1 2012-11-26 00:16:49 <jgarzik> ;;bc,halfreward
  2 2012-11-26 00:16:50 <gribble> Estimated time of bitcoin block reward halving: Wed Nov 28 13:46:00 2012 | Time remaining: 2 days, 20 hours, 30 minutes, and 0 seconds
  3 2012-11-26 00:16:51 <gribble> Estimated time of bitcoin block reward halving: Wed Nov 28 13:46:00 2012 | Time remaining: 2 days, 20 hours, 30 minutes, and 0 seconds
  4 2012-11-26 03:11:40 <jgarzik> hah!
  5 2012-11-26 03:11:41 <jgarzik> http://bitcoinclock.com/
  6 2012-11-26 03:11:42 <jgarzik> http://bitcoinclock.com/
  7 2012-11-26 03:11:49 <jgarzik> counting down to 25 BTC reward
  8 2012-11-26 03:15:09 <jgarzik> https://bitcointalk.org/index.php?topic=127604.msg1357522#msg1357522
  9 2012-11-26 03:15:10 <jgarzik> https://bitcointalk.org/index.php?topic=127604.msg1357522#msg1357522
 10 2012-11-26 03:15:15 <jgarzik> noooooooooooooooooo   "In fact, a new version of the draft is already with the IETF Secretariat that should be published in the next few hours that covers Tonal Bitcoin (TBC) by request of another user on this forum."
 11 2012-11-26 03:39:30 <rex> c:\\bitcoin-master\\src>mingw32-make -f makefile.mingw mingw32-make: *** No rule to make target `leveldb/libleveldb.lib', needed by `ob j/leveldb.o'.  Stop.
 12 2012-11-26 03:40:16 <rex> I recompiled leveldb, now i'm trying to compile bitcoin
 13 2012-11-26 03:42:47 <rex> I don't know how to solve this
 14 2012-11-26 03:42:48 <rex> I don't know how to solve this
 15 2012-11-26 03:45:17 <muhoo> tonal? bitcoin . oooh kaaay...
 16 2012-11-26 03:45:18 <muhoo> tonal? bitcoin . oooh kaaay...
 17 2012-11-26 03:49:16 <rex> thanks for any help. it seems that target obj/leveldb.o needs leveldb/libleveldb.lib, but leveldb/libleveldb.lib is not generated...(don't quite understand this)
 18 2012-11-26 03:52:08 <rex> in folder bitcoin/src/leveldb, i managed to generate libleveldb.a libmemenv.a by TARGET_OS=OS_WINDOWS_CROSSCOMPILE make libleveldb.a libmemenv.a, there's no libleveldb.lib
 19 2012-11-26 04:21:00 <rex> ok, i managed to build bitcoind.exe by replacing makefile.mingw line 100 obj/leveldb.o: leveldb/libleveldb.lib with obj/leveldb.o: leveldb/libleveldb.a. It builds successfully, but just can not run. error: Bitcoin: Error initializing database environment
 20 2012-11-26 04:21:01 <rex> ok, i managed to build bitcoind.exe by replacing makefile.mingw line 100 obj/leveldb.o: leveldb/libleveldb.lib with obj/leveldb.o: leveldb/libleveldb.a. It builds successfully, but just can not run. error: Bitcoin: Error initializing database environment
 21 2012-11-26 06:15:00 <Diablo-D3> wrf
 22 2012-11-26 06:15:01 <Diablo-D3> wrf
 23 2012-11-26 06:15:03 <Diablo-D3> wtf
 24 2012-11-26 06:15:07 <Diablo-D3> so I fell asleep in my chair
 25 2012-11-26 06:15:15 <Diablo-D3> I assume bitcoin finished
 26 2012-11-26 06:15:20 <Diablo-D3> it "ran out of disk space" in the ram drive
 27 2012-11-26 06:15:22 <Diablo-D3> 4.8gb used
 28 2012-11-26 06:15:23 <Diablo-D3> 4.8gb used
 29 2012-11-26 06:15:25 <Diablo-D3> wtf did it do
 30 2012-11-26 06:17:13 <Diablo-D3> 2gb blk0001, 1.6gb blk0002, 1.2gb blkindex.dat
 31 2012-11-26 06:17:14 <Diablo-D3> 2gb blk0001, 1.6gb blk0002, 1.2gb blkindex.dat
 32 2012-11-26 06:21:09 <midnightmagic> is github bitcoin HEAD in midst of flux at the moment which is known to be unstable?
 33 2012-11-26 06:21:22 <Diablo-D3> possibly
 34 2012-11-26 06:21:23 <Diablo-D3> possibly
 35 2012-11-26 06:21:36 <Diablo-D3> they're still trying to get this shit in order for 0.8
 36 2012-11-26 06:21:39 <midnightmagic> i know about possibly. i mean known..
 37 2012-11-26 06:21:40 <midnightmagic> i know about possibly. i mean known..
