1 2012-04-02 00:35:59 <Graet> http://blockchain.info/ here we go again? i hope not
  2 2012-04-02 00:36:46 <forsetifox> Transactions aren't put in orphans are they?
  3 2012-04-02 00:37:18 <forsetifox> This web page shows they are but that sounds like a very bad idea to me.
  4 2012-04-02 00:37:31 <doublec> Graet: ouch, would that be two of your recent blocks orphaned?
  5 2012-04-02 00:39:23 <gmaxwell> forsetifox: What you're saying makes no sense at all.
  6 2012-04-02 00:40:04 <gmaxwell> You don't know your block is going to be orphaned when you creat it. An orphan is just a block that didn't end up becoming part of the longest chain, because some other block at the same height did again.
  7 2012-04-02 00:40:25 <gmaxwell> Graet: what code are you running? 0.6 or one of luke's backports? or?
  8 2012-04-02 00:40:33 <forsetifox> Are transactions normally put in orphaned blocks?
  9 2012-04-02 00:40:58 <copumpkin> orphaned blocks are normal blocks
 10 2012-04-02 00:41:10 <copumpkin> orphaning is a property of how the network treats the block, not the block itself
 11 2012-04-02 00:42:17 <gmaxwell> oh this looks boring
 12 2012-04-02 00:42:57 <copumpkin> Graet: here we go again about what?
 13 2012-04-02 00:43:12 <copumpkin> oh, another duplicate?
 14 2012-04-02 00:43:19 <gmaxwell> ozcoin had a block race, but it appears to have won.
 15 2012-04-02 00:43:28 <gmaxwell> (lets see if a fork forms...)
 16 2012-04-02 00:44:36 <doublec> copumpkin: I assumed it was "here we go again, ozcoins last block was orphaned, looks like this one will be too"
 17 2012-04-02 00:44:59 <copumpkin> oh okay
 18 2012-04-02 00:45:50 <gmaxwell> oh.. interesting!
 19 2012-04-02 00:45:56 <gmaxwell> both side have been extended.
 20 2012-04-02 00:46:38 <doublec> that makes for a confusing block list
 21 2012-04-02 00:46:51 <luke-jr> O.o
 22 2012-04-02 00:47:44 <da2ce7> lawl: https://twitter.com/#!/da2ce7/status/186645872933736449
 23 2012-04-02 00:47:48 <Graet> gmaxwell, look at blockchain.info now
 24 2012-04-02 00:47:53 <Graet> there are dupes everywhere
 25 2012-04-02 00:48:09 <gmaxwell> Again, MM is involved.
 26 2012-04-02 00:48:24 <gmaxwell> :-/
 27 2012-04-02 00:48:35 <Graet> http://blockchain.info/orphaned-blocks
 28 2012-04-02 00:48:38 <Graet> this is bullshit
 29 2012-04-02 00:48:39 <gmaxwell> Graet: what code are you running?
 30 2012-04-02 00:49:18 <doublec> so a 3 block fork again?
 31 2012-04-02 00:49:18 <Graet> checking
 32 2012-04-02 00:49:26 <luke-jr> someone wake up p2k :p
 33 2012-04-02 00:49:32 <gmaxwell> And again starting with Graet.
 34 2012-04-02 00:49:41 <gmaxwell> And again with MM extending Graet's earlier chain.
 35 2012-04-02 00:49:56 <luke-jr> is there anything wrong with Graet's block, per human analysis?
 36 2012-04-02 00:50:07 <doublec> is MM connected directly to Graet and getting their blocks first to build on
 37 2012-04-02 00:50:22 <luke-jr> doublec: Graet's finding the blocks first
 38 2012-04-02 00:50:35 <doublec> luke-jr: right, then MM builds on that
 39 2012-04-02 00:50:40 <gmaxwell> 04/02/12 02:22:15 ERROR: ConnectInputs() : 4005d6bea3 P2SH VerifySignature failed
 40 2012-04-02 00:50:43 <gmaxwell> 04/02/12 02:22:15 InvalidChainFound: invalid block=00000000000001bd778c  height=173957  work=281775367106871011598
 41 2012-04-02 00:50:53 <gmaxwell> Ahem.
 42 2012-04-02 00:51:09 <copumpkin> hmm
 43 2012-04-02 00:52:00 <luke-jr> http://blockchain.info/tx-index/3618498/4005d6bea3a93fb72f006d23e2685b85069d270cb57d15f0c057ef2d5e3f78d2
 44 2012-04-02 00:52:01 <luke-jr> wtf?
 45 2012-04-02 00:52:06 <luke-jr> input: OP_1 029c7187ecea7f09146820075c3a8de5d33ffbc293b63228ea1667c8d3796aff3f OP_1 OP_CHECKMULTISIG
 46 2012-04-02 00:52:22 <gmaxwell> Graet: I bet you if you keep mining those obviously broken blocks you can get people to chip in to replace the lost income since every time you do you manage to make a couple other miners extend your crap chain.
 47 2012-04-02 00:53:14 <nanotube> haha
 48 2012-04-02 00:53:21 <Graet> ohy ffrs
 49 2012-04-02 00:53:49 <forsetifox> Is that Austrailian for "Oh, $hit"?
 50 2012-04-02 00:53:54 <luke-jr> is e8c300c87986efa84c37c0519929019ef86eb5b4 the hash160 of 1?
 51 2012-04-02 00:54:01 <copumpkin> forsetifox: I think so
 52 2012-04-02 00:56:52 <gmaxwell> [Tycho]: you have non-p2sh nodes running.
 53 2012-04-02 00:57:02 <gmaxwell> ?
 54 2012-04-02 00:57:05 <luke-jr> gmaxwell: are they really Deepbit?
 55 2012-04-02 00:57:10 <gmaxwell> Yea, fair point.
 56 2012-04-02 00:57:16 <luke-jr> I think MM is still relaying via Deepbit sometimes
 57 2012-04-02 00:58:33 <gmaxwell> Well, these blocks don't look like MM they have txns (obviously since they're carring the bad one)
 58 2012-04-02 00:58:45 <gmaxwell> But yea, probably not deepbit.
 59 2012-04-02 00:58:52 <luke-jr> oh, true
 60 2012-04-02 00:58:56 <luke-jr> actaulyl
 61 2012-04-02 00:59:09 <luke-jr> the fact that deepbit relayed the blocks means deepbit is broken
 62 2012-04-02 00:59:11 <luke-jr> I think
 63 2012-04-02 00:59:35 <luke-jr> sounding like *some* backport has issues?
 64 2012-04-02 00:59:38 <gmaxwell> luke-jr: he may have just updated his relay nodes.
 65 2012-04-02 00:59:58 <gmaxwell> er NOT upgraded.
 66 2012-04-02 01:00:04 <gmaxwell> sorry. I'm tired apparently.
 67 2012-04-02 01:00:07 <luke-jr> maybe
 68 2012-04-02 01:00:16 <etotheipi_> luke-jr, isn't that script failing due to the OP_CHECKMULTISIG bug?
 69 2012-04-02 01:00:25 <etotheipi_> it needs an extra op-code at the beginning
 70 2012-04-02 01:00:45 <luke-jr> etotheipi_: it has one?
 71 2012-04-02 01:01:00 <etotheipi_> is the OP_1 PUBKEY OP_1 declaring 1-of-1
 72 2012-04-02 01:01:13 <etotheipi_> *isn't *?
