1 2013-10-11 02:24:52 <BingoBoingo> How common is it for one bitcoin public key to has to multiple addresses?
  2 2013-10-11 02:27:45 <Jere_Jones> BingoBoingo: That isn't possible since an address is just a hash of a public key. Explain what happened without trying to troubleshoot it and you may have more luck finding out what really happened.
  3 2013-10-11 02:29:52 <BingoBoingo> I signed a message with an address several weeks ago. I put the signed message in the Brainwallet.org verifier today (someone else actually tried it first). Brainwallet.org verified it to an address other than the one I used to sign the message.
  4 2013-10-11 02:30:29 <Jere_Jones> And Bitcoin-qt verified it to a completely different (third) address.
  5 2013-10-11 02:33:05 <BingoBoingo> Jere_Jones: And mathematics seems broken. Or RipeMD 160 seems to have collisions. Maybe it is the case that this is just karma punishing me for a vanity address with firstbits 1lvcunts. Whatever it is seems dangerous.
  6 2013-10-11 02:41:48 <edcba> https://bitcointalk.org/index.php?topic=251037.0 ?
  7 2013-10-11 02:42:24 <edcba> anyway maths are not broken nor ripemd160
  8 2013-10-11 02:43:04 <edcba> also bugs exist and more likely in brainwallet than in bitcoin
  9 2013-10-11 02:45:54 <BingoBoingo> edcba: Bitcoin-qt offers a third address that could have validly signed the message.
 10 2013-10-11 02:49:50 <edcba> ok i read about brainwallet
 11 2013-10-11 02:50:05 <edcba> and if it does generate addresses from passphrases
 12 2013-10-11 02:50:15 <edcba> then maybe you have chosen a weak passphrase
 13 2013-10-11 02:50:32 <edcba> (if you used brainwallet to generate public key)
 14 2013-10-11 02:56:02 <BingoBoingo> edcba: The address wasn't generated from brainwallet though. I just tried verifying the signature there.
 15 2013-10-11 04:03:47 <BingoBoingo> jgarzik: I have a problem with bitcoin address message signing.
 16 2013-10-11 04:09:13 <jgarzik> BingoBoingo, what is the problem?   bitcoin version and OS/platform version?
 17 2013-10-11 04:12:40 <BingoBoingo> Address generated with vanitygenocl on Ubuntu 12.04 LTS. Brainwallet.org and Bitcoin-QT's verify signed message functions both verify signed messages to differing addresses than the one I've had imported to Multibit for quite some time.
 18 2013-10-11 04:14:15 <jgarzik> BingoBoingo, there are two bitcoin addresses for each private key, because there can be two public key datastreams: a compressed public key, and an uncompressed public key, each with their own pubkeyhash
 19 2013-10-11 04:14:16 <gmaxwell> BingoBoingo: what agrees with what?
 20 2013-10-11 04:14:54 <gmaxwell> jgarzik: kinda, except a correctly encoded private key indicates which should be used.
 21 2013-10-11 04:15:02 <jgarzik> BingoBoingo, This is most likely a mismatch between multi bit wanting uncompressed and bitcoin-qt wanting compressed
 22 2013-10-11 04:15:15 <jgarzik> gmaxwell, I doubt vanitygen and its ilk do that
 23 2013-10-11 04:15:20 <BingoBoingo> Original vanitygen'd address was 1LvCuntsJyFFQsLuJhBXBPokbQa7SAyMr4 Brainwallet.org verifies the struck though signature http://www.thedrinkingrecord.com/contact-and-credentials/ to the address 1LC9GzYK1zjrG9zAvMPExcwHTP8aPYPMyY a third result was apparently returned by verifying the signature in Bitcoin-QT but I don't have the patience for that.
 24 2013-10-11 04:15:34 <gmaxwell> jgarzik: they do. (the format makes it impossible not to)
 25 2013-10-11 04:18:00 <gmaxwell> BingoBoingo: your message is >254 characters long, a _lot_ of tools create corrupted signatures in that case.
 26 2013-10-11 04:18:21 <gmaxwell> BingoBoingo: can you try signing something short like "test"
 27 2013-10-11 04:18:49 <gmaxwell> if bitcoin-qt and multibit agree then (I'll test bitcoin-qt for you) then multibit is buggy.
 28 2013-10-11 04:22:24 <BingoBoingo> gmaxwell: I can verify Multibit 0.5.11 reports the message as valid from 1LvCuntsJyFFQsLuJhBXBPokbQa7SAyMr4, but other tools that don't request the address have suggested at least two other messages so far when the message in question is "I may often be found on CoinChat using the nick BingoBoingo and on bitcointalk.org as Atruk. I use the bitcoin address 1LvCuntsJyFFQsLuJhBXBPokbQa7SAyMr4 as a message signing address. It
 29 2013-10-11 04:22:25 <BingoBoingo> isn’t my oldest address or my only address, but it does have a history of transactions visible on Blockchain.info going back to March 14th, 2013 and has been mentioned by my screennames at various points."
 30 2013-10-11 04:27:57 <BingoBoingo> Maybe address signed messages should have something roughly equivalent to the armoured ascii format? I've seen jgarzik had threads on Reddit about using address signatures as an alternative to password logins for websites, but maybe this shit should be at least equal to GPG first?