 38 2012-11-26 06:21:53 <dparrish> i'll be happy when "bitcoind move" doesn't hang...
 39 2012-11-26 06:21:54 <dparrish> i'll be happy when "bitcoind move" doesn't hang...
 40 2012-11-26 06:24:12 <Diablo-D3> okay so
 41 2012-11-26 06:24:18 <Diablo-D3> stuff copied back out of ramdisk
 42 2012-11-26 06:24:19 <Diablo-D3> stuff copied back out of ramdisk
 43 2012-11-26 06:24:24 <Diablo-D3> lets see where it left me
 44 2012-11-26 06:24:37 <Diablo-D3> 206383
 45 2012-11-26 06:24:39 <Diablo-D3> ;;bc,blocks
 46 2012-11-26 06:24:40 <gribble> 209634
 47 2012-11-26 06:24:49 <Diablo-D3> hah it segfaulted almost at the very end
 48 2012-11-26 06:24:50 <Diablo-D3> hah it segfaulted almost at the very end
 49 2012-11-26 06:25:29 <Diablo-D3> 3333 blocks to go lol
 50 2012-11-26 06:48:05 <swulf--> has anyone noticed a deadlock with the 'move' command lately?
 51 2012-11-26 06:48:06 <swulf--> has anyone noticed a deadlock with the 'move' command lately?
 52 2012-11-26 06:48:42 <swulf--> in git head
 53 2012-11-26 06:52:40 <midnightmagic> ah thank you: full block hash in debug.log is very awesome and much appreciated. :)
 54 2012-11-26 06:52:41 <midnightmagic> ah thank you: full block hash in debug.log is very awesome and much appreciated. :)
 55 2012-11-26 07:14:27 <thermoman> swulf--: this was fixed in git
 56 2012-11-26 07:14:28 <thermoman> swulf--: this was fixed in git
 57 2012-11-26 07:15:02 <thermoman> swulf--: https://github.com/bitcoin/bitcoin/pull/2009
 58 2012-11-26 07:16:30 <thermoman> dparrish: ^
 59 2012-11-26 07:16:31 <thermoman> dparrish: ^
 60 2012-11-26 07:24:34 <dparrish> thanks!
 61 2012-11-26 07:38:06 <Diablo-D3> hrm
 62 2012-11-26 07:38:07 <Diablo-D3> hrm
 63 2012-11-26 07:38:08 <Diablo-D3> hey midnightmagic
 64 2012-11-26 07:38:39 <Diablo-D3> whats your ~/.bitcoin/database/log.* number?
 65 2012-11-26 07:38:40 <midnightmagic> hey
 66 2012-11-26 07:38:41 <midnightmagic> hey
 67 2012-11-26 07:38:47 <midnightmagic> hrm..
 68 2012-11-26 07:38:48 <midnightmagic> hrm..
 69 2012-11-26 07:38:54 <Diablo-D3> well, the highest
 70 2012-11-26 07:38:58 <Diablo-D3> you probably have more than one
 71 2012-11-26 07:39:19 <midnightmagic> 181
 72 2012-11-26 07:39:20 <midnightmagic> 181
 73 2012-11-26 07:39:26 <Diablo-D3> huh
 74 2012-11-26 07:39:31 <Diablo-D3> I guess that number isnt portabl
 75 2012-11-26 07:39:36 <Diablo-D3> *portable
 76 2012-11-26 07:39:51 <Diablo-D3> Im on 753
 77 2012-11-26 07:39:53 <midnightmagic> that is on the bitcoin instance where I'm tracking -current btw.
 78 2012-11-26 07:40:01 <Diablo-D3> oh
 79 2012-11-26 07:40:06 <Diablo-D3> erm
 80 2012-11-26 07:40:24 <Diablo-D3> but leveldb doesnt use those files
 81 2012-11-26 07:40:25 <Diablo-D3> but leveldb doesnt use those files
 82 2012-11-26 07:40:35 <midnightmagic> I have a "real" one with 366. but those are just db logs aren't they
 83 2012-11-26 07:40:36 <midnightmagic> I have a "real" one with 366. but those are just db logs aren't they
 84 2012-11-26 07:40:47 <Diablo-D3> write intent logs
 85 2012-11-26 07:40:48 <Diablo-D3> write intent logs
 86 2012-11-26 07:40:57 <Diablo-D3> or really, its a wall log
 87 2012-11-26 07:41:02 <midnightmagic> so what do you mean they aren't portable?
 88 2012-11-26 07:41:13 <Diablo-D3> well, Im on 753
 89 2012-11-26 07:41:14 <Diablo-D3> you're not
 90 2012-11-26 07:41:15 <Diablo-D3> you're not
 91 2012-11-26 07:41:49 <midnightmagic> yeah so?
 92 2012-11-26 07:42:10 <Diablo-D3> so you'd think doing the same things would produce the same numbers
 93 2012-11-26 07:42:11 <Diablo-D3> so you'd think doing the same things would produce the same numbers
 94 2012-11-26 07:42:37 <midnightmagic> seems to me the age is more a factor there..?