 73 2012-04-02 01:01:54 <luke-jr> hmm
 74 2012-04-02 01:02:01 <luke-jr> yeah, how is that getting into *any* blocks? ;/
 75 2012-04-02 01:02:21 <luke-jr> 0day vuln in BIP16 backport?
 76 2012-04-02 01:02:22 <copumpkin> so Graet occasionally spits out a bad block?
 77 2012-04-02 01:02:56 <gmaxwell> copumpkin: no, all of his blocks are bad.
 78 2012-04-02 01:03:04 <copumpkin> how has nobody noticed that?
 79 2012-04-02 01:03:07 <gmaxwell> Because he keeps trying to reinsert http://blockchain.info/tx-index/3618498/4005d6bea3a93fb72f006d23e2685b85069d270cb57d15f0c057ef2d5e3f78d2
 80 2012-04-02 01:03:11 <gmaxwell> copumpkin: because he's only had two
 81 2012-04-02 01:03:14 <copumpkin> oh
 82 2012-04-02 01:03:20 <gmaxwell> And I looked at the first one and missed the transaction.
 83 2012-04-02 01:03:37 <etotheipi_> luke-jr, nm -- it's on the signature list you need the extra OP_0
 84 2012-04-02 01:03:40 <copumpkin> but blockchain.info says it's okay! it's even in green
 85 2012-04-02 01:03:40 <gmaxwell> Unfortunately he's not responded about _what_ code he's actually running.
 86 2012-04-02 01:03:59 <luke-jr> etotheipi_: that IS the scriptSig
 87 2012-04-02 01:04:20 <gmaxwell> copumpkin: blockchain.info is only very mildly smart.
 88 2012-04-02 01:04:37 <copumpkin> ah, so it doesn't have the full set of checks that it should?
 89 2012-04-02 01:05:24 <gmaxwell> I expect it shows everything as green.
 90 2012-04-02 01:05:51 <gmaxwell> I like how it says it has 5 confirmations.
 91 2012-04-02 01:06:14 <copumpkin> there's a list of things it rejected, though
 92 2012-04-02 01:06:15 <copumpkin> http://blockchain.info/rejected
 93 2012-04-02 01:06:26 <copumpkin> so I'd assume it isn't just accepting everything
 94 2012-04-02 01:06:54 <gmaxwell> copumpkin: all due to ConnectInputs failed .
 95 2012-04-02 01:07:14 <copumpkin> yeah
 96 2012-04-02 01:07:36 <gmaxwell> Graet: it would be really great if you'd tell us exactly what code you're running.
 97 2012-04-02 01:08:21 <gmaxwell> fwiw,
 98 2012-04-02 01:08:23 <gmaxwell> 04/01/12 14:56:41  195.72.201.1:8333  got inventory: tx 4005d6bea3a93fb72f00  new
 99 2012-04-02 01:08:26 <gmaxwell> 04/01/12 14:56:41 askfor tx 4005d6bea3a93fb72f00   0
100 2012-04-02 01:08:28 <gmaxwell> 04/01/12 14:56:41  195.72.201.1:8333 sending getdata: tx 4005d6bea3a93fb72f00
101 2012-04-02 01:08:31 <gmaxwell> 04/01/12 14:56:41 04/01/12 14:56:41 sending: getdata (37 bytes)
102 2012-04-02 01:08:34 <gmaxwell> 04/01/12 14:56:42 04/01/12 14:56:42 received: tx (123 bytes)
103 2012-04-02 01:08:36 <gmaxwell> 04/01/12 14:56:42 ERROR: AcceptToMemoryPool() : nonstandard transaction input
104 2012-04-02 01:08:40 <gmaxwell> is the first time I saw that txn.
105 2012-04-02 01:08:43 <copumpkin> ;;seen roconnor
106 2012-04-02 01:08:43 <gribble> roconnor was last seen in #bitcoin-dev 2 weeks, 5 days, 8 hours, 19 minutes, and 35 seconds ago: <roconnor> gavinandresen: given it a very quick glance, it all seems plausible.  I've been trying to figure how to anonymize transactions using homomorphic encryption for a little while.
107 2012-04-02 01:08:59 <luke-jr> I don't see how4005d6bea3a93fb72f006d23e2685b85069d270cb57d15f0c057ef2d5e3f78d2 could ever be valid
108 2012-04-02 01:09:48 <Graet> gmaxwell, as soon as i can i will
109 2012-04-02 01:09:57 <Graet> shutdown mining for now
110 2012-04-02 01:09:59 <Graet> :(
111 2012-04-02 01:10:38 <luke-jr> :/
112 2012-04-02 01:10:42 <gmaxwell> unfortunately I don't have my .19 node running at the moment, so I can't see if it'll accept it.
113 2012-04-02 01:11:21 <luke-jr> can we get the raw txn data anywhere?
114 2012-04-02 01:11:55 <bitfoo> this is the best I could find: https://blockchain.info/tx-index/3618498/4005d6bea3a93fb72f006d23e2685b85069d270cb57d15f0c057ef2d5e3f78d2?show_adv=true
115 2012-04-02 01:12:43 <luke-jr> we should get gavin or tcatm to send out an alert about now imo
116 2012-04-02 01:12:59 <gmaxwell> huh?
117 2012-04-02 01:13:00 <gmaxwell> No way.
118 2012-04-02 01:13:03 <luke-jr> the risk of high-confirmation double-spends seems high
119 2012-04-02 01:13:09 <gmaxwell> What would the alert say?
120 2012-04-02 01:13:22 <luke-jr> not sure :/
121 2012-04-02 01:13:27 <gmaxwell> I don't see any reason to think that the risk of high confirmation double-spends is high.
122 2012-04-02 01:13:35 <luke-jr> but someone could easily use this situation to get a double-spend in
123 2012-04-02 01:13:52 <luke-jr> gmaxwell: spend the output of the invalid txn
124 2012-04-02 01:13:57 <doublec> gmaxwell: I have an old node running that asked for that transaction but didn't add to the memory pool due to not enough fees
125 2012-04-02 01:14:09 <doublec> if I understand the debug.log correctly
126 2012-04-02 01:14:47 <gmaxwell> I just started up .19 but it'll take a bit to sync up the chain.
127 2012-04-02 01:14:58 <gmaxwell> (up to 100k now)
128 2012-04-02 01:15:27 <gmaxwell> All my nodes that heard it rejected it.. but they're all ~0.6 vintage git trunk of some version or another.
129 2012-04-02 01:15:37 <gmaxwell> luke-jr: check your logs?
130 2012-04-02 01:15:48 <luke-jr> gmaxwell: no sign of it in Eligius's logs at all
131 2012-04-02 01:15:51 <doublec> gmaxwell: http://pastebin.com/1RbeYVZJ
132 2012-04-02 01:16:15 <gmaxwell> doublec: indeed, what version is that node?
133 2012-04-02 01:16:16 <doublec> gmaxwell: running bitcoin version 31900
134 2012-04-02 01:16:44 <gmaxwell> okay thats not helpful, in 319 the fees check is right after connect.
135 2012-04-02 01:17:32 <gmaxwell> oh actually no..
136 2012-04-02 01:20:44 <gmaxwell> so.. yea.. that error is after signature validation.. so...