 31 2013-10-11 04:30:46 <BingoBoingo> I mean not having a container like gpg's armoured ascii was reasonable. When nobody can sign messages the same way from different clients this shit looks ridiculous.
 32 2013-10-11 04:31:33 <gmaxwell> BingoBoingo: no amount of container stuff has ANYTHING to do with software which is just implemented _wrong_.
 33 2013-10-11 04:31:59 <gmaxwell> BingoBoingo: if you're not going to conduct the simple test I suggested then you might as well leave the channel.
 34 2013-10-11 04:37:13 <dexX7> what test did you suggest?
 35 2013-10-11 04:37:51 <BingoBoingo> gmaxwell: Bitcoin-qt/bitcoind and Multibit report different addresses. Brainwallet.org presents a third (though I tested it second and chronologically its deviance is second). This is honestly weird an worrisome. I know Address signed messages have quirks after some ~250 characters, but I am disturbed different clients would assume different addresses for the same public key.
 36 2013-10-11 04:40:00 <BingoBoingo> I'd try this in Electrum as well, but they seem to have buried their message signing and verification functions.
 37 2013-10-11 04:40:00 <gmaxwell> oh he left before I could kick him for not running my @##$@ test.
 38 2013-10-11 04:40:05 <gmaxwell> there.
 39 2013-10-11 04:40:23 <gmaxwell> oh he didn't.
 40 2013-10-11 04:41:11 <gmaxwell> dexX7: I asked him to sign a short string like "test" to see if multibit has broken varint coding.
 41 2013-10-11 04:41:52 <dexX7> did you see this thread? https://bitcointalk.org/index.php?topic=309179.msg3317290#msg3317290
 42 2013-10-11 04:42:04 <TheLordOfTime> gmaxwell, kickban a habit click for you?  (purely curious hence the question)
 43 2013-10-11 04:42:20 <TheLordOfTime> (especially because banned, kicked, unbanned)
 44 2013-10-11 04:43:06 <gmaxwell> TheLordOfTime: It just prevents the instant rejoin many clients have.
 45 2013-10-11 04:43:21 <gmaxwell> kick alone isn't even noticed by users sometimes.
 46 2013-10-11 04:43:29 <gmaxwell> esp in channels with high join/part rates.
 47 2013-10-11 04:43:47 <TheLordOfTime> gmaxwell, remove works too
 48 2013-10-11 04:44:04 <TheLordOfTime> and actually breaks most autorejoins
 49 2013-10-11 04:48:23 <gmaxwell> BingoBoingo: Have you performed the test I asked for yet?
 50 2013-10-11 04:49:11 <gmaxwell> For your reference Bitcoin-qt does not "present a third", as it does display any address at all, it prompts you to provide an address.
 51 2013-10-11 04:49:58 <dexX7> <BingoBoingo> dexX7: THe signature for "test" is Gwn1aQana5m/0k0UvrMyrDHroyFfcJVVBovdKjB97rrsxr9yc0vQD+2PHIE1KK6Tq01bLAJ5tUkRY0v8R683Mg8= << if that's the answer for your test
 52 2013-10-11 04:50:03 <gmaxwell> Every validatly encoded signature is a signature for _some_ address. The fact that you can get an address out of a broken implementation that tells you the address rather than asking for one is itself not concerning.
 53 2013-10-11 04:51:08 <dexX7> the problem is solely this special char + more than 249 other (normal) chars, everything else works as expected
 54 2013-10-11 04:51:31 <gmaxwell> bitcoind verifymessage 1LvCuntsJyFFQsLuJhBXBPokbQa7SAyMr4 'Gwn1aQana5m/0k0UvrMyrDHroyFfcJVVBovdKjB97rrsxr9yc0vQD+2PHIE1KK6Tq01bLAJ5tUkRY0v8R683Mg8=' 'test'
 55 2013-10-11 04:51:38 <dexX7> correct
 56 2013-10-11 04:51:42 <gmaxwell> dexX7: why to you say any "special" character is required.
 57 2013-10-11 04:52:06 <dexX7> i don't know. but that's the case when this strange behaviour appears.
 58 2013-10-11 04:52:16 <gmaxwell> dexX7: I doubt it.
 59 2013-10-11 04:52:35 <dexX7> please take a look at this thread to reproduce it: https://bitcointalk.org/index.php?topic=309179.msg3317290#msg3317290
 60 2013-10-11 04:52:51 <dexX7> i
 61 2013-10-11 04:52:56 <dexX7> i'm here, because i'm curious :)
 62 2013-10-11 04:53:23 <gmaxwell> I see the brainwallet code is simply wrong, it's not encoding a varint at all: https://github.com/brainwallet/brainwallet.github.com/blob/master/js/bitcoinsig.js#L11
 63 2013-10-11 04:55:27 <dexX7> isn't that what msg_bytes(message) does?
 64 2013-10-11 04:56:16 <dexX7> i have no idea, but if you can say "that's the source", well.. then that's it.
 65 2013-10-11 04:57:31 <gmaxwell> I don't know how to link the the right thing, grr github.
 66 2013-10-11 04:57:38 <gmaxwell> In any case, the byte order is swapped.