 95 2012-11-26 07:42:38 <midnightmagic> seems to me the age is more a factor there..?
 96 2012-11-26 07:42:59 <Diablo-D3> yeah, but wouldnt age drive it up?
 97 2012-11-26 07:43:00 <Diablo-D3> yeah, but wouldnt age drive it up?
 98 2012-11-26 07:43:00 <midnightmagic> whoaaaaahhh   bedtime. :(
 99 2012-11-26 07:43:01 <midnightmagic> whoaaaaahhh   bedtime. :(
100 2012-11-26 07:43:09 <Diablo-D3> this is a brand new db
101 2012-11-26 07:43:10 <Diablo-D3> this is a brand new db
102 2012-11-26 07:43:11 <midnightmagic> yeah, and the other node is young.
103 2012-11-26 07:43:12 <midnightmagic> yeah, and the other node is young.
104 2012-11-26 07:43:14 <midnightmagic> oh
105 2012-11-26 07:43:15 <Diablo-D3> had to redownload because fs fuckup
106 2012-11-26 07:43:15 <midnightmagic> oh
107 2012-11-26 07:43:16 <Diablo-D3> had to redownload because fs fuckup
108 2012-11-26 07:43:22 <Diablo-D3> infact its not even done
109 2012-11-26 07:43:23 <Diablo-D3> infact its not even done
110 2012-11-26 07:43:25 <Diablo-D3> 1442 blocks to go
111 2012-11-26 07:43:26 <Diablo-D3> 1442 blocks to go
112 2012-11-26 07:44:39 <Diablo-D3> how big are your blk files?
113 2012-11-26 07:44:40 <Diablo-D3> how big are your blk files?
114 2012-11-26 07:47:55 <thermoman> the database log file numbers are not portable and are not important
115 2012-11-26 07:47:56 <thermoman> the database log file numbers are not portable and are not important
116 2012-11-26 07:48:11 <thermoman> it's just an increasing number for the database journal
117 2012-11-26 08:12:32 <t7> how many blocks 'til the change
118 2012-11-26 08:13:45 <t7> 2 days
119 2012-11-26 08:34:24 <sipa> ;;bc,halfreward
120 2012-11-26 08:34:29 <gribble> Error: invalid syntax (<string>, line 1)
121 2012-11-26 08:34:30 <gribble> Error: invalid syntax (<string>, line 1)
122 2012-11-26 08:35:36 <t7> price keeps going up :( why did i sell
123 2012-11-26 08:35:37 <t7> price keeps going up :( why did i sell
124 2012-11-26 08:35:47 <t7> hopefully a big crash when reward drops
125 2012-11-26 08:35:48 <t7> hopefully a big crash when reward drops
126 2012-11-26 08:35:59 <jeremias> speculation elsewhere plz
127 2012-11-26 09:05:42 <vazakl> macy's day parade confetti reconstructed to reveal satoshi's identity
128 2012-11-26 09:08:45 <sipa> ?
129 2012-11-26 09:27:00 <t7> lel
130 2012-11-26 09:56:29 <thermoman> vazakl: /quit
131 2012-11-26 09:56:30 <thermoman> vazakl: /quit
132 2012-11-26 10:05:08 <sipa> midnightmagic: git head is quite stable now - there is one known bug (segfault...) that may occur if you leave it running for a long time and mine on it
133 2012-11-26 11:55:58 <Ferroh> Is 0.7.99 using levelDB?
134 2012-11-26 11:57:03 <sipa> yes
135 2012-11-26 11:57:08 <Ferroh> awesome
136 2012-11-26 11:57:30 <sipa> but 0.7.99 doesn't really exist as a release, it's the internal version number used while we're working on 0.8
137 2012-11-26 11:57:36 <sipa> but current git head uses leveldb, yes
138 2012-11-26 11:57:37 <sipa> but current git head uses leveldb, yes
139 2012-11-26 11:57:45 <Ferroh> ok, that's what I was really asking
140 2012-11-26 12:00:25 <sipa> not only that, git head also uses a very different database layout, which doesn't require access to the full block chain during block/tx validation
141 2012-11-26 12:00:26 <sipa> not only that, git head also uses a very different database layout, which doesn't require access to the full block chain during block/tx validation
142 2012-11-26 13:57:30 <kjj_> are the values stored in the wallet.db documented anywhere?
143 2012-11-26 14:14:11 <Jouke> hmm, backupwallet doesn't overwrite an existing file?
144 2012-11-26 14:14:12 <Jouke> hmm, backupwallet doesn't overwrite an existing file?
145 2012-11-26 14:17:39 <helo> 
ACTION tests and sees that it does not

146 2012-11-26 14:18:25 <Jouke> right. Is that a bug or a feature? :)
147 2012-11-26 14:19:19 <helo> an error would be nice :/
148 2012-11-26 14:19:20 <helo> an error would be nice :/
149 2012-11-26 14:23:41 <Jouke> Myeah, submitted a ticket just to be sure.
