:38 jaromil jgarzik: thanks. i needed autotools myself. BTW i need rebase that branch and provide some documentation for mingw32 compiling
13:38 CrazyThinker sipa but if I mention the email address and other details in the message sent using Bitcoin client, can people read this information?
13:38 sipa there is no message
13:38 TD good day everyone
13:39 BlueMatt TD: good day
13:39 gavinandresen Wow, the gang's all here....
13:39 TD gavinandresen: could you maybe include the "if (!fTestNet)" conditions on the tx replacement and other disabled features before 0.3.21?
13:39 TD so everything that is currently disabled for safety can be played with on the testnet ....
13:39 luke-jr gavinandresen: jgarzik just left. he isn't part of the gang? :P
13:39 CrazyThinker sipa, In the bitcoin client it shows message and From text boxes
13:39 sipa CrazyThinker: that's only for the outdated send-to-IP feature
13:39 gavinandresen TD: I'd rather not-- I want to get 0.3.21 out the door
13:39 sipa CrazyThinker: the bitcoin network doesn't use them
13:39 TD fair enough
13:40 CrazyThinker sipa, okay
13:40 B0g4r7 TD: Verying behavior based on whether or not "test mode" is being used sounds like a bad idea to me.
13:40 gavinandresen TD: It'll be top of the list for 0.4 (assuming somebody submits a PULL or I remember)
13:40 devrandom jgarzik - are you interested in joining the gitian builders?
13:40 TD great.
13:40 B0g4r7 varying
13:40 luke-jr &
13:40 gavinandresen jgarzik just left
13:40 BlueMatt devrandom: hes off
13:40 sipa TD: by the way, i have an openssl-based ecdsa key recovery working
13:41 TD B0g4r7: it's already done. they are separate networks
13:41 TD sipa: nice!
13:41 devrandom ok
13:41 B0g4r7 mmm...well...yes.
13:41 gavinandresen sipa: but... why?
13:41 gavinandresen (bottleneck is already CPU, why make it worse?)
13:41 gavinandresen ("because I can" is a valid response, by the way)
13:41 sipa gavinandresen: i haven't benchmarked it, but key recovery shouldn't be slower than verification, actually
13:42 BlueMatt gavinandresen: probably no worse for cpu
13:42 sipa obviously, i still need to do benchmarks
13:42 gavinandresen But you have to recover the CPU AND THEN STILL check the signature, right?
13:42 sipa no
13:42 gavinandresen (err, recover the key)
13:42 BlueMatt gavinandresen: theoretically you dont have to check the sig
13:42 sipa if you recover a key, and that key's hash matches your address, you're done
13:42 sipa the algorithm will only produce keys for which that sig is a valid sig
13:43 TD bottleneck is CPU today. in future it might move to become the RAM required to store the block chain (assuming that big miners have many cores)
13:43 TD so shrinking the block chain is still useful. but the change isn't backwards compatible.
13:44 TD if it's included today, perhaps in a year or two years if miners start complaining that the storage overheads are getting too heavy, it could be activated
13:44 TD as by that point hopefully everyone is upgraded
13:44 TD ie - kept around in our back pocket in case it's needed
13:44 sipa for a typical 2-txin-to-2-txout transaction, the size (of the whole tx) would go from 438 to 290 bytes
13:44 BlueMatt TD: that is what I was saying
13:44 TD ah, sorry for duplicating :) i wasn't watching irc before
13:44 BlueMatt keep it off until strictly necessary
13:45 BlueMatt TD: never a problem, I do that all the time
13:45 sipa it does put stronger constraints on alternative implementations
13:45 luke-jr ;;later tell jgarzik is there a reason pushpool uses sbin instead of bin?
13:45 gribble The operation succeeded.
13:45 TD sipa: how so?