 67 2013-10-11 07:06:01 <Luke-Jr> hmm, seems if you suspend bitcoind for a few minutes and come back up with a new IP, it never recovers blockchain sync (fully synced prior to suspend) O.o
 68 2013-10-11 07:06:10 <Luke-Jr> all the blocks coming in are being marked orphan
 69 2013-10-11 07:06:17 <Luke-Jr> and it never tries to resolve it
 70 2013-10-11 07:07:50 <gmaxwell> hm? I've never seen it do that and I suspect my bitcoind twice a day for my ~10 minute trip to and from the office. What version?
 71 2013-10-11 07:07:56 <gmaxwell> s/suspect/suspend/ :P
 72 2013-10-11 07:08:08 <gmaxwell> ;;bc,blocks
 73 2013-10-11 07:08:09 <gribble> 262893
 74 2013-10-11 07:09:48 <Luke-Jr> gmaxwell: 0.8.4
 75 2013-10-11 07:10:12 <Luke-Jr> gmaxwell: I SIGSTOP'd it to shut debug.log up, did a live migration on the server to a new datacenter, and SIGCONT'd
 76 2013-10-11 07:10:35 <Luke-Jr>     "blocks" : 262873,
 77 2013-10-11 07:12:05 <Luke-Jr> (this is #bitcoin-watch's node fwiw)
 78 2013-10-11 07:12:51 <Luke-Jr> just started syncing again
 79 2013-10-11 07:13:56 <gmaxwell> on its own?
 80 2013-10-11 07:14:13 <Luke-Jr> unless someone here connected to it and pushed the block
 81 2013-10-11 07:14:44 <Luke-Jr> but I didn't post the IP either, so that's probably not likely
 82 2013-10-11 07:17:01 <gmaxwell> okay, well, that behavior I can't confirm or deny, its possible that my laptop sometimes gets stuck for a bit after a suspend/resume cycle— I wouldn't notice that.
 83 2013-10-11 10:44:43 <jouke> when a transaction is made before the client is fully synced (so not transmitted to the network), is it possible to let bitcoin-qt forget about that transaction?
 84 2013-10-11 10:48:39 <michagogo> jouke: Yes
 85 2013-10-11 10:48:44 <michagogo> Oh, except...
 86 2013-10-11 10:48:51 <michagogo> It *does* get breadcast to the network
 87 2013-10-11 10:49:04 <michagogo> (it also gets broadcast)
 88 2013-10-11 10:49:25 <sipa> there is no way afaik to remove wallet transactions
 89 2013-10-11 10:49:35 <michagogo> Well, there are 2 ways
 90 2013-10-11 10:49:39 <sipa> unless they are coinbases that get reverted
 91 2013-10-11 10:50:01 <michagogo> One is to edit the wallet
 92 2013-10-11 10:50:21 <michagogo> either with pywallet, or with db_dump or whatever it's called
 93 2013-10-11 10:50:44 <sipa> yeah sure, that works
 94 2013-10-11 10:51:05 <michagogo> The other, which is easier but will forget all unconfirmed transactions, address labels, and transaction timestamps, is salvagewallet
 95 2013-10-11 11:18:54 <jouke> michagogo: thanks.
 96 2013-10-11 11:19:51 <michagogo> np
 97 2013-10-11 11:45:07 <melvster> FYI: i posted about the vocab I was talking about yesterday, there's no changes needed from the dev team, only something extra for those working on the intersection of bitcion and the semantic web ...
 98 2013-10-11 11:45:54 <melvster> hopefully can help bitcoin to a wider audience
 99 2013-10-11 11:49:01 <michagogo> Why was alert 1033 set to expire after a day?
100 2013-10-11 13:21:27 <jgarzik> mornin'
101 2013-10-11 13:31:43 <michagogo> jgarzik: o/
102 2013-10-11 17:22:52 <pierce> a question someone asked me recently that I wasn't able to answer : what is stopping an attacker from starting when the difficulty was much lower, and creating a super long blockchain, and have it overtake the existing blockchain?
103 2013-10-11 17:23:47 <pierce> like, assuming they have a client modified to think time is moving super fast, so they could crank out blocks quickly without increasing the difficulty
104 2013-10-11 17:24:17 <pierce> is there some limit on how far the blockchain will rewind?
105 2013-10-11 17:27:43 <sipa> it's the total work on a chain that counts
106 2013-10-11 17:27:44 <pierce> a sneaky attacker could even emulate all the old transactions, so everyone would have the same amount of coins, except for a select few untouched mined blocks which would go to the attacker
107 2013-10-11 17:27:46 <sipa> not its length
108 2013-10-11 17:28:09 <pierce> sipa: ah, interesting, so difficulty*lengh essentially?
109 2013-10-11 17:28:10 <sipa> you'd need to redo all the work that went into building the current chain
110 2013-10-11 17:28:12 <sipa> yes
111 2013-10-11 17:28:39 <sipa> well, sum(block.difficulty, block in blocks)
112 2013-10-11 17:29:57 <pierce> makes sense.  I had always been under the impression that it was length, which is almost always the same thing, except in the scenario I proposed
113 2013-10-11 17:30:36 <gmaxwell> pierce: it's not the same in any other forking situation right around a retargeting either.