137 2012-04-02 01:22:49 <gmaxwell> I've compiled 0.3.19 without the fees. Lets see if it takes it.. (if it hears it. :( )
138 2012-04-02 01:23:05 <luke-jr> gmaxwell: you're probably onto something with the voting-but-not-supporting thing actually
139 2012-04-02 01:23:34 <luke-jr> no mentions of coinbaseflags or coinbaseaux in ecoinpool src
140 2012-04-02 01:23:57 <luke-jr> which means p2k would have had to configure the flag in it independently from bitcoind
141 2012-04-02 01:42:19 <gmaxwell> okay, Just an update.
142 2012-04-02 01:42:28 <gmaxwell> My hypothesis is currently:
143 2012-04-02 01:42:53 <gmaxwell> blockchain.info is being "smart" with p2sh transactions: it's decoding the inner script and displaying that instead of the real script.
144 2012-04-02 01:43:28 <gmaxwell> ... And, with that assumption: the TXN in question is a really a p2sh spend although it doesn't look like one on blockchain.info.
145 2012-04-02 01:43:58 <gmaxwell> Which is consistent with how my nodes rejected it: They say it fails full validation but it passes non-p2sh validation. Also consistent with it passing validation for doublec's 0.3.19.
146 2012-04-02 01:44:33 <gmaxwell> The reason it fails is because the inner script (that we're seeing on blockchain.info) can't possibly be right. And couldn't possibly pass our validation code.
147 2012-04-02 01:45:47 <gmaxwell> The only open question is to why ozcoin would accept that transaction. My theory is that ozcoin isn't really validating p2sh for some reason. Perhaps it just missing the code. As luke points out the pool software it uses couldn't pass through the flag. So I can only assume the flag is hardcoded.
148 2012-04-02 01:46:15 <Graet> yep thats why i'm trying to get onbto p2k but its night in eu
149 2012-04-02 01:46:36 <gmaxwell> Graet: is there any chance you had updated code but hadn't restarted your node?
150 2012-04-02 01:46:57 <Graet> p2k did restart when he patched
151 2012-04-02 01:47:12 <Graet> i remember announcing it in the irc chan
152 2012-04-02 01:47:21 <gmaxwell> Perhaps some chance he started the wrong copy up? (e.g. do you have multiple versions laying around?)
153 2012-04-02 01:47:59 <Graet> not on the mining nodes
154 2012-04-02 01:48:17 <Graet> we just have an empty bitcoiond just for mining
155 2012-04-02 01:48:29 <Graet> empty as in no coins ever
156 2012-04-02 01:49:43 <gmaxwell> the fact that you're the only "p2sh" node that extended that fork at least suggests that it's some problem specific to you. (whew)
157 2012-04-02 01:49:50 <Graet> yes
158 2012-04-02 01:50:04 <Graet> thast why i turned our nodes off until its resolved
159 2012-04-02 01:51:51 <gribble> New news from bitcoinrss: timmerov opened issue 1024 on bitcoin/bitcoin <https://github.com/bitcoin/bitcoin/issues/1024>
160 2012-04-02 01:57:11 <nanotube> ooh, round issue number!
161 2012-04-02 01:58:07 <gmaxwell> nanotube: geek  (I thought that too)
162 2012-04-02 02:01:22 <nanotube> lol
163 2012-04-02 02:01:54 <nanotube> i think it's fair to say that P(geek|hanging out on #bitcoin-dev) >> P(geek) :)
164 2012-04-02 02:02:12 <doublec> Graet: do you use that erlang pool software?
165 2012-04-02 02:02:18 <Graet> yes
166 2012-04-02 02:02:22 <Graet> ecoinpool
167 2012-04-02 02:02:27 <doublec> does that use getmemorypool and create it's own blocks?
168 2012-04-02 02:02:36 <Graet> yes
169 2012-04-02 02:04:06 <doublec> I don't see anything in their git repository about bip 16 support
170 2012-04-02 02:07:40 <Graet> p2k patched our bitcoinds, he has been busy on another (paying) project so he may not have updated yet
171 2012-04-02 02:09:50 <gmaxwell> doublec: thus my guess above about the flag being hardcoded
172 2012-04-02 02:10:03 <doublec> right
173 2012-04-02 02:10:34 <gmaxwell> Graet: back when there was that issue with nodes getting orphaned a few weeks ago due to the p2sh taking effect too early for them, were you one of the impacted parties?
174 2012-04-02 02:10:45 <Graet> no
175 2012-04-02 02:11:11 <Graet> but we changed the deadline thing before that
176 2012-04-02 02:11:30 <gmaxwell> Hm, okay. (my thinking there was if you were perhaps it was solved by turning it off completely, or incorrectly updating the deadline)
177 2012-04-02 02:12:15 <Graet> someone thought one was us, but it was a mistake
178 2012-04-02 02:12:24 <Graet> we made it thgru that one opk ;)
179 2012-04-02 03:26:44 <DBordello> Something appears to be up, multiple people are reporting bitcoin-qt not syncing this last block
180 2012-04-02 03:27:08 <SomeoneWeird> probably because of the retarded split
181 2012-04-02 03:27:30 <denisx> hi
182 2012-04-02 03:27:49 <denisx> my bitcoind also not accepting this last block #173965
183 2012-04-02 03:28:00 <DBordello> ooh, did BIP XX just go live?
184 2012-04-02 03:28:14 <da2ce7> DBordello: yep.
185 2012-04-02 03:28:15 <copumpkin> I just installed bitcoin-qt on a different computer and it's been stuck on a block from 241 days for ages now. Is the progress reporting messed up or is something wrong?
186 2012-04-02 03:28:30 <copumpkin> (syncing for the first time)
187 2012-04-02 03:29:49 <gmaxwell> copumpkin: whats the current height? (mouseover)
188 2012-04-02 03:30:04 <copumpkin> 139647
189 2012-04-02 03:30:09 <copumpkin> it's been there for ages
190 2012-04-02 03:31:15 <gmaxwell> copumpkin: so, if the peer you were pulling from stops feeding (for whatever reason) it will get stuck until the next block on the network.
191 2012-04-02 03:31:22 <gmaxwell> And there hasn't been a block for an hour.
192 2012-04-02 03:31:27 <copumpkin> ah, fair enough
193 2012-04-02 03:31:39 <copumpkin> gotta love these dry spells
194 2012-04-02 03:32:01 <copumpkin> denisx: are you sure that isn't just bitcoin-qt reporting it's out of sync after a while, because there have been no blocks recently?
195 2012-04-02 03:32:03 <copumpkin> I've seen mine do that
196 2012-04-02 03:32:33 <twobitcoins> Block 173965 contains a funny transaction 4005d6bea3a93fb72f006d23e2685b85069d270cb57d15f0c057ef2d5e3f78d2 that keeps ending up in orphaned blocks.
197 2012-04-02 03:32:34 <gmaxwell> denisx: why do you say its not accepting 173965?
198 2012-04-02 03:32:42 <denisx> I see in debug.log that other peers have 173965
199 2012-04-02 03:32:48 <twobitcoins> I don't quite understand what is up with that transaction and why it is causing forking though.
200 2012-04-02 03:32:58 <gmaxwell> oh because it isn't.
201 2012-04-02 03:33:00 <gmaxwell> Thats fine.
202 2012-04-02 03:33:12 <gmaxwell> 173965 will get orphaned and life will go on.