150 2012-11-26 14:23:42 <Jouke> Myeah, submitted a ticket just to be sure.
151 2012-11-26 14:27:29 <Luke-Jr> kjj_: whether they are or not, it'll probably change in 0.8
152 2012-11-26 14:27:40 <Luke-Jr> or 0.9
153 2012-11-26 14:28:41 <sipa> Luke-Jr: hardly, i think
154 2012-11-26 14:28:42 <sipa> Luke-Jr: hardly, i think
155 2012-11-26 14:29:14 <Luke-Jr> sipa: when append-only wallet format? :/
156 2012-11-26 14:29:15 <Luke-Jr> sipa: when append-only wallet format? :/
157 2012-11-26 14:29:26 <sipa> bip32 support probably just adds an extra field, append-only formay is just a drop-in replacement for bdb
158 2012-11-26 14:30:56 <Luke-Jr> o
159 2012-11-26 14:30:57 <Luke-Jr> o
160 2012-11-26 14:32:43 <sipa> it's just a key=value list with checksums basically
161 2012-11-26 14:32:44 <sipa> it's just a key=value list with checksums basically
162 2012-11-26 14:33:29 <sipa> though there probably will be some change to the cwallettx serialization, to prevent those from being rewritten on every txout spend
163 2012-11-26 14:33:30 <sipa> though there probably will be some change to the cwallettx serialization, to prevent those from being rewritten on every txout spend
164 2012-11-26 14:33:32 <denisx> sipa: and thats the last nail for the bdb coffin?
165 2012-11-26 14:33:33 <denisx> sipa: and thats the last nail for the bdb coffin?
166 2012-11-26 14:33:42 <sipa> denisx: *YES*
167 2012-11-26 14:33:45 <denisx> yeah!
168 2012-11-26 14:33:46 <denisx> yeah!
169 2012-11-26 14:34:01 <gmaxwell> denisx: we'll still have to have it around for upgrades and imports. Though perhaps it could be pulled into a seperate binary.
170 2012-11-26 14:34:02 <gmaxwell> denisx: we'll still have to have it around for upgrades and imports. Though perhaps it could be pulled into a seperate binary.
171 2012-11-26 14:34:51 <kjj_> Luke-Jr: just because we may upgrade doesn't mean that the old backed up files are going away
172 2012-11-26 14:37:43 <sipa> i'm certainly in favor of moving it to a separate binary
173 2012-11-26 14:37:45 <sipa> i'm certainly in favor of moving it to a separate binary
174 2012-11-26 14:38:37 <kjj_> I have a cron job that makes copies of my wallet files every night.  it would be REALLY nice to be able to suck all of the keys out of those and into a central repository
175 2012-11-26 14:38:38 <kjj_> I have a cron job that makes copies of my wallet files every night.  it would be REALLY nice to be able to suck all of the keys out of those and into a central repository
176 2012-11-26 14:41:05 <Luke-Jr> a wallet conversion CLI tool that GUIs can invoke sounds ideal
177 2012-11-26 14:41:27 <Luke-Jr> kjj_: jgarzik has a pullreq for a simple HTTP encrypted wallet download builtin to the bitcoind server
178 2012-11-26 14:42:00 <sipa> i should try to get walletdump rebased and merged
179 2012-11-26 14:42:16 <kjj_> yeah, that came from a discussion we had about how the backupwallet RPC command didn't work like most people would expect
180 2012-11-26 14:43:24 <BlueMatt> Luke-Jr: did you really need your own standard in ISO42170-A3 for tonal???
181 2012-11-26 14:43:25 <BlueMatt> Luke-Jr: did you really need your own standard in ISO42170-A3 for tonal???
182 2012-11-26 14:44:31 <Luke-Jr> BlueMatt: more important than decimal
183 2012-11-26 14:45:20 <BlueMatt> yes, because you, as an individual, are more important than a few thousand bitcoin users...
184 2012-11-26 14:45:22 <Luke-Jr> certainly more than scamcoins
185 2012-11-26 14:45:30 <BlueMatt> well, that one may be
186 2012-11-26 14:45:35 <Luke-Jr> BlueMatt: what do I as an individual have to do with anythign?
187 2012-11-26 14:45:36 <Luke-Jr> BlueMatt: what do I as an individual have to do with anythign?
188 2012-11-26 14:45:47 <BlueMatt> because you are the only one using tonal bitcoin
189 2012-11-26 14:45:53 <Luke-Jr> not quite.
190 2012-11-26 14:46:05 <BlueMatt> oh, did you finally find a second?