114 2013-10-11 17:31:14 <pierce> gmaxwell: for sure, but at least it's going to be pretty darn close
115 2013-10-11 17:40:44 <skinnkavaj> https://blockchain.info/
116 2013-10-11 17:41:00 <skinnkavaj> Block 262974 was found 12 minutes ago
117 2013-10-11 17:41:05 <skinnkavaj> while block 262975 was found 14 minutes ago
118 2013-10-11 17:41:13 <skinnkavaj> what is wrong?
119 2013-10-11 17:42:07 <helo> blockchain.info
120 2013-10-11 17:42:16 <TheLordOfTime> what helo said
121 2013-10-11 17:42:46 <helo> i saw them the same second on my node ^^
122 2013-10-11 17:43:33 <sipa> saw what?
123 2013-10-11 17:44:23 <TheLordOfTime> sipa: block 262974 and 262975
124 2013-10-11 17:44:30 <TheLordOfTime> blockchain.info displays the found times wrong :p
125 2013-10-11 17:44:51 <sipa> ok?
126 2013-10-11 17:44:51 <TheLordOfTime> sipa: (see skinnkavaj's messages a little further up the screen)
127 2013-10-11 17:45:00 <helo> height=262974  log2_work=72.658804  tx=25227928  date=2013-10-11 17:28:39
128 2013-10-11 17:45:01 <helo> height=262975  log2_work=72.658961  tx=25228170  date=2013-10-11 17:26:42
129 2013-10-11 17:45:03 <TheLordOfTime> (that was to answer you is all :) )
130 2013-10-11 17:45:10 <helo> so the timestamps were out of order :)
131 2013-10-11 17:45:12 <TheLordOfTime> *goes back to poking nginx with a stick*
132 2013-10-11 17:45:34 <helo> blockchain.info orders by timestamp instead of height :/
133 2013-10-11 17:45:48 <sipa> :o
134 2013-10-11 22:33:27 <skinnkavaj> gmaxwell
135 2013-10-11 22:33:35 <skinnkavaj> any idea when block will be produced more stable?
136 2013-10-11 22:33:54 <skinnkavaj> everytime i send some bitcoins and need it fast i have always so unlucky to be in a time when it takes like 50 minutes
137 2013-10-11 22:34:36 <skinnkavaj> right now litecoin is better in this way because of blocks is produced more stable aka every 2 minutes.
138 2013-10-11 22:35:18 <skinnkavaj> i know you have to wait 10 min for litecoin to be as secure, but bitcoin blocks is so unstable thats what annoys m
139 2013-10-11 22:38:08 <gmaxwell> skinnkavaj: Are you confusing 6 confirmations with a single block? In any case it won't ever be any more "stable" the instability is essential for the network to be convergent, otherwise forks that tie would forever stay split.
140 2013-10-11 22:38:35 <skinnkavaj> gmaxwell: but before asic appered
141 2013-10-11 22:38:43 <skinnkavaj> didnt blocks get solved in 10 minutes more likely?
142 2013-10-11 22:39:16 <gmaxwell> TD: In bitcoinj why is the varint encode function called encodeBE?  (VarInt.java)
143 2013-10-11 22:39:28 <gmaxwell> skinnkavaj: I've seen no suggestion that the variance has increased in recent times.
144 2013-10-11 22:40:22 <skinnkavaj> gmaxwell: I have read about it several times in forums that after a while when we have had asic, time for blocks will be 10 minutes more likely.
145 2013-10-11 22:40:23 <gmaxwell> If it has, then it could only be because there is more consolodation in pools and miners now, so outages have a bigger impact. .. but I've not seen anything to suggest that we're seeing that.
146 2013-10-11 22:40:56 <skinnkavaj> gmaxwell: Also, i sent this transaction https://blockchain.info/address/1LH14TGeCaA3wzUiTGs6QQJJFBDT2LdKSv before any new block was solved.. but then asicminer mines a block with only 27 transactions, why didnt include mine?
147 2013-10-11 22:41:00 <gmaxwell> skinnkavaj: you're reading about stuff that is saying that the time will go _UP_ to 10 minutes, the average is below 10 minutes now due to hashrate growth.
148 2013-10-11 22:42:05 <gmaxwell> I think asicminer only includes their own transactions or something like that. They include very few.
149 2013-10-11 22:42:10 <skinnkavaj> gmaxwell: Yeah right it goes faster.. you are right about that, but that dont annoy me.. the thing that annoy me is that somestimes it is 1 minute, another time it is 50 minutes.. and i always manage to send transactions when next block takes 50 minutes... :p
150 2013-10-11 22:42:18 <skinnkavaj> wtf?
151 2013-10-11 22:42:25 <skinnkavaj> why?
152 2013-10-11 22:42:32 <skinnkavaj> should that be allowed?
153 2013-10-11 22:42:37 <skinnkavaj> what if everyone did that..
154 2013-10-11 22:42:56 <skinnkavaj> and why does asicminer do it, they miss out on fee profits.
155 2013-10-11 22:43:25 <gmaxwell> skinnkavaj: It can't be prevented.
156 2013-10-11 22:43:42 <gmaxwell> Don't ask me— but you can observe it for yourself.