203 2012-04-02 03:33:53 <gmaxwell> Block from 50 BTC, I mentioned them in my eulogy.
204 2012-04-02 03:34:10 <denisx> gmaxwell: I hope you'r right
205 2012-04-02 03:34:23 <copumpkin> that's how it works, right?
206 2012-04-02 03:34:42 <gmaxwell> denisx: I've been right the ~5 other times its happened today.
207 2012-04-02 03:35:13 <gmaxwell> denisx: the other peers that have 173965 are just nodes that aren't upgraded to enforce BIP16.
208 2012-04-02 03:35:37 <denisx> which versions have BIP16?
209 2012-04-02 03:35:59 <gmaxwell> 0.6 and older versions that have the backports appilied.
210 2012-04-02 03:36:03 <gmaxwell> applied too.
211 2012-04-02 03:36:21 <DBordello> wait, shouldn't it be impossible for the BIP 16 chain to get BEHIND the non BIP16 chain?
212 2012-04-02 03:36:36 <DBordello> won't older clients reject BIP16 transactions?
213 2012-04-02 03:36:43 <denisx> this is a hard cut
214 2012-04-02 03:36:48 <gmaxwell> DBordello: You ... sound profoundly confused.
215 2012-04-02 03:36:50 <denisx> a lot of pools will maybe die
216 2012-04-02 03:36:53 <gmaxwell> There aren't two chains.
217 2012-04-02 03:37:00 <DBordello> gmaxwell, I am, I am reading the wiki :)
218 2012-04-02 03:37:00 <gmaxwell> denisx: almost all of them were upgraded.
219 2012-04-02 03:37:14 <DBordello> gmaxwell, well, we can split
220 2012-04-02 03:37:17 <gmaxwell> The named pools that didn't appear upgraded that I could find were Donate@Home, BTC Warp, Arsbitcoin, 50BTC, nmcbit, and Polmine
221 2012-04-02 03:37:45 <gmaxwell> DBordello: not really, we'll orphan blocks if they contain invalid BIP16 spends. But they won't form their own fork at least not persistantly.
222 2012-04-02 03:38:10 <gmaxwell> DBordello: the chain produced by the supermajority of BIP16 nodes is acceptable to the non-BIP16 nodes.
223 2012-04-02 03:38:38 <DBordello> gmaxwell, but there are blocks that non-BIP 16 nodes will accept but the supermajority of BIP 16 nodes will orphan?
224 2012-04-02 03:38:38 <gmaxwell> DBordello: the non-BIP16 nodes however, may produce/extend a chain which is not acceptable to the BIP16 nodes, and if they do those blocks will be orphaned.
225 2012-04-02 03:38:43 <copumpkin> what would have to happen for the tree to fork permanently and break into separate currencies? is that even possible without a major implementation fork?
226 2012-04-02 03:38:45 <gmaxwell> Correct.
227 2012-04-02 03:38:49 <denisx> so if I find another 173965 the BIP16 pools will accept that?
228 2012-04-02 03:38:53 <DBordello> gmaxwell, what kind of transaction would cause that?
229 2012-04-02 03:39:08 <gmaxwell> denisx: sure, so long as it's a BIP16 valid block.
230 2012-04-02 03:39:15 <gmaxwell> DBordello: 4005d6bea3a93fb72f006d23e2685b85069d270cb57d15f0c057ef2d5e3f78d2
231 2012-04-02 03:39:24 <gmaxwell> https://blockchain.info/tx-index/3618498/4005d6bea3a93fb72f006d23e2685b85069d270cb57d15f0c057ef2d5e3f78d2?show_adv=true
232 2012-04-02 03:39:25 <denisx> iam running 60006
233 2012-04-02 03:39:50 <gmaxwell> unfortunately blockchain.info is too smart for our own good: it's decoding the BIP16 input there and only displaying the serialized script.
234 2012-04-02 03:39:58 <gmaxwell> And that script is invalid as the day is long.
235 2012-04-02 03:40:18 <DBordello> ./bitcoind gettransaction 4005d6bea3a93fb72f006d23e2685b85069d270cb57d15f0c057ef2d5e3f78d2
236 2012-04-02 03:40:19 <DBordello> error: {"code":-5,"message":"No information available about transaction"}
237 2012-04-02 03:40:26 <copumpkin> doesn't blockexplorer show it?
238 2012-04-02 03:40:30 <gmaxwell> DBordello: yes, your upgraded node doesn't have it.
239 2012-04-02 03:40:38 <DBordello> gmaxwell, interesting :)
240 2012-04-02 03:40:56 <gmaxwell> Pre-BIP16 nodes accept it because they don't parse the seralized inner script, which is where the invalidity is.
241 2012-04-02 03:41:23 <DBordello> ah
242 2012-04-02 03:41:28 <gmaxwell> As you can see from that link, that txn has been mined 5 times today.
243 2012-04-02 03:41:32 <DBordello> who is the jerk who made this transaction?
244 2012-04-02 03:41:52 <copumpkin> DBordello: let me just write a VB GUI to find out
245 2012-04-02 03:41:55 <gmaxwell> DBordello: it actually could have been an accident  its invalid in a way that could have arisen by mistake.
246 2012-04-02 03:42:04 <DBordello> copumpkin, make sure it blames luke
247 2012-04-02 03:42:49 <gmaxwell> (the reason it's invalid is actually a well known bug in bitcoin checkmultisig pops consumes too many elements... so to use it you have to insert some padding, that txn lacks the padding.)
248 2012-04-02 03:43:01 <gmaxwell> (it's not a bug that can be fixed without creating forking risk)
249 2012-04-02 03:43:11 <copumpkin> boo :P
250 2012-04-02 03:43:36 <forsetifox> Never seen that network graph look like that before. http://www.bitcoinwatch.com/
251 2012-04-02 03:43:36 <gmaxwell> It's harmless in any case.
252 2012-04-02 03:43:58 <denisx> someone found another 173965
253 2012-04-02 03:43:59 <copumpkin> gmaxwell: but it makes every software engineer cringe :)
254 2012-04-02 03:44:12 <gmaxwell> denisx: as expected.
255 2012-04-02 03:44:31 <gmaxwell> Deepbit.
256 2012-04-02 03:45:01 <gmaxwell> Now the fun bit will be the mysteryminer will extend the dead chain and burn a couple blocks uselessly.
257 2012-04-02 03:45:19 <DBordello> gmaxwell, ah hah!  That should silence him for a while
258 2012-04-02 03:45:32 <gmaxwell> DBordello: he's lost like .. 7 or 8 blocks today, I think.
259 2012-04-02 03:45:42 <DBordello> gmaxwell, good, maybe he won't notice :)
260 2012-04-02 03:45:47 <ELT> lol
261 2012-04-02 03:45:55 <DBordello> blockchain.info is too awesome.
262 2012-04-02 03:47:50 <DBordello> It looks like deepbit had 173948 orphaned?  How come?
263 2012-04-02 03:48:20 <gmaxwell> DBordello: not a deepbit blocks.
264 2012-04-02 03:48:31 <DBordello> gmaxwell, ah, okay, the "relayed by" fooled me
265 2012-04-02 03:48:42 <gmaxwell> yea. ::nods::
266 2012-04-02 03:48:50 <Cory> Does Deepbit report their blocks somewhere?
267 2012-04-02 03:49:00 <gmaxwell> or if it is, it didn't show on the deepbit website, and it doesn't have p2sh in its coinbase like all the ones that show on the deepbit website.