191 2012-11-26 14:46:18 <Luke-Jr> but I'm not going to waste time arguing; nothing productive can come from it
192 2012-11-26 14:46:19 <Luke-Jr> but I'm not going to waste time arguing; nothing productive can come from it
193 2012-11-26 14:48:12 <Luke-Jr> BlueMatt: your goals for Bitcoin and mine are different. different goals aren't a problem, so long as they're legal
194 2012-11-26 14:50:10 <BlueMatt> Luke-Jr: fair, but getting it standardized as a separate currency doesnt make sense
195 2012-11-26 14:51:25 <kjj_> my understanding is that ISO4217 includes a radix size
196 2012-11-26 14:51:26 <kjj_> my understanding is that ISO4217 includes a radix size
197 2012-11-26 14:54:18 <kjj_> and it is really about documents describing currency transactions, rather than currency itself.  so, in a way it makes sense
198 2012-11-26 14:54:19 <kjj_> and it is really about documents describing currency transactions, rather than currency itself.  so, in a way it makes sense
199 2012-11-26 15:09:50 <Luke-Jr> motion to make -printpriority work without -debug
200 2012-11-26 15:09:51 <Luke-Jr> motion to make -printpriority work without -debug
201 2012-11-26 15:11:54 <epscy> denied
202 2012-11-26 15:11:55 <epscy> denied
203 2012-11-26 16:29:15 <jgarzik> ;;bc,halfreward
204 2012-11-26 16:29:16 <gribble> Estimated time of bitcoin block reward halving: Wed Nov 28 12:29:00 2012 | Time remaining: 2 days, 3 hours, and 0 seconds
205 2012-11-26 17:38:39 <BitDev> hi all
206 2012-11-26 17:38:40 <BitDev> hi all
207 2012-11-26 17:39:10 <BitDev> can some one help me
208 2012-11-26 17:39:11 <BitDev> can some one help me
209 2012-11-26 17:40:22 <BitDev> if my client just started and haven't got any block header - how can i ask first block header? i think i must send getblocks packet - but what hash i must put in?
210 2012-11-26 17:43:36 <BitDev> sipa can you help me?
211 2012-11-26 17:43:37 <BitDev> sipa can you help me?
212 2012-11-26 17:44:14 <sipa> you know the genesis block and its hash
213 2012-11-26 17:44:33 <sipa> you should not download the genesis block, but hardxode it
214 2012-11-26 17:44:34 <sipa> you should not download the genesis block, but hardxode it
215 2012-11-26 17:44:56 <BitDev> ow, i must hardcode some block
216 2012-11-26 17:45:11 <BitDev> and where can i find this code?
217 2012-11-26 17:45:15 <BitDev> *hash
218 2012-11-26 17:46:59 <BitDev> i think it must be in sources, but in what file?
219 2012-11-26 17:48:17 <maaku> `bitcoind getbalance` and `bitcoind getbalance \\*` don't return the same value when there are immature coinbase transactions; is this a known issue/expected behavior?
220 2012-11-26 17:48:18 <maaku> `bitcoind getbalance` and `bitcoind getbalance \\*` don't return the same value when there are immature coinbase transactions; is this a known issue/expected behavior?
221 2012-11-26 17:49:58 <gmaxwell> maaku: I was aware of it.
222 2012-11-26 17:49:59 <gmaxwell> maaku: I was aware of it.
223 2012-11-26 17:51:32 <maaku> the fix is simple enough; i'll submit a patch
224 2012-11-26 17:52:59 <gmaxwell> I don't think it is??? rather, I don't think the correct behavior is obvious.
225 2012-11-26 17:53:03 <BitDev> some one knows this genesis block?
226 2012-11-26 17:53:24 <sturles> Everyone know it.
227 2012-11-26 17:53:26 <gmaxwell> BitDev: are you asking about software you wrote?
228 2012-11-26 17:53:27 <gmaxwell> BitDev: are you asking about software you wrote?
229 2012-11-26 17:53:51 <BitDev> no, i am asking about hardcoded genesis block hash
230 2012-11-26 17:53:52 <BitDev> no, i am asking about hardcoded genesis block hash
231 2012-11-26 17:53:58 <BitDev> so i can get all others hashes
232 2012-11-26 17:54:14 <gmaxwell> Get the others using software you wrote. Right?
233 2012-11-26 17:55:36 <BitDev> i writing my software and i want to get all blocks, but for this is need genesis hash to send in getdata packet
234 2012-11-26 17:56:19 <maaku> BitDev: it's at the top of main.cpp
235 2012-11-26 17:57:16 <BitDev> thnx!
236 2012-11-26 17:58:11 <BitDev> i found it ^)
237 2012-11-26 17:58:19 <maaku> gmaxwell: why is that? the comments in the code explicitly state that `getbalance` and `getbalance *` should always be the same (they aren't), and
238 2012-11-26 17:58:20 <maaku> gmaxwell: why is that? the comments in the code explicitly state that `getbalance` and `getbalance *` should always be the same (they aren't), and
239 2012-11-26 17:58:37 <nibor_> http://blockchain.info/block-index/1/000000000019d6689c085ae165831e934ff763ae46a2a6c172b3f1b60a8ce26f
240 2012-11-26 17:59:16 <maaku> maybe it's not obvious whether getbalance should include immature blocks or not, but that choice should be the same whether you include the '*' or not
241 2012-11-26 17:59:23 <gmaxwell> maaku: making them the same is easy (eliminate the special case code so they're actually handled the same). But e.g. the handling of change / unconfirmed change, makes getbalance's correct behavior non-obvious.