157 2013-10-11 22:43:54 <skinnkavaj> here i am waiting for a transaction to confirm
158 2013-10-11 22:44:02 <skinnkavaj> and then asicminer doesnt want to transfer my money
159 2013-10-11 22:44:23 <skinnkavaj> oh wait
160 2013-10-11 22:44:25 <skinnkavaj> look now gmaxwell
161 2013-10-11 22:44:29 <gmaxwell> I'd say that they take ones with high enough fees, except they mine some rather large no-fee transactions too... ::shrugs::
162 2013-10-11 22:44:29 <skinnkavaj> they solved another block
163 2013-10-11 22:44:36 <skinnkavaj> and included 586 transcations
164 2013-10-11 22:44:42 <skinnkavaj> see blockchain.info now
165 2013-10-11 22:44:49 <gmaxwell> that one has a lot of transactions, thats unusual for them.
166 2013-10-11 22:45:52 <gmaxwell> in their last 5 blocks before that one none have more than 100 transactions.
167 2013-10-11 22:46:14 <gmaxwell> maybe they have multiple bitcoinds and one is set to not include transactions. ::shrugs::
168 2013-10-11 22:46:42 <MC1984> surely they can spit out a full solved block fast enough to just be good guys and process what they can
169 2013-10-11 22:47:29 <skinnkavaj> gmaxwell my windows 8 crashed (great microsoft)... what did you write?
170 2013-10-11 22:48:02 <gmaxwell> 15:44 <@gmaxwell> that one has a lot of transactions, thats unusual for them.
171 2013-10-11 22:48:06 <gmaxwell> 15:45 <@gmaxwell> in their last 5 blocks before that one none have more than 100 transactions.
172 2013-10-11 22:48:09 <gmaxwell> 15:46 <@gmaxwell> maybe they have multiple bitcoinds and one is set to not include transactions. ::shrugs::
173 2013-10-11 22:49:00 <skinnkavaj> gmaxwell: i have always liked asicminer, seems like a well run company by a nice guy... if they dont include transactiions that makes me upset, i dont understand why either.. makes no sense
174 2013-10-11 22:52:26 <gmaxwell> skinnkavaj: lots of transactions increases orphaning rate.
175 2013-10-11 22:52:28 <skinnkavaj> gmaxwell: another thing that was asked in a "We are the litcoin team AMA" @ reddit, would it be possible to make bitcoin proof of stake + proof of work?
176 2013-10-11 22:52:31 <gmaxwell> So there is a tradeoff.
177 2013-10-11 22:53:03 <skinnkavaj> gmaxwell: i have never understood why a block gets orphaned, care to explain?
178 2013-10-11 22:53:58 <gmaxwell> generally proof of stake as proposed appears unworkable. It misses the fundimental point of mining which is to put at stake a finite resource (energy) which you lose if you bet it on a consensus which doesn't win.  All PoS systems I've seen have the problem that nothing is at stake.. you can mine in a PoS fork for free, and if that fork loses, no biggie you still have your stake in all other forks.
179 2013-10-11 22:54:19 <gmaxwell> skinnkavaj: because it takes time to forward blocks, because the speed of light and data transmission is finite.. and in that time someone else can find a block.
180 2013-10-11 22:56:49 <skinnkavaj> gmaxwell: isn't that a good reason to have both proof of stake + proof of work? one improves the other one.
181 2013-10-11 22:57:36 <gmaxwell> That isn't clear to me but I've not seen any really good proposals of it, most of them look like they'd just create big centeralization incentives. e.g. you can't participate at all except by joining in some mega pool that holds a lot of stake
182 2013-10-11 22:58:16 <gmaxwell> I have an alternative proposal which is soft forking compatible with bitcoin, and is a much smaller change, but it won't do anything useful until most mining is for fees.
183 2013-10-11 22:59:20 <gmaxwell> Which is to include a 32 bit blockhash in every transaction, and a miner can only collect the fee for that transaction if they are extending a chain which is consistent with that hash. That lets all transacting bitcoin users 'vote with their fees' for the identity of the chain they like.
184 2013-10-11 23:00:18 <gmaxwell> but the consensus there is still proof of work. It just means a malicious deep rewriter can't collect many/any fees.
185 2013-10-11 23:01:03 <skinnkavaj> gmaxwell: sounds like a good idea but when there is only fees rewarding the miners this "problem" will be solved already
186 2013-10-11 23:01:12 <gmaxwell> skinnkavaj: uh? why do you say that?
187 2013-10-11 23:01:37 <gmaxwell> oh you mean miners not including transactions.
188 2013-10-11 23:02:05 <gmaxwell> POS wouldn't help with that in the first place.
189 2013-10-11 23:02:56 <skinnkavaj> gmaxwell: are you a pool operator?
190 2013-10-11 23:03:04 <skinnkavaj> or have you been?
191 2013-10-11 23:03:57 <gmaxwell> skinnkavaj: Why do you ask?
192 2013-10-11 23:04:04 <skinnkavaj> your speciality is mining so i guess so
193 2013-10-11 23:04:12 <gmaxwell> It is? :P
194 2013-10-11 23:04:22 <skinnkavaj> yeah
195 2013-10-11 23:04:47 <gmaxwell> I think its more generally the case that I'm a fucking know it all who thinks hes a specialist at everything.
196 2013-10-11 23:05:28 <gmaxwell> In any case I have never run a public pool, though I've helped with some of them. (and I've been a fairly large solo miner, with >1% of the network hashrate at points)
197 2013-10-11 23:05:33 <skinnkavaj> gmaxwell: is there something that you would like to test on litecoin?