268 2012-04-02 03:49:08 <Cory> Ah.
269 2012-04-02 03:49:31 <gmaxwell> (and thats why it was orphaned it included 4005d6bea3a93fb72f006d23e2685b85069d270cb57d15f0c057ef2d5e3f78d2)
270 2012-04-02 03:49:46 <DBordello> Have we had any multiple-block forks orphaned?
271 2012-04-02 03:49:57 <gmaxwell> sure.
272 2012-04-02 03:50:01 <gmaxwell> There was one with four blocks.
273 2012-04-02 03:50:29 <Cory> It kept pace for four blocks?
274 2012-04-02 03:50:39 <gmaxwell> thats how the MM is getting so screwed.. he doesn't mine any txn himself so he can't make an invalid block, but he can extend an invalid chain when someone else does!
275 2012-04-02 03:51:07 <DBordello> gmaxwell, hah
276 2012-04-02 03:51:23 <gmaxwell> Cory: the first was ozcoin then three MM blocks. (he had a lucky run)
277 2012-04-02 03:51:32 <Cory> Yikes.
278 2012-04-02 03:51:41 <gmaxwell> he was _late_ on all of them but you only switch when the other chain is longer.
279 2012-04-02 03:51:50 <gmaxwell> Not where there is a tine.
280 2012-04-02 03:51:52 <gmaxwell> er tie.
281 2012-04-02 03:52:05 <gmaxwell> so he kept extending the dead chain because the other one wasn't longer yet.
282 2012-04-02 03:52:06 <Cory> Yeah. Soon all of the pools losing money will stop accepting those transactions, though.
283 2012-04-02 03:52:26 <DBordello> I bet this fixes itself in a hurry
284 2012-04-02 03:52:26 <gmaxwell> nah, any pool losing money except ozcoin is dead at the wheel.
285 2012-04-02 03:52:42 <DBordello> What is the deal with ozcoin?
286 2012-04-02 03:52:52 <DBordello> Graet doesn't strike me as someone to get caught by surprise
287 2012-04-02 03:53:24 <gmaxwell> He thought he was upgraded. But it looks like the poolserver he uses can't actually copy the p2sh coinbase flag from bitcoin so it was simply hardcoded.
288 2012-04-02 03:53:42 <gmaxwell> This may have been masking the running of non-p2sh node software.
289 2012-04-02 03:53:52 <gmaxwell> luke-jr was looking into it.
290 2012-04-02 03:54:43 <gmaxwell> us figuring out the problem was delayed by the fact that blockchain.info is effectively disguising p2sh transactions. :(
291 2012-04-02 03:54:51 <DBordello> ah, he was running the erlang poolserver.  I have played with it, it is pretty slick
292 2012-04-02 03:54:51 <denisx> deepbit found another one
293 2012-04-02 03:55:36 <gmaxwell> denisx: yea, there you go. fork resolved.
294 2012-04-02 03:55:45 <gmaxwell> it shouldn't get any longer.
295 2012-04-02 03:55:55 <gmaxwell> (well, unless someone else just found a block now)
296 2012-04-02 03:56:27 <gmaxwell> copumpkin: sync moving now?
297 2012-04-02 03:56:45 <copumpkin> yep, thanks :)
298 2012-04-02 03:57:08 <ThomasV> gmaxwell: so the MM just lost 150 btc?
299 2012-04-02 03:57:44 <gmaxwell> ThomasV: I don't know what the total is... I think he's lost more like 6 blocks today.
300 2012-04-02 03:58:04 <gmaxwell> though he won't keep bleeding at quite that rate now that ozcoin is offline until its fixed.
301 2012-04-02 03:58:27 <ThomasV> but he probably doesn't pay for electricity (if he is a botnet)
302 2012-04-02 03:58:37 <Tril> it's normal to see tons of ORPHAN BLOCK when downloading the chain these days?
303 2012-04-02 04:02:44 <Tril> oh I think it's just downloading them out of order?
304 2012-04-02 04:04:13 <gribble> New news from bitcoinrss: Diapolo opened pull request 1025 on bitcoin/bitcoin <https://github.com/bitcoin/bitcoin/pull/1025>
305 2012-04-02 04:06:58 <gmaxwell> Tril: you got it!
306 2012-04-02 04:22:45 <gmaxwell> Blockexplorer is broken.
307 2012-04-02 04:22:46 <gmaxwell> http://blockexplorer.com/block/00000000000001d181b4842f31c7c0d56b36f0da90bbc78a657b2eefa97d3d11
308 2012-04-02 04:22:50 <gmaxwell> ^ it says that block doesn't exist.
309 2012-04-02 04:23:05 <nanotube> also, there is no spoon.
310 2012-04-02 04:23:06 <gmaxwell> It does, it's 958 http://blockchain.info/block-index/202232/00000000000001d181b4842f31c7c0d56b36f0da90bbc78a657b2eefa97d3d11 in the main chain.
311 2012-04-02 04:23:38 <gmaxwell> And, also, it reveals that deepbit's stats page is based on blockexplorer... since the stats page is wrong about that block.
312 2012-04-02 04:23:54 <forsetifox> 173970, another orphan.
313 2012-04-02 04:24:11 <denisx> no
314 2012-04-02 04:24:14 <ThomasV> nanotube: there is no spoon but there is a fork
315 2012-04-02 04:24:28 <nanotube> ThomasV: ha! :)
316 2012-04-02 04:24:31 <gmaxwell> forsetifox: yea. that was pointed out before in #bitcoin
317 2012-04-02 04:24:54 <denisx> oh, yes
318 2012-04-02 04:25:08 <gmaxwell> forsetifox: any block that contains https://blockchain.info/tx-index/3618498/4005d6bea3a93fb72f006d23e2685b85069d270cb57d15f0c057ef2d5e3f78d2?show_adv=true will be orphaned.
319 2012-04-02 04:25:35 <gmaxwell> 'included in blocks' field gives you that nice graveyard.
320 2012-04-02 04:26:04 <forsetifox> Ah, that's the strange one that ozcoin was getting.
321 2012-04-02 04:27:12 <forsetifox> That hard coded opcode is causing the orphans gmaxwell?
322 2012-04-02 04:27:22 <gmaxwell> ???
323 2012-04-02 04:27:37 <gmaxwell> The transaction is invalid.
324 2012-04-02 04:27:59 <forsetifox> What is it that's creating the orphans?
325 2012-04-02 04:28:20 <gmaxwell> It's a p2sh spend (which you can't tell, because blockchain.info only shows you the inner script), and it misuses checkmultisig. The script there fails to validate.
326 2012-04-02 04:28:28 <forsetifox> Gotcha.
327 2012-04-02 04:38:54 <gmaxwell> sipa: is it possible that the people with the crashing wallets have the wrong passphrase... but its one that makes the private key decrypt to a null byte first... which is then read as zero length... and then things go pear-shaped?
328 2012-04-02 04:39:14 <gmaxwell> (I ask because I haven't spent as much time staring at that code today as you have and I'm about to go to bed)
329 2012-04-02 04:43:56 <forsetifox> gmaxwell. Are you still here?
330 2012-04-02 04:44:03 <gmaxwell> Yes.
331 2012-04-02 04:44:11 <forsetifox> Why do some of the blocks only have one transaction?