13:45 sipa TD: as they would also need to implement the additional opcodes, obviousl
13:46 TD afaik the only alternative implementation today that tries to verify txns is bitdollar and it doesn't even run scripts
13:46 BlueMatt just more stuff to implement
13:46 BlueMatt assuming you want to verify txes yourself
13:46 TD right, but as no such implementation exists today, it's not a big deal to give them even more work :-)
13:46 BlueMatt true
13:46 gavinandresen sipa: ok. carry on then. But a factor of less-than-two storage savings... at the cost of an incompatible addition to Script.cpp ... I'm not enthused.
13:46 TD actually i'm wrong ... i'd have to support it too
13:47 luke-jr TD: but nothing could stop a rogue miner from using it in blocks early, and forking the chain
13:47 devrandom the miner would not get their blocks accepted, wasting gpus
13:47 sipa a rogue miner who starts to use it before the official "introduction time", would just be ignored by others
13:47 TD as even thin clients need to understand how to extract the relevant addresses/pubkeys from scripts
13:47 TD even though they don't run them
13:47 BlueMatt luke-jr: could be disabled till block x
13:47 luke-jr BlueMatt: that would do it I guess
13:48 devrandom I like the enable after block x solution for upgrades... maybe should also have an alert system for "your version is considered obsolete by n peers"
13:48 luke-jr devrandom: there is such an alert system already
13:49 luke-jr devrandom: except it's monarchial
13:49 BlueMatt Kiba: potato chip seems kinda random there, just one?
13:49 CIA-89 bitcoin: Luke Dashjr * rdf81e33c4c2f gentoo/net-p2p/pushpool/ (Manifest pushpool-0.3.ebuild pushpool-9999.ebuild): net-p2p/pushpool: oh yeah, I forgot about econf... http://tinyurl.com/42t3axq
13:49 devrandom luke-jr - the satoshi alert?
13:49 luke-jr yeah
13:49 Kiba BlueMatt: naw, vingear and salt chips, my favorite
13:50 luke-jr gavinandresen: what would you think of modifying the alert to relay signed-by-other-key messages on the same terms as a free txn?
13:50 BlueMatt gavinandresen: can we get you to add a new alert key for yourself...or get satoshi to give you his?
13:50 devrandom I'm thinking about something that encourages people to upgrade if their version is obsolete. based on security upgrades or upcoming breaking changes
13:50 luke-jr gavinandresen: so, for example, other clients can broadcast alerts only to their client users
13:51 gavinandresen luke-jr: interesting idea
13:52 gavinandresen BlueMatt: interesting you should mention that, Satoshi sent me the alert key this morning.
13:52 BlueMatt gavinandresen: ah, good then
13:52 Kiba will satoshi come back, gavinandresen?
13:52 BlueMatt Kiba: last I heard, he had "moved on to other projects"
13:53 BlueMatt and that bitcoin is "in good hands"
13:53 Kiba totally inaccurate rumor methink
13:53 BlueMatt Kiba: that is what TD told me satoshi said in an email
13:53 Kiba gavinandresen: so...what's official statement
13:53 gavinandresen Kiba: I don't know, he says he's moving on to other things, so expect him to fade away even more.
13:53 eps is that possible?
13:53 Kiba so we will never knows the true identity of the project?