198 2013-10-11 23:05:39 <skinnkavaj> anything?
199 2013-10-11 23:06:17 <skinnkavaj> gmaxwell: i dont understand why you are not a pool operator, seems like you were born to be one.
200 2013-10-11 23:06:20 <gmaxwell> skinnkavaj: I've already told warren what I'd want tested there.
201 2013-10-11 23:06:36 <gmaxwell> The legalities of operating a pool are unclear.
202 2013-10-11 23:06:58 <sipa> skinnkavaj: what you just said sounds a lot like "You seem to know a lot about blood. You should be a doctor."
203 2013-10-11 23:07:01 <gmaxwell> (which was the initial motivation behind eligius having the coinbaser payments, fwiw)
204 2013-10-11 23:07:56 <gmaxwell> I also don't believe centeralized pools should exist.
205 2013-10-11 23:08:18 <skinnkavaj> gmaxwell: but they do is it a problem?
206 2013-10-11 23:09:14 <skinnkavaj> and what would you like to have tested on litecoin? please tell. warren is really cool, he teached me to use GPG and i dont know even know the guy before he got into development.
207 2013-10-11 23:09:20 <gmaxwell> Well, "problem" is hard to define. Right now the security assumptions of bitcoin have clearly been undermined by the existance of very large pools, but it hasn't seemed to have many pratical consequences.
208 2013-10-11 23:10:35 <gmaxwell> Right now you could coerce two or three guys and then successfully reverse transactions a dozen blocks deep. This is clearly not how bitcoin was intended to work. But the motivations for doing so are apparently few and far between.
209 2013-10-11 23:10:47 <gmaxwell> Fortunately security tends to work better in practice than in theory.
210 2013-10-11 23:10:49 <skinnkavaj> sipa: just like you are born to organize statics
211 2013-10-11 23:11:04 <sipa> statics? :o
212 2013-10-11 23:11:08 <gmaxwell> skinnkavaj: you do realize the at this point sipa has written most of bitcoind right?
213 2013-10-11 23:11:18 <gmaxwell> And you credit him for being a hashrate bookie? :P
214 2013-10-11 23:12:02 <sipa> gmaxwell: don't exaggerate :)
215 2013-10-11 23:12:30 <gmaxwell> sipa: I thought we looked at it before and it was close? I'm interpolating not exaggerating. :P
216 2013-10-11 23:12:33 <skinnkavaj> gmaxwell: please tell me what bitcoind is, isn't that the software satoshi developed?
217 2013-10-11 23:12:57 <skinnkavaj> what is the difference between bitcoin and bitcoind?
218 2013-10-11 23:13:11 <gmaxwell> skinnkavaj: I was saying bitcoind there in order to exclude the gui.
219 2013-10-11 23:13:14 <sipa> bitcoin is a currency, a system, a network, a protocol
220 2013-10-11 23:13:22 <gmaxwell> (and the currency, netowkr, etc.)
221 2013-10-11 23:13:30 <sipa> bitcoind is the reference client binary name, without the gui
222 2013-10-11 23:13:37 <sipa> bitcoin-qt is the reference client binary name, with the gui
223 2013-10-11 23:13:57 <skinnkavaj> so bitcoind didnt exist 2008 when satoshi realsed bitcoin?
224 2013-10-11 23:14:26 <sipa> the actual answer to your question is "indeed not", for multiple reasons
225 2013-10-11 23:14:40 <sipa> but for all intents and purposes, the answer you want is likely "yes it did"
226 2013-10-11 23:14:59 <sipa> satoshi release bitcoin (a windows-only, gui-only, wx-based) tool in early 2009
227 2013-10-11 23:15:21 <sipa> bitcoind was later added, as a new binary built from the same source code, after removing the GUI
228 2013-10-11 23:15:38 <sipa> when wx was replaced by qt as GUI tooltik, the name of the program changed to bitcoin-qt
229 2013-10-11 23:16:12 <sipa> but what gmaxwell was referring to (i don't think it's true), is that for the core code of the reference client, i by now have (re)written most of its code
230 2013-10-11 23:16:22 <MC1984> was it really windows only at the start? damn
231 2013-10-11 23:16:48 <skinnkavaj> do you think satoshi is overrated or is he a genius? because some people say he is brillant and some people say it was just crap when bitcoin was released.. and the fact that is was windows only
232 2013-10-11 23:16:59 <gmaxwell> (for i in *.cpp ; do git blame $i ; done) | cut -d'(' -f2 | cut -d' ' -f1 | tr '[:upper:]' '[:lower:]' | sort | uniq -c | sort -n | awk '{aa+=$1; print $0} END {print aa}'    ... well it's 31% but it's more than the next two people combined.
233 2013-10-11 23:17:21 <gmaxwell> and close to the next three people combined.
234 2013-10-11 23:17:27 <sipa> moving other people's code around helps :D
235 2013-10-11 23:17:35 <gmaxwell> Indeed. :P
236 2013-10-11 23:17:53 <MC1984> satoshi had vision, but he did put together a lot of stuff otehr people had come up with
237 2013-10-11 23:17:58 <MC1984> he didnt invent it all
238 2013-10-11 23:18:19 <MC1984> some of the concepts in bitcoin were decades old i think
239 2013-10-11 23:18:56 <sipa> the overal design is genius imho, even if he didn't foresee some of the economics that appeared later (pooled mining, ...)