332 2012-04-02 04:44:32 <gmaxwell> Because sometimes there just aren't any transactions waiting.
333 2012-04-02 04:44:37 <forsetifox> Ah.
334 2012-04-02 04:44:45 <Habbie> and some miners just don't bother with including transactions
335 2012-04-02 04:44:47 <gmaxwell> Especially when blocks come in rapid succession.
336 2012-04-02 04:45:10 <gmaxwell> And as Habbie says well, at least one appears to not include transactions.
337 2012-04-02 04:45:48 <gmaxwell> There are also some (e.g. eligius) which won't mine zero fee transactions. (though eligius will mine transactions with infinitesmal fee, or ones handed to eligius by special arrangement)
338 2012-04-02 04:49:50 <forsetifox> Can you explain "mining transactions"?
339 2012-04-02 04:51:09 <Tril> forsetifox: it means they will attempt to put the transaction in a block
340 2012-04-02 04:51:21 <Tril> if it does not have a fee, they will disregard it
341 2012-04-02 04:51:38 <forsetifox> I think I get it.
342 2012-04-02 04:52:03 <forsetifox> When I transfer funds to a Gox or whatever a pool has to pick that up and apply it to one of thier blocks?
343 2012-04-02 04:52:23 <Tril> yes
344 2012-04-02 04:52:26 <forsetifox> Sweet.
345 2012-04-02 04:52:36 <forsetifox> Thanks for the information guys. =3
346 2012-04-02 04:54:17 <forsetifox> So if someone is only mining transactions but not blocks how do they get the transactions into a block?
347 2012-04-02 04:54:53 <Tril> you are always mining blocks.. you might want to check out the wiki.
348 2012-04-02 04:54:59 <forsetifox> k
349 2012-04-02 05:14:41 <Graet> is -payscripthashtime=1333238400 obsolete with this 0.6 rc3?
350 2012-04-02 05:18:08 <Cory> More orphaned blocks.
351 2012-04-02 05:38:31 <Graet> ozcoin is back up with 0.6 bitcoin and other nupdates
352 2012-04-02 05:38:59 <Cory> Yay!
353 2012-04-02 05:39:41 <da2ce7> :)
354 2012-04-02 05:39:56 <da2ce7> good on Graet and #ozcoin
355 2012-04-02 05:40:47 <Graet> :)
356 2012-04-02 06:33:03 <[Tycho]> gmaxwell: no.
357 2012-04-02 06:36:12 <Graet> hi [Tycho]
358 2012-04-02 06:40:18 <[Tycho]> Graet: hello.
359 2012-04-02 06:40:48 <Graet> :)
360 2012-04-02 07:02:56 <[Tycho]> How that 4005d6be TX should be failed in debug.log ?
361 2012-04-02 07:03:43 <[Tycho]> It would be better if I disabled mining of P2SH TXes at all...
362 2012-04-02 07:16:30 <[Tycho]> Anyone here ?
363 2012-04-02 07:26:17 <sipa> yes
364 2012-04-02 07:26:38 <sipa> wow, invalid P2SH spend, mined by ozcoin, and extended by MM?
365 2012-04-02 07:26:51 <[Tycho]> sipa: should that TX be failed by ConnectInputs or something else ?
366 2012-04-02 07:28:11 <Graet> 2 times iirc sipa
367 2012-04-02 07:28:28 <Graet> updated properly now
368 2012-04-02 07:28:49 <sipa> [Tycho]: by ConnectInputs, I'd say
369 2012-04-02 07:29:20 <[Tycho]> Heh, shows both branches as orphaned :) https://blockchain.info/orphaned-blocks
370 2012-04-02 07:30:19 <[Tycho]> I wonder who generates all that failed blocks if it's not OzCoin and MM
371 2012-04-02 07:30:59 <Graet> p2pool node not updated ;)
372 2012-04-02 07:36:36 <TuxBlackEdo> really interesting stuff going on in the blockchain
373 2012-04-02 07:36:38 <TuxBlackEdo> right?
374 2012-04-02 07:36:55 <TuxBlackEdo> how does an invalid p2sh spend happen?
375 2012-04-02 07:37:14 <TuxBlackEdo> and why doesn't the pool reject that transaction
376 2012-04-02 07:37:18 <TuxBlackEdo> bad code or something?
377 2012-04-02 07:38:34 <sipa> because it didn't upgrade
378 2012-04-02 07:39:11 <sipa> that's exactly the problem: most miners reject it, but some don't
379 2012-04-02 07:41:27 <TuxBlackEdo> the pool didn't update to 0.6.0?
380 2012-04-02 07:41:43 <sipa> yes
381 2012-04-02 07:41:53 <sipa> most pools run far from the latest version
382 2012-04-02 07:42:12 <TuxBlackEdo> so in older versions of bitcoin p2sh spends were allowed?
383 2012-04-02 07:42:17 <TuxBlackEdo> i am confused
384 2012-04-02 07:42:39 <_W_> so should we expect incoming drama from miners now not making coin?
385 2012-04-02 07:43:33 <[Tycho]> Graet: no, they don't have that much hashrate
386 2012-04-02 07:43:41 <sipa> TuxBlackEdo: of, course BIP16 is only implemented in 0.6.0
387 2012-04-02 07:46:51 <TuxBlackEdo> why isn't a p2sh transaction backwards compatiable in 0.6.0?
388 2012-04-02 07:47:09 <sipa> that's the point: it is backward compatible
389 2012-04-02 07:47:33 <sipa> the only way to make new features backward compatible is by making sure that old clients accept them
390 2012-04-02 07:47:43 <sipa> you can only make the rules stricter
391 2012-04-02 07:49:07 <TuxBlackEdo> so if ozcoin didn't include that p2sh transaction, their block would have been accepted by the network?
392 2012-04-02 07:49:33 <Graet> yes x 2,
393 2012-04-02 07:50:29 <TuxBlackEdo> i still don't understand, sipa said "it is backward compatible" yet their block didn't get accepted by the network...?
394 2012-04-02 07:51:19 <TuxBlackEdo> sorry i am really trying to understand... thanks for trying to explain it to me
395 2012-04-02 07:51:59 <sipa> TuxBlackEdo: old miners will see every BIP16 transaction as valid, new miners won't
396 2012-04-02 07:52:22 <sipa> sorry; new miners will accept them only if they are valid BIP16 spends
397 2012-04-02 07:56:17 <sipa> that way, it's not a problem if not everyone upgrades, as long as the majority does
398 2012-04-02 07:57:00 <TuxBlackEdo> i am reading the bip16 page on bitcoin.it
399 2012-04-02 07:57:46 <sipa> not sure how detailed that principle is explained there, but it's a general rule: if you want a new protocol rule, all it can do it make previously-valid things invalid, and not the other way around
400 2012-04-02 07:58:35 <sipa> because what if a majority upgraded to allow something that was previously not allowed? older nodes would not accept the longest chain anyway, and end up on a permanent fork
401 2012-04-02 08:01:16 <[Tycho]> 20 invalids since 01.04, wow
402 2012-04-02 08:01:30 <TuxBlackEdo> why couldn't the client be written in a way where it would process protocol commands it does understand and ignore the ones it doesn't?
403 2012-04-02 08:01:51 <sipa> TuxBlackEdo: everyone needs to agree on the blockchain
404 2012-04-02 08:01:59 <sipa> you can't have anyone who ignores part of it
405 2012-04-02 08:19:23 <sturles> 20 invalid blocks due to invalid BIP 16 transactions included?