13:53 Kiba err
13:53 noagendamarket :(
13:53 Kiba founder
13:53 TD yeah, i asked him that and he gave me the answer BlueMatt said. he's gone for good, but still answers questions from time to time
13:54 Kiba so. we will
13:54 TD some of them at least .... if they don't involve him doing much work :-) i asked him if he'd be willing to review a wiki page and he ignored that part of my mail, heh
13:54 Kiba never know how the true origin of bitcoin..how its inventor come up with a system
13:54 eps well the cryptocurrency is not a new idea
13:54 BlueMatt Kiba: if it catches on, someone will do the research
13:55 eps but the execution of bitcoin is very impressive
13:55 Kiba satoshi is a true hero in any case
13:55 BlueMatt eps: no, but satoshi was one of the first to come up with the idea of proof of work
13:55 Kiba no
13:55 BlueMatt he wasnt the first
13:55 Kiba he just integrate other people's idea
13:55 BlueMatt but the first to implement it very well afaik
13:55 xxxxxxx BlueMatt, implementing proof of work is ez as shit
13:55 eps i always say execution beats ideas
13:55 TD using partial pre-images of zero as a proof of work is from Dr Adam Back
13:56 eps i have ideas
13:56 TD the block chain is (afaik) a new algorithm
13:56 BlueMatt xxxxxxx: never said it was hard, but bitcoin is done VERY well
13:56 xxxxxxx u might be confusing i w the block chain
13:56 eps but i don't necessarily execute them very well
13:56 Kiba satoshi will remain...a legend
13:56 MacRohard bootstrapping the currency by allowing everyone to mine coins might be a satoshi inention
13:56 MacRohard invention
13:56 Kiba his invention is ingenuiously puting together all the ideas
13:56 DingoRabiit BlueMatt: Hey man you active? i lost your Repay address, Could you please send me it ina PM? It's JackRabiit.
13:56 xxxxxxx BlueMatt, proof of work is like a few lines of code and was implemented multiple times to prevent spam a long time ago
13:57 Kiba satoshi might come up with a new project under a different name in a few years
13:57 gavinandresen We'll know it's him by all the pszStrings in the code....
13:57 eps heh
13:57 Kiba lol
13:57 BlueMatt xxxxxxx: but using it in a currency like satoshi and his clear talent in the code...
13:58 xxxxxxx BlueMatt, im not saying anyhting about satoshi or bitcoin, just proof o work
13:58 eps there is nothing new under the sun
13:58 BlueMatt Kiba: I have a feeling satoshi will come back to bitcoin under a different name if it catches on
13:58 BlueMatt xxxxxxx: what eps said
13:58 eps everything is built on existing ideas
13:58 TD he may well end up as a regular user of it
13:59 BlueMatt there hasnt been anything big new in comp sci that is big in a long time
13:59 TD whether he takes part in the community beyond that is something we'll never know. personally i doubt it.
13:59 TD he was working on it for nearly 4 years now. i know from past experience that after 4 years of a project you do want to move on
13:59 Kiba we will never figure out
13:59 Kiba and even if he told us
13:59 Kiba we would never believe him :)
13:59 BlueMatt TD: I agree
14:00 devrandom Kiba - he has a pubkey, so he can prove his identity if he wants
14:00 BlueMatt devrandom: I doubt he would
14:00 devrandom right
14:00 BlueMatt the mystery is great
14:01 Blitzboom im wondering if he will ever sell his coins
14:01 Kiba probably won't
14:01 Kiba for a long time
14:01 Blitzboom or wait until he can purchase whatever he wants with them
14:02 Kiba and nobody will be able to know that he's the world's richest man by that point
14:02 devrandom were the genesis coins ever spent? I guess I can easily check that in blockexplorer
14:02 BlueMatt he wont sell the ones from the first blocks that we know are his
14:02 BlueMatt devrandom: no, and I highly doubt satoshi will ever spend those
14:02 eps can you choose which bitcoins to spend?
14:02 BlueMatt if he still has the keys to them
14:02 BlueMatt eps: not in the current client
14:03 eps i guess you could put them in different wallets
14:03 BlueMatt but satoshi easily could
14:03 BlueMatt he knows the software pretty well ;)
14:03 Kiba he can launder his money
14:03 TD i think we should open bets on which "who is satoshi" conspiracy theory becomes the most popular
14:03 xelister gavinandresen: yeah, how could you type all this hungarian notation :<
14:03 Blitzboom i hope satoshi hasnt lost his keys :D
14:03 xelister TD: I know
14:03 BlueMatt TD: Its gavin ;)
14:03 TD my bet is on gavin too :-)
14:03 eps some say he started writing code at the age of 2
14:04 Kiba lol
14:04 xelister information on who is Satoshi (except speculation about gavinandresen)
14:04 xelister for 10 BTC
14:04 xelister or next higher better =) (that can choose to get it in private)
14:04 eps some say he is descended from adam smith
14:04 TD haha
14:04 TD chuck norris has nothing on satoshi
14:04 eps all we know is... his name is satoshi, seriously, that is all we know
14:05 xelister more exactly, I found a (fictional) charactere that closelly reassembles Satoshi AND is quite known
14:05 BlueMatt some say he is a russian coder who has spent his entire life as an orphan in a code house (apparently they have those in russia)
14:05 xelister guys?