242 2012-11-26 17:59:24 <gmaxwell> maaku: making them the same is easy (eliminate the special case code so they're actually handled the same). But e.g. the handling of change / unconfirmed change, makes getbalance's correct behavior non-obvious.
243 2012-11-26 17:59:52 <gmaxwell> maaku: Right, I agree with you??? I'm just saying that this is only one way in which getbalance gives weird results.
244 2012-11-26 18:01:23 <Luke-Jr> next-test 2012-11-25: https://bitcointalk.org/?topic=127864
245 2012-11-26 18:01:53 <gmaxwell> Luke-Jr: wow, you're a rebasing madman to get all that in.
246 2012-11-26 18:01:54 <gmaxwell> Luke-Jr: wow, you're a rebasing madman to get all that in.
247 2012-11-26 18:02:33 <Luke-Jr> gmaxwell: actually, only 5 of the non-disputed/abandoned ones need rebasing right now ;)
248 2012-11-26 18:02:44 <Luke-Jr> but yeah, some of them are annoying to merge
249 2012-11-26 18:06:05 <Luke-Jr> gmaxwell: trying to prepare a master fork for Eligius to test now
250 2012-11-26 18:06:06 <Luke-Jr> gmaxwell: trying to prepare a master fork for Eligius to test now
251 2012-11-26 18:08:24 <Luke-Jr> gmaxwell: my plan is to randomly choose between bitcoind 0.6.0, master, and bitsofproof for making the block, then BIP23-propose it to all of them
252 2012-11-26 18:08:25 <Luke-Jr> gmaxwell: my plan is to randomly choose between bitcoind 0.6.0, master, and bitsofproof for making the block, then BIP23-propose it to all of them
253 2012-11-26 18:08:34 <Luke-Jr> if any fail, log it and failover to whatever 0.6.0 gives
254 2012-11-26 18:08:35 <Luke-Jr> if any fail, log it and failover to whatever 0.6.0 gives
255 2012-11-26 18:08:53 <Luke-Jr> (or maybe just failover until one provides a working template)
256 2012-11-26 18:08:54 <Luke-Jr> (or maybe just failover until one provides a working template)
257 2012-11-26 18:09:56 <Luke-Jr> https://gist.github.com/4149986
258 2012-11-26 18:11:45 <gmaxwell> Luke-Jr: what happened to the pull for the priority tweaking stuff. I recall we went off on shed painting but I forget where it ended up.
259 2012-11-26 18:11:46 <gmaxwell> Luke-Jr: what happened to the pull for the priority tweaking stuff. I recall we went off on shed painting but I forget where it ended up.
260 2012-11-26 18:12:13 <Luke-Jr> gmaxwell: it's still sitting there ready to merge: https://github.com/bitcoin/bitcoin/pull/1583
261 2012-11-26 18:12:32 <Luke-Jr> you asked for a rebase 3 months ago and I took 2 months <.<
262 2012-11-26 18:12:33 <Luke-Jr> you asked for a rebase 3 months ago and I took 2 months <.<
263 2012-11-26 18:12:42 <gmaxwell> heh
264 2012-11-26 18:12:43 <gmaxwell> heh
265 2012-11-26 18:13:14 <sipa> probably less work than having it kept rebased the whole time :)
266 2012-11-26 18:13:15 <sipa> probably less work than having it kept rebased the whole time :)
267 2012-11-26 18:13:38 <Luke-Jr> it's on my eligius checklist, so if you want to merge it now, let me know and I'll hold off on my eligius fork :P
268 2012-11-26 18:13:39 <Luke-Jr> it's on my eligius checklist, so if you want to merge it now, let me know and I'll hold off on my eligius fork :P
269 2012-11-26 18:14:06 <Luke-Jr> (same goes for any on that gist)
270 2012-11-26 18:14:07 <Luke-Jr> (same goes for any on that gist)
271 2012-11-26 18:14:35 <Luke-Jr> in fact, if you want to merge BlueMatt's addnoderpc, I'll even rebase that myself just to get it in
272 2012-11-26 18:14:36 <Luke-Jr> in fact, if you want to merge BlueMatt's addnoderpc, I'll even rebase that myself just to get it in
273 2012-11-26 18:14:42 <gmaxwell> I have my own patches to tend to first when I get a chance.
274 2012-11-26 18:14:43 <gmaxwell> I have my own patches to tend to first when I get a chance.