240 2013-10-11 23:19:06 <skinnkavaj> soon fbi raid satoshis home in japan because they found out an email adress "hey look at this cool program bitcoin" and another "wanna create a new currency? my private mail is satoshi@
241 2013-10-11 23:19:10 <gmaxwell> skinnkavaj: I don't know why you'd single out windows only there. Bitcoin was so many novel things at once, I don't think you could have asked it to be much more... and while it was initially windows only, the software was at least portable (unlike virtually all other windows only software)
242 2013-10-11 23:19:14 <sipa> and he is also clearly a very good programmer, though not really a good software designer
243 2013-10-11 23:19:33 <sipa> and the coding style was ancient :)
244 2013-10-11 23:20:18 <gmaxwell> well, it was pre-RAII style, but it was way more modern than it would have been if I'd written it.
245 2013-10-11 23:20:21 <MC1984> how much can you really tell of a person from the code they write
246 2013-10-11 23:20:22 <gmaxwell> :P
247 2013-10-11 23:20:54 <skinnkavaj> you can tell gmaxwell was destined to become a pool operator but decided to help other instead
248 2013-10-11 23:21:06 <gmaxwell> I mean, ... it didn't write its own string class so even in C++ terms it wasn't _that_ ancient. :)
249 2013-10-11 23:21:06 <sipa> gmaxwell: not sure you'd come up with a main.cpp that did all validation logic, all mining, all wallet management, and did direct calls to the GUI :p
250 2013-10-11 23:22:03 <Luke-Jr> gmaxwell: eh, it wasn't that portable from what I can tell :P
251 2013-10-11 23:22:06 <skinnkavaj> sipa: everyone here say blockchain.info sucks, you with all your statics, cant you setup a better alternative?
252 2013-10-11 23:22:09 <gmaxwell> Well. okay. Yea I wouldn't have done that, though I also wouldn't have done it the way I'd do it now. (I'd make validation logic standalone, a seperate library, and formally verify as much of it as possible)
253 2013-10-11 23:22:26 <sipa> skinnkavaj: dude
254 2013-10-11 23:22:43 <sipa> skinnkavaj: i wrote some perl scripts to generate some graphs 2 years ago
255 2013-10-11 23:22:48 <sipa> and the site is still running
256 2013-10-11 23:23:00 <skinnkavaj> sipa: exactly how hard was it for you :D
257 2013-10-11 23:23:05 <skinnkavaj> you get my point
258 2013-10-11 23:23:08 <Luke-Jr> skinnkavaj: it shouldn't be a website at all :P
259 2013-10-11 23:23:13 <Luke-Jr> should be built into Bitcoin-Qt
260 2013-10-11 23:23:15 <sipa> ok, some of the math to compute it efficiently was interesting
261 2013-10-11 23:23:16 <gmaxwell> sipa: no no, but you have the power of THE CALCULUS.  :P
262 2013-10-11 23:24:01 <skinnkavaj> Luke-Jr: that would be so cool! i would be like "hey check out this statics software" of the bitcoin program.. now i don even want that slow program on my computer because it doesnt have any cool functions.. no offense.
263 2013-10-11 23:24:52 <sipa> but i've probably spent like 100 times more work on the client's code than on that website :)
264 2013-10-11 23:25:15 <skinnkavaj> bitcoin program on windows is really ugly
265 2013-10-11 23:25:31 <skinnkavaj> and no cool functions like statics
266 2013-10-11 23:25:37 <sipa> (statistics, you mean?)
267 2013-10-11 23:25:47 <sipa> i don't care about the GUI, tbh
268 2013-10-11 23:25:49 <skinnkavaj> oh yes lol sorry
269 2013-10-11 23:26:02 <skinnkavaj> sipa: ofc you dont care about the GUI
270 2013-10-11 23:26:04 <skinnkavaj> no nerd does
271 2013-10-11 23:26:10 <sipa> hahaha
272 2013-10-11 23:26:12 <skinnkavaj> but the averge joe does
273 2013-10-11 23:26:21 <skinnkavaj> the truth:)
274 2013-10-11 23:26:23 <sipa> sure
275 2013-10-11 23:26:52 <skinnkavaj> maybe the average joe doesnt have to touch bitcoin, but isnt it good for network propogation?
276 2013-10-11 23:27:02 <skinnkavaj> isnt it good to just run the program?
277 2013-10-11 23:27:08 <skinnkavaj> even if you dont mine bro.
278 2013-10-11 23:27:22 <sipa> 
ACTION doesn't mine bro

279 2013-10-11 23:28:18 <gmaxwell> Anyone going to be at linuxcon europe / linux kernel summit in a couple weeks in Edinburgh?  I'm speaking (not about bitcoin stuff) at the gstreamer summit there.
280 2013-10-11 23:29:40 <sipa> hmmno
281 2013-10-11 23:29:48 <sipa> nobody
282 2013-10-11 23:29:49 <sipa> sorry
283 2013-10-11 23:31:16 <skinnkavaj> gmaxwell: what are you speaking about? dont say its a suprise i kill you
284 2013-10-11 23:31:52 <gmaxwell> skinnkavaj: I'm on the schedule, you could look it up. (though, uh, maybe I should send them an abstract)
285 2013-10-11 23:32:33 <skinnkavaj> sipa: you and Luke-Jr go build blockchain.info statistics into bitcoin gui
286 2013-10-11 23:34:44 <gmaxwell> skinnkavaj: why don't you do it?