14:05 xelister *** I found a (fictional) charactere that closelly reassembles Satoshi AND is quite known ****
14:05 BlueMatt who?
14:05 xelister more then 1:1,000,000 close match
14:06 gavinandresen Satoshi did suggest this morning that I (we) should try to de-emphasize the whole "mysterious founder" thing when talking publically about Bitcoin. It plays into the "bitcoin is pirate money" meme.
14:06 xelister it will cost to find out, starting price 10 btc =)
14:06 xelister gavinandresen: I thought only americafags use that
14:06 TD yeah
14:06 xelister war on children - think about the pirates
14:06 xelister no wait, or viceversa
14:06 BlueMatt gavinandresen: true, for common people its probably best not to, but it will catch on eventually (if bitcoin catches on)
14:06 TD between the anonymous founder and the underground black markets there is enough already ....
14:07 gavinandresen The press loves a mystery-- gives them a good hook for interesting stories.
14:07 devrandom first used key is in block 9.... http://blockexplorer.com/a/fuWLnx1cN
14:07 eps anyone who understands what open source is won't be bothered by satoshi's secret identity
14:07 Kiba we know next to nothing about satoshi
14:07 eps but it will probably take a while for the average joe to get to that point
14:07 xelister jp male
14:07 xelister coder
14:08 Blitzboom i dont want to. satoshis anonymity is a necessity to protect him
14:08 xelister hacker
14:08 xelister realted to economy
14:08 MacRohard maybe bitcoin was setup by the narcocartels as a way to modernize their distribution system
14:08 xelister related to programming
14:08 gavinandresen He also suggested that I give you'all more credit-- and I agree, you'all deserve more credit....
14:08 xelister will be made RICH slowly if his idea works out
14:08 Kiba bitcoin is skynet and we are its machine
14:08 xelister and his program has a theme of micro payments, and ~0.01 payments
14:08 BlueMatt gavinandresen: the problem is there are many of us, and one of him so more credit to us still means he gets the most ;)
14:08 xelister I FOUND A CHARACTER MATCHING ALL ABOVE CRITERIA :)
14:09 Kiba Is it L?
14:09 xelister no
14:09 Kiba japanaese male. world's best detective. Half british.
14:09 Kiba love sugar and candy
14:09 xelister Satosi is not a fucking detective
14:09 Kiba lol
14:10 xelister he is a hacker if at all
14:10 BlueMatt so...japanese sherlock?
14:10 xelister and a bit of revolutionist
14:10 gavinandresen no, he's a rogue british secret agent....
14:10 xelister and I found person matching the above :)
14:10 gavinandresen (damn it, I just did it again.... too much fun speculating....)
14:10 xelister gavinandresen: 1000 BTC to know what the person is? Surly its change money for you ;)
14:10 Kiba no
14:10 Blitzboom satoshi = secret military project
14:10 Kiba Satoshi is a half Japanese hacker girl pretending to be a male
14:11 midnightmagic i understand what open source is, and satoshi's identity is irrelevant. what's more important is knowing more about the rationale behind the design, and that i don't see.
14:11 BlueMatt I thought someone did an analysis of post times and determined hes probably american?
14:11 Kiba American or British
14:11 midnightmagic might just be in the U.S.
14:11 BlueMatt Kiba: very, very different time zones
14:12 TD midnightmagic: i agree, that's why i've been drilling satoshi with questions for the past months. i'm sure gavin has been doing the same
14:12 midnightmagic does he answer you?