406 2012-04-02 08:19:48 <[Tycho]> Yes.
407 2012-04-02 08:20:21 <lh77> is this a problem?
408 2012-04-02 08:20:46 <sturles> Wow!  Difficulty may even go down a notch then, unless people update fast.
409 2012-04-02 08:20:55 <conman> pretty sure it will
410 2012-04-02 08:21:11 <conman> the number of people who mine blindly unattended without thought is clearly high
411 2012-04-02 08:22:40 <t7> :O
412 2012-04-02 08:22:43 <t7> quick
413 2012-04-02 08:45:50 <[Tycho]> Hmm, mtgox fee is low again. Or it's a 1st april joke ?
414 2012-04-02 08:46:13 <Graet> nah 1st birthday promotion
415 2012-04-02 08:46:14 <conman> isn't it a bit late for that?
416 2012-04-02 08:46:38 <conman> april fools I mean
417 2012-04-02 09:25:40 <sje> i'm still seeing all my outgoing connections time out - updated to latest git just now and did a clean and rebuild...
418 2012-04-02 09:26:07 <sje> is there a log message that would should if my client was just trying to connect to itself? i.e. if my firewall rules were stuffed?
419 2012-04-02 09:32:21 <sje> ffs that was it
420 2012-04-02 09:33:24 <Joric> yeah
421 2012-04-02 09:47:55 <[Tycho]> nmcbit just did it...
422 2012-04-02 09:51:04 <Joric> [Tycho], are you hiring? i just figured i'm spending most of the time on bitcoins
423 2012-04-02 09:52:02 <Joric> cs degree and shit
424 2012-04-02 09:52:47 <conman> heh
425 2012-04-02 09:53:08 <coingenuity> lol
426 2012-04-02 09:53:28 <conman> shame I haven't got a cs degree and shit
427 2012-04-02 09:53:36 <[Tycho]> No :)
428 2012-04-02 09:53:37 <coingenuity> [Tycho]: is probably not hiring...since....his stuff runs mostly on autopilot ftw
429 2012-04-02 09:53:38 <conman> I seem to just have the shit
430 2012-04-02 09:53:46 <luke-jr> CS degree *is* crap
431 2012-04-02 09:53:49 <coingenuity> same here, conman
432 2012-04-02 09:54:00 <conman> not just the shit but tEh shit
433 2012-04-02 09:54:45 <sje> CS degree is awesome - 4 years drinking beer :)
434 2012-04-02 09:55:34 <Graet> i managed to drink beer for more than 4 years, without getting a degree
435 2012-04-02 09:55:35 <Graet> :P
436 2012-04-02 09:55:40 <conman> lol
437 2012-04-02 09:55:44 <sje> ha
438 2012-04-02 09:56:00 <Graet> now i prefer rum :)
439 2012-04-02 09:56:05 <sje> yeah i took a bit longer too - no need to rush those things
440 2012-04-02 09:56:11 <Graet> :)
441 2012-04-02 09:56:31 <sje> i did get my degree in the end but i blame my wife for that
442 2012-04-02 09:56:33 <Graet> i looki forward to your shout conman :P
443 2012-04-02 09:56:44 <sje> turns out the psychology people were waiting for me to write something 'nice'
444 2012-04-02 09:56:45 <conman> gladly
445 2012-04-02 09:57:42 <luke-jr> I continue to be amazed at all the people posting on the Ars thread "durr, pool is down?"
446 2012-04-02 09:57:49 <conman> heh yeah
447 2012-04-02 09:58:09 <Graet> +1 thier irc has been a bit constant too
448 2012-04-02 09:58:36 <Graet> thanks for the new link to share, the other i was using wasnt so amusing :)
449 2012-04-02 09:59:13 <Graet> ;;bc,nethash
450 2012-04-02 09:59:13 <gribble> 11830.262935960041
451 2012-04-02 09:59:40 <conman> has dropped
452 2012-04-02 09:59:45 <conman> looks like they didn't even have backups set
453 2012-04-02 10:00:05 <conman> which doesn't surprise if they were mining blindly to begin with
454 2012-04-02 10:00:11 <Graet> yep
455 2012-04-02 10:02:05 <Joric> i'm about to celebrate 10 years anniversary from teh graduation this summer
456 2012-04-02 10:02:29 <Joric> gonna troll them saying i turned into christianity
457 2012-04-02 10:03:41 <Joric> it's kindof modern now
458 2012-04-02 10:04:02 <[Tycho]> What happened to ArsBitcoin ?
459 2012-04-02 10:04:13 <conman> they did nothing
460 2012-04-02 10:04:16 <conman> and then it crashed
461 2012-04-02 10:04:26 <conman> and now it's both down and off the bip16 chain path
462 2012-04-02 10:06:05 <conman> Burning toad said he was planning on taking it down at some stage
463 2012-04-02 10:06:11 <conman> looks like his hand was forced  now
464 2012-04-02 10:59:52 <sipa> gmaxwell: any valid encrypted record should decrypt to a 32-byte result
465 2012-04-02 11:00:07 <sipa> even under a wrong key
466 2012-04-02 11:15:28 <gmaxwell> sipa: yea.. N-am brainfart. I was thinking the keys were in base-58 encoded form, when I should have remembered they weren't just based on their length.
467 2012-04-02 11:15:56 <user> gavinandresen: my comments here maake sense? https://gist.github.com/830ca16758fb9ad496d7
468 2012-04-02 11:16:41 <sipa> gavinandresen: time to open 0.7.0's merge windows?
469 2012-04-02 11:17:06 <gavinandresen> user: yes. Although that document is supposed to be all about 2-party escrow (3-party escrow will be like you describe, involving a website, I think)
470 2012-04-02 11:17:12 <gavinandresen> sipa: yes
471 2012-04-02 11:17:55 <gavinandresen> sipa: version number for the merge window period?  0.6.0.99 ?
472 2012-04-02 11:17:56 <user> ok
473 2012-04-02 11:18:38 <luke-jr> 0.6.99 makes more sense imo
474 2012-04-02 11:19:50 <sipa> 0.6.99, indeed, as it will have more features than any stable branch we fork off 0.6.0 now
475 2012-04-02 11:20:11 <gavinandresen> What new features?  I see... coin control stuff....
476 2012-04-02 11:20:39 <luke-jr> gavinandresen: get my email? :P
477 2012-04-02 11:20:49 <sipa> gavinandresen: well, that's not known yet, i just mean: everything in the 0.7 merge window will have new features, why 0.6.x will not
478 2012-04-02 11:20:56 <sipa> *while
479 2012-04-02 11:21:40 <gavinandresen> Well, that's what I'm asking: should the next release be 0.6.1, where we do lots of internal clean-ups but no major new features, or 0.7
480 2012-04-02 11:22:23 <Blitzboom> will 0.7 have multisig tx in the GUI?
481 2012-04-02 11:23:09 <gavinandresen> Blitzboom: unlikely
482 2012-04-02 11:23:24 <gavinandresen> Blitzboom: .... but what do you mean when you say 'multisig' ?
483 2012-04-02 11:23:24 <sipa> gavinandresen: Oh, that's also a possibility; what kind of cleanups were you thinking about?