275 2012-11-26 18:15:00 <kjj> ooh.  it would be nifty if the RPC had commands for adding and removing connections
276 2012-11-26 18:15:07 <gmaxwell> wrt the addnode one, I'm still really unhappy about the remove semantics with hostnames.
277 2012-11-26 18:15:08 <Luke-Jr> kjj: that's what addnoderpc is
278 2012-11-26 18:15:09 <Luke-Jr> kjj: that's what addnoderpc is
279 2012-11-26 18:15:37 <kjj> that's what I figured, that's why I was excited to see it
280 2012-11-26 18:16:20 <sipa> i'd really just prefer a call that said "connect now to X"
281 2012-11-26 18:16:50 <Luke-Jr> sipa: it does that too :p
282 2012-11-26 18:17:06 <Luke-Jr> addnode onetry|add|remove <hostname>
283 2012-11-26 18:17:27 <sipa> i'll have a look soon
284 2012-11-26 18:17:49 <gmaxwell> The thing I don't like about it is that if you remove <hostname> it resolves the hostname and then kills everyting it returns.
285 2012-11-26 18:17:50 <gmaxwell> The thing I don't like about it is that if you remove <hostname> it resolves the hostname and then kills everyting it returns.
286 2012-11-26 18:18:00 <Luke-Jr> gmaxwell: it doesn't look like it does to me
287 2012-11-26 18:18:23 <Luke-Jr> vAddedNodes.erase(it);
288 2012-11-26 18:18:24 <Luke-Jr> vAddedNodes.erase(it);
289 2012-11-26 18:18:25 <gmaxwell> oh, well I thought it did. Been a long time since I looked at it.
290 2012-11-26 18:18:28 <Luke-Jr> oh
291 2012-11-26 18:18:37 <Luke-Jr> maybe before BlueMatt changed stuff based on your comments ;P
292 2012-11-26 18:18:38 <Luke-Jr> maybe before BlueMatt changed stuff based on your comments ;P
293 2012-11-26 18:20:43 <BlueMatt> gmaxwell: did it ever do that? oh well, whatever
294 2012-11-26 18:21:03 <gmaxwell> BlueMatt: maybe? we're talking about my memory here. It's not known to be very reliable.
295 2012-11-26 18:21:04 <gmaxwell> BlueMatt: maybe? we're talking about my memory here. It's not known to be very reliable.
296 2012-11-26 18:21:08 <Luke-Jr> XD
297 2012-11-26 18:21:09 <Luke-Jr> XD
298 2012-11-26 18:21:13 <BlueMatt> heh
299 2012-11-26 18:21:19 <gmaxwell> I remember that there was something about the behavior I didn't like but also didn't see a clean way of fixing.
300 2012-11-26 18:21:20 <gmaxwell> I remember that there was something about the behavior I didn't like but also didn't see a clean way of fixing.
301 2012-11-26 18:21:33 <gmaxwell> But I might have also misunderstood it. (though IIRC, I did also try it out)
302 2012-11-26 18:21:34 <gmaxwell> But I might have also misunderstood it. (though IIRC, I did also try it out)
303 2012-11-26 18:21:44 <BlueMatt> yes, I remember a few of those, but I dont remember the what color the bike shed ended up on that pull...
304 2012-11-26 18:21:44 <Luke-Jr> BlueMatt: should I pastebin the rebase so you can push it? :P
305 2012-11-26 18:21:45 <BlueMatt> yes, I remember a few of those, but I dont remember the what color the bike shed ended up on that pull...
306 2012-11-26 18:22:51 <BlueMatt> Luke-Jr: feel free to post a link to your branch on the pull comments, but I may not push an update for a while (if anyone wants it they can manually pull...)
307 2012-11-26 18:22:52 <BlueMatt> Luke-Jr: feel free to post a link to your branch on the pull comments, but I may not push an update for a while (if anyone wants it they can manually pull...)
308 2012-11-26 18:23:27 <Luke-Jr> BlueMatt: I'd have to make the branch. I'll wait until someone wants it I guess
309 2012-11-26 18:23:28 <Luke-Jr> BlueMatt: I'd have to make the branch. I'll wait until someone wants it I guess
310 2012-11-26 18:23:34 <Luke-Jr> (right now, it's a diff of the merge into next)
311 2012-11-26 18:23:53 <BlueMatt> mmm
312 2012-11-26 18:51:29 <jgarzik> Luke-Jr: I'm not understanding your github comments.  What GetMinFee(), DoS-related stuff was eliminated in __block__ acceptance?
313 2012-11-26 18:51:41 <jgarzik> seems like blocks accept whatever the miner stuffs in there, fees or so
314 2012-11-26 18:51:51 <jgarzik> *no
315 2012-11-26 18:53:29 <Luke-Jr> jgarzik: acceptance of transactions into blocks we create
316 2012-11-26 19:42:38 <Luke-Jr> BlueMatt: when you get some time, gmaxwell suggests I include your Java full node thing in Eligius's round robin of block makers
317 2012-11-26 19:42:39 <Luke-Jr> BlueMatt: when you get some time, gmaxwell suggests I include your Java full node thing in Eligius's round robin of block makers
318 2012-11-26 19:47:52 <helo> you get you cycle through getting your blocks from different implementations?