14:12 TD parts of the design are definitely clearer in my mind now. it's unfortunate that nearly nothing about bitcoin was documented.
14:12 TD beyond the paper
14:12 TD yes he does answer me, though sometimes it takes a few days
14:13 midnightmagic are you writing this stuff down somewhere?
14:13 devrandom TD - what remains unclear? economic dynamics?
14:13 TD the emails come in at basically random times. he uses outlook express on a mac, i suspect he artificially delays responses to make it harder to figure out where he is :-)
14:13 TD no, some details of how best to use the more obscure features
14:13 xelister SATOSHI
14:13 xelister USES
14:13 xelister OUTLOOK EXPRESS
14:13 xelister ????????????????????????????????????????
14:13 xelister what the fuck
14:13 luke-jr lol
14:13 xelister what the fuck the fuck.
14:13 luke-jr that explains a bit
14:13 xelister only retard use OE
14:13 BlueMatt probably faking th headers
14:13 xelister *retards
14:13 luke-jr OE isn't all that bad
14:13 luke-jr but it IS Windows
14:14 luke-jr xelister: he probably emails from a library
14:14 TD he didn't comment when i asked him about the dynamics of mining in a fully fee based system. but i already decided it won't likely be an issue so i'm not too worried about that ;-)
14:14 devrandom TD - curious to hear which ones?
14:14 B0g4r7 It would be good to preserve those emails.
14:14 xelister well
14:14 TD i've been asking mostly about things like sequence numbers, nLockTime, designing distributed contracts and some details of how he imagined thin clients to work.
14:14 gavinandresen TD: actually, I haven't been grilling him-- I've been too busy with other stuff to dive deep into the code. You should definitely write up what you've learned.
14:15 xelister I know a personal that matches Satoshi as closelly as 1 in 10 milion or so =)
14:15 TD yeah. i'm going to write up a wiki page on the distributed contracts stuff
14:15 TD like i said, i asked if he'd review that but he ignored that part of my mail. i think he wants to stay as hands off as possible now.
14:15 Kiba distributed contracts
14:15 TD i'll write up what i've learned when i feel i understand it a bit better
14:15 midnightmagic TD: can i request that you write these things down somewhere so it's not just stuck in the half-life of someone's brain? :(
14:16 midnightmagic are you barred from posting his discussion with you?
14:16 BlueMatt TD: does he still use the @gmx.com email?
14:16 BlueMatt midnightmagic: you think TD throws away emails, he uses gmail, wasnt that the whole point google pushed when they started gmail?
14:16 BurtyB oi xelister I wouldnt class myself as a retarded OE user :p
14:17 luke-jr BlueMatt: so Google throws away emails for you?
14:17 BlueMatt luke-jr: point was they dont
14:17 BlueMatt ie gmail was designed so that you never delete email
14:17 phantomcircuit there's a thunderbird extension that forges headers, it defaults to outlook express
14:17 midnightmagic BlueMatt: first off, i didn't know that. secondly, it doesn't matter: if TD gets hit by a bus, all that conversation "goes away".
14:18 BlueMatt midnightmagic: it was mostly sarcastic to point out google's odd habits as a company
14:18 midnightmagic more sarcasm, doh. of course. it's BM.
14:18 midnightmagic my mistake
14:18 TD i keep all the mail, obviously
14:19 BlueMatt yep, sorry...I really need to stop doing that online
14:19 TD and he still responds to me on the gmx.com address. but i don't know if he replies to everyone. i think justmoon contacted him asking about official approval for the video and didn't get a response.
14:19 midnightmagic BlueMatt: I don't think so, i say keep it up, it's more interesting that way
14:19 TD my impression is he's willing to discuss the technical details, perhaps with a few people, but otherwise doesn't want to be involved with the project any more. which makes sense.
14:20 BlueMatt midnightmagic: :)
14:20 genjix TD: do you use sync threads or async reactors?