484 2012-04-02 11:23:54 <gavinandresen> sipa: there are a bunch of pull requests that are cleanups
485 2012-04-02 11:24:08 <Blitzboom> transactions that will only be spent when multiple parties agree
486 2012-04-02 11:24:40 <Blitzboom> what is mainly planned then for the next major release, if i may ask?
487 2012-04-02 11:24:51 <gavinandresen> Blitzboom: that's what this conversation is about
488 2012-04-02 11:24:59 <sipa> ok, first 0.6.0.99 then, pull cleanups, move to 0.6.1 (possibly release candidates, but hopefully nit many) 0.6.1.1, and then 0.6.99?
489 2012-04-02 11:25:02 <Blitzboom> oh, okay :D
490 2012-04-02 11:25:40 <sipa> wait, what? 50 pull requests open
491 2012-04-02 11:25:49 <gavinandresen> sipa: "we" could insist on cleaning up some longstanding bugs/issues, like the slow shutdown
492 2012-04-02 11:26:09 <gavinandresen> sipa: and the reports of encrypted wallets crashing when unlocked
493 2012-04-02 11:27:49 <gmaxwell> I like this plan.
494 2012-04-02 11:27:54 <gavinandresen> sipa: So, in my mind, the first question is:  should the next release be bug-fix-and-cleanup only?  That'll make some people who want development to move faster unhappy.
495 2012-04-02 11:28:17 <sipa> No, it may be time for that; as long as 0.6.1 doesn't take 4 months.
496 2012-04-02 11:28:36 <luke-jr> does CBlockStore count as cleanup? :p
497 2012-04-02 11:30:13 <gmaxwell> I think CBlockStore could be taken in a bug-fix-and-cleanup cycle simply because it's not at all user/interface/network visible and could be backed out if it had any issues, although I might lean a little against it.
498 2012-04-02 11:30:44 <sipa> I've started testing blockstore and change a few things myself
499 2012-04-02 11:30:50 <gmaxwell> Certantly CoinControl would not count as cleanup.
500 2012-04-02 11:31:57 <gavinandresen> I haven't looked at the difference between CBlockStore and Michael Gronager's refactoring(s)
501 2012-04-02 11:32:16 <sipa> gavinandresen: libcoin is far far more intrusive
502 2012-04-02 11:33:06 <gmaxwell> luke-jr: Did you figure anything out about Ozcoin's bitcoind?
503 2012-04-02 11:33:25 <luke-jr> gmaxwell: I fell asleep.
504 2012-04-02 11:33:51 <luke-jr> gavinandresen: afaict, libcoin effectively reverts most cleanups since 0.4
505 2012-04-02 11:34:34 <luke-jr> as much as I think their general direction is a good idea, I think it needs to be taken in steps
506 2012-04-02 11:34:42 <sipa> gavinandresen: libcoin is mostly about splitting the code in a few layers, and reduce lock contention maximally (and though it seems well thought out, I'm still not convinced it's entirely safe; for example, he was not aware that reading a stl structure required a lock if it could be written to at the same time)
507 2012-04-02 11:35:34 <gavinandresen> right, I'm just wondering if we like his block-chain abstraction better than Matt's
508 2012-04-02 11:35:53 <gmaxwell> Probably time to run a lock profiler against bitcoin again with the bdb speedups...
509 2012-04-02 11:36:16 <sipa> blockstore only moves globals from main (partially, it's not 100% yet) to a separate class, and adds a callback processor that runs in one (or more!) separate threads
510 2012-04-02 11:37:37 <luke-jr> so are you guys decided on 0.6.1 vs 0.7.0?
511 2012-04-02 11:37:49 <luke-jr> (so I can make the 0.6.0.x or 0.6.x stable branches)
512 2012-04-02 11:38:09 <sipa> gavinandresen: i took me a few mails with him in private before i understood why acceptblock didn't need to take a lock
513 2012-04-02 11:38:54 <sipa> it sounded safe afterwards, but understanding every single detail will take a lot of time
514 2012-04-02 11:39:38 <gmaxwell> Personally I don't see a lot of value to removing locks that can't be shown to be either contended or in very frequent fastpath code.
515 2012-04-02 11:40:16 <gavinandresen> sipa gmaxwell : I'm inclined to do a cleanup/bug-fix 0.6.1, with a target rc1 date in... oh, I dunno, 3 or 4 weeks ?
516 2012-04-02 11:41:12 <gmaxwell> That would also give some time for actual features to get more developed we don't really have a backlog of pull-ready features.
517 2012-04-02 11:43:32 <gmaxwell> http://bitcoinstatus.rowit.co.uk/versions.html  < 3.24 still the single most popular listening version?
518 2012-04-02 11:43:54 <luke-jr> I think 4005d6bea3 may need more attention.
519 2012-04-02 11:44:08 <[Tycho]> luke-jr: why ?
520 2012-04-02 11:44:24 <luke-jr> stable backport is giving me ERROR: ConnectInputs() : 4005d6bea3 VerifySignature failed
521 2012-04-02 11:44:35 <gmaxwell> It sould be.
522 2012-04-02 11:44:35 <luke-jr> 0.6 gives: ERROR: ConnectInputs() : 4005d6bea3 P2SH VerifySignature failed
523 2012-04-02 11:44:38 <gmaxwell> Oh.
524 2012-04-02 11:45:01 <sipa> gavinandresen: ok, but i'd try not to delay things further then
525 2012-04-02 11:45:04 <luke-jr> so either our assumptions about 4005d6 are wrong, or there's something wrong with the stable backport
526 2012-04-02 11:45:13 <gavinandresen> sipa: delay what things?
527 2012-04-02 11:45:17 <luke-jr> has anyone managed to get a copy of the txn yet?
528 2012-04-02 11:45:30 <gmaxwell> Does it even have any code for "P2SH VerifySignature failed"?
529 2012-04-02 11:46:19 <gmaxwell> Okay, I have a 0.3.19 node which has AcceptToMemoryPool(): accepted 4005d6bea3
530 2012-04-02 11:46:24 <gmaxwell> So I have it in memory. :)
531 2012-04-02 11:46:31 <sipa> gavinandresen: have a 3-week merge window for 0.6.1, but if certain things aren't ready by then, don't delay rc1 further
532 2012-04-02 11:46:47 <gavinandresen> sipa: ACK
533 2012-04-02 11:47:21 <gavinandresen> sipa:  You can always pick 2 of 3 for a release:  features, quality, time.  For 0.6.1 we'll pick quality, time.
534 2012-04-02 11:47:28 <gmaxwell> Frankly, I think the adoption curves would be helped by cutting some post 0.6 point releases even if they change very little.
535 2012-04-02 11:47:57 <gavinandresen> gmaxwell: agreed.
536 2012-04-02 11:49:23 <gavinandresen> So then we get to Blitzboom's question:  what features would we like to see for a 0.7 ?
537 2012-04-02 11:50:38 <sipa> coin control, ipv6, deterministic wallets
538 2012-04-02 11:50:41 <gavinandresen> I'd vote deterministic wallet with an automatic backup scheme
539 2012-04-02 11:51:03 <Blitzboom> would it be possible to connect it to dropbox and do encryption?
540 2012-04-02 11:51:26 <gavinandresen> I like the idea of it emailing an encrypted copy of the master root key to me
541 2012-04-02 11:51:47 <sipa> gavinandresen: encrypted using your key, or their key? ;)