287 2013-10-11 23:34:56 <gmaxwell> (other than that the Bc.i stats are pretty lame)
288 2013-10-11 23:35:36 <skinnkavaj> gmaxwell: i already have it written and done its just that i would like sipa and Luke-Jr to get some credits ;)
289 2013-10-11 23:38:42 <skinnkavaj> just being cocky, but really you guys should
290 2013-10-11 23:38:59 <skinnkavaj> gmaxwell: what would you prefer to be in bitcoin 1.0?
291 2013-10-11 23:39:04 <skinnkavaj> when will it be "finished"?
292 2013-10-11 23:45:31 <MC1984> 0.9 rolls over into 0.10
293 2013-10-11 23:45:34 <MC1984> trollface.png
294 2013-10-11 23:46:32 <gmaxwell> MC1984: what, you thought it was otherwise?
295 2013-10-11 23:46:54 <sipa> 0.9, 0.9.9, 0.9.9.9, ...
296 2013-10-11 23:46:55 <MC1984> never had a doubt
297 2013-10-11 23:47:36 <sipa> it converges to (e^pi - pi)/20
298 2013-10-11 23:47:48 <gribble> 0.999954998959
299 2013-10-11 23:47:48 <sipa> ;;calc (e**pi - pi)/20
300 2013-10-11 23:48:01 <gmaxwell> oh thats a tricky number.
301 2013-10-11 23:48:37 <skinnkavaj> he who have written gribble, nanotube, is he spanish? does anyone know? he is so hard to reach.
302 2013-10-11 23:49:20 <warren> we need more bits for the block version number to accommodate that much innovation
303 2013-10-11 23:49:33 <skinnkavaj> warren :D
304 2013-10-11 23:53:05 <skinnkavaj> gmaxwell: has there been any big discussion on what to include in bitcoin 1.0?
305 2013-10-11 23:53:15 <skinnkavaj> or are we too far away?
306 2013-10-11 23:53:28 <MC1984> another year at least imo
307 2013-10-11 23:53:34 <skinnkavaj> another year!?
308 2013-10-11 23:53:36 <skinnkavaj> crazy
309 2013-10-11 23:53:40 <warren> MC1984: hey I actually did that thing you asked for
310 2013-10-11 23:53:43 <MC1984> whats the rush
311 2013-10-11 23:53:52 <gmaxwell> skinnkavaj: it's not anywhere near at all.
312 2013-10-11 23:53:52 <skinnkavaj> i was thinking another decade or so
313 2013-10-11 23:53:56 <gmaxwell> skinnkavaj: right.
314 2013-10-11 23:54:09 <MC1984> warren ??
315 2013-10-11 23:54:12 <gmaxwell> anyways maybe something gets called 1.0 before then, marketing is a funny thing.
316 2013-10-11 23:54:24 <gmaxwell> But it will not be finished anytime soon.
317 2013-10-11 23:54:36 <MC1984> surely it will never be finished
318 2013-10-11 23:54:44 <MC1984> man i remember when vlc hit 1.0
319 2013-10-11 23:55:00 <warren> MC1984: http://sourceforge.net/mailarchive/message.php?msg_id=31503819
320 2013-10-11 23:55:18 <gmaxwell> skinnkavaj: there are constantly new things being invented which could radically improve bitcoin once implemented. Most of them will never find a place in bitcoin, but some of them may.
321 2013-10-11 23:55:36 <skinnkavaj> i bet it will be hard to have everyone agree that 1.0 is it, thats why i think it will never be 1.0
322 2013-10-11 23:56:02 <skinnkavaj> there will always be people who want to add more or change something
323 2013-10-11 23:56:13 <MC1984> oh thanks man
324 2013-10-11 23:56:55 <MC1984> 1.0 implies some sort of warranty
325 2013-10-11 23:57:01 <skinnkavaj> gmaxwell: looking forward to sunnykings next coin
326 2013-10-11 23:57:05 <MC1984> just keep adding nines forever
327 2013-10-11 23:57:07 <Luke-Jr> MC1984: nonsense
328 2013-10-11 23:57:09 <skinnkavaj> he seems to like to innovate
329 2013-10-11 23:57:29 <skinnkavaj> who doesnt
330 2013-10-11 23:57:52 <MC1984> 1.0 is analagous to initial release
331 2013-10-11 23:58:13 <warren> Then (e**pi - pi)/20 sounds like a good idea.
332 2013-10-11 23:58:42 <skinnkavaj> 1.0 will never be agreed on when money is involved, vlc is totally different since it does not involve money
333 2013-10-11 23:59:01 <skinnkavaj> too many bright minds wanna have their say
334 2013-10-11 23:59:06 <MC1984> hmm when did bitcoin get to 22mb
335 2013-10-11 23:59:18 <skinnkavaj> thats why bitcoin always will be GOLD =D
336 2013-10-11 23:59:21 <MC1984> google kept gmail in beta for a long time for this reason
337 2013-10-11 23:59:58 <skinnkavaj> MC1984: but gmail is not open source