14:21 TD genjix: i'm lacking context. use in what ?
14:21 TD bitcoinj? gmail?
14:21 genjix for connections in bitcoinj
14:21 devrandom BlueMatt - is autotools going into .21?
14:21 genjix i think you use threads
14:21 TD threads
14:21 genjix kk
14:21 BlueMatt devrandom: no, just 0.4
14:21 TD by the way, satoshi anticipated the rise of GPU mining right from the start
14:21 devrandom ok
14:21 genjix ohh interesting
14:39 kile BlueMatt: when i mine using a pooler my miner sends hashes to the pool right?
14:40 lfm kile: not really, you send the block headers with nonces which produce the hashes wanted
14:41 kile lfm: im trying to understand how poolers create shares
14:41 lfm kile: well first do you understand how solo mining works?
14:42 kile lfm: yes can you explain me a bit?
14:42 kile lfm: i dont understand very good how blocks are passed away and stuff
14:42 kile lfm: i think each bitcoin client gets a few blocks to solve ( calculate a valid hash for the block )
14:42 kile thats all i understand
14:43 xelister kile: you get milions of milions hashes to solve, one of them will get you a 50 btc rewar
14:43 xelister reward
14:43 lfm ok, you're trying to find the 80 byte block headers that produce the hashes which pass the target constraints(difficulty)
14:43 xelister depending on luck it can take weeks or months
14:43 kile xelister: i see so my bitcoin client gets millions of millons of hashes to solve or blocks?
14:43 kile xlister: i think my bitcoin client gets blocks and is asked to hash them, right ?
14:44 lfm solving isnt really what your doing. it more like searching
14:44 xelister kile: find N so that sha256( sha256( data + N ) ) < target. So your comkputer tries milions of N and if he finds the 'correct' one it gets a bonus
14:44 xxxxxxx kile, you try to find a random string which hashes to a value less than X, say X is 10, which is really hard to do, when u coop mine, you send the mine pool any value that is less than 100 instead of 10,which is easier to do, and not small enough to solve a block, but small enuf to prove to them you are in fact mining
14:45 xelister I even have a pool that mathematicall calculates
14:45 xelister if you should be solo mining or not
14:45 xxxxxxx so they know what % of hashpower ur contributing to the pool
14:45 xelister it consist just few simple questions
14:45 xelister actually it is a really easy questionnare to determin if you should solo mine
14:46 kile xxxxxxx: why would a pool want the value that be less than 100 ? if it only needs the one thats less than 10 ?
14:46 xxxxxxx kile, to prove to them ur actually trying to mien stuff
14:46 kile i think the pool would discard those >10
14:46 xxxxxxx well, to prove what speed ur hasing at
14:46 lfm kila just so it can tell you are trying
14:46 xxxxxxx yea it discards them, but this way it knows how much to pay who when someone finally solves a block
14:47 xxxxxxx so if i send in hashes < 100 at a rate of 1 per hour and you send tem in at 2 per hour
14:47 xxxxxxx ur doing twice as much work as me
14:48 xelister well just do the questionnare
14:48 kile xxxxxxx: what if everyone on the pool ( all the miners ) just send values that are 80<x<90 ? they will still get shares? and what if there is only one guy that sends the < 10 hash ?
14:48 lfm the two numbers are actually more like 2^32 and (2^32 * 100000)
14:48 xxxxxxx kile, yea they will
14:48 kile it takes the same speed to calculate a <10 hash that to calculate a 80 hash right?
14:48 xxxxxxx kile, and the guy that acutally solves the hash will only get whatever his contribution was, he doesnt get anyhting special for finding it
14:57 kile xxxxxxx: and if some other pool solved the block first?
14:57 xxxxxxx kile, thats documented somewhere
14:57 kile what happens? the block gets orphan and all the miners wont earn anything?
14:57 xxxxxxx u get a split chain until teh next block
14:57 xxxxxxx or something