319 2012-11-26 19:47:53 <helo> you get you cycle through getting your blocks from different implementations?
320 2012-11-26 19:48:00 <helo> s/you get //
321 2012-11-26 19:48:01 <helo> s/you get //
322 2012-11-26 19:48:16 <Luke-Jr> helo: that's the plan
323 2012-11-26 19:48:17 <Luke-Jr> helo: that's the plan
324 2012-11-26 19:48:22 <gmaxwell> helo: then he proposes the blocks to the other implementations to see if they're happy with them.
325 2012-11-26 19:48:23 <gmaxwell> helo: then he proposes the blocks to the other implementations to see if they're happy with them.
326 2012-11-26 19:48:23 <Luke-Jr> helo: setting it up with 0.6 and master now
327 2012-11-26 19:48:24 <Luke-Jr> helo: setting it up with 0.6 and master now
328 2012-11-26 19:48:45 <helo> ahh, the proposing part ++
329 2012-11-26 19:48:50 <Luke-Jr> yeah, for testing the implementation
330 2012-11-26 19:49:05 <Luke-Jr> this way, if master or bitsofproof make blocks 0.6 rejects (or vice versa), we find out
331 2012-11-26 19:49:06 <Luke-Jr> this way, if master or bitsofproof make blocks 0.6 rejects (or vice versa), we find out
332 2012-11-26 19:49:09 <Luke-Jr> without orphans :P
333 2012-11-26 19:49:10 <Luke-Jr> without orphans :P
334 2012-11-26 19:49:26 <Luke-Jr> I should throw pynode in that mix too
335 2012-11-26 19:49:27 <Luke-Jr> I should throw pynode in that mix too
336 2012-11-26 19:49:39 <helo> kind of vital to avoid forking badness if different implementations become popular, isn't it?
337 2012-11-26 19:49:39 <Luke-Jr> but first I need to port to Python3 probably
338 2012-11-26 19:49:40 <helo> kind of vital to avoid forking badness if different implementations become popular, isn't it?
339 2012-11-26 19:49:40 <Luke-Jr> but first I need to port to Python3 probably
340 2012-11-26 19:49:43 <Luke-Jr> yep
341 2012-11-26 19:49:44 <Luke-Jr> yep
342 2012-11-26 19:49:53 <Luke-Jr> that's why Satoshi was very opposed to third-party implementations
343 2012-11-26 19:50:16 <sipa> does bitsofproof and bitcoinj have gbt?
344 2012-11-26 19:50:20 <sipa> *do
345 2012-11-26 19:50:32 <TD> no
346 2012-11-26 19:50:36 <Luke-Jr> sipa: bitsofproof has it on their near-term plan at least
347 2012-11-26 19:50:37 <Luke-Jr> sipa: bitsofproof has it on their near-term plan at least
348 2012-11-26 19:50:44 <sipa> yes, i know that
349 2012-11-26 19:50:45 <sipa> yes, i know that
350 2012-11-26 19:50:55 <Luke-Jr> hence why I'm just doing 0.6 and master right now :P
351 2012-11-26 19:50:56 <Luke-Jr> hence why I'm just doing 0.6 and master right now :P
352 2012-11-26 19:51:11 <helo> he's still kind of right, even with round-robin testing... just because past blocks wouldn't have caused a fork, doesn't mean future blocks won't
353 2012-11-26 19:51:12 <helo> he's still kind of right, even with round-robin testing... just because past blocks wouldn't have caused a fork, doesn't mean future blocks won't
354 2012-11-26 19:51:25 <helo> but there's no way to prevent third-party implementations, so might as well mitigate what we can
355 2012-11-26 19:51:43 <Luke-Jr> helo: yes, but better to catch some bugs than miss them all
356 2012-11-26 19:51:44 <Luke-Jr> helo: yes, but better to catch some bugs than miss them all
357 2012-11-26 19:51:58 <Luke-Jr> by using proposals, we catch them much faster and cheaper
358 2012-11-26 19:51:59 <Luke-Jr> by using proposals, we catch them much faster and cheaper
359 2012-11-26 19:52:10 <Luke-Jr> since we DO catch future blocks
360 2012-11-26 19:52:11 <Luke-Jr> since we DO catch future blocks
361 2012-11-26 19:52:41 <Luke-Jr> eg, if any implementation rejects the proposed block, Eligius just won't mine it
362 2012-11-26 19:52:42 <Luke-Jr> eg, if any implementation rejects the proposed block, Eligius just won't mine it
363 2012-11-26 19:53:18 <jgarzik> Luke-Jr: next-test sure does have a lot of pull requests from me, that I've long closed and abandoned...
364 2012-11-26 19:53:28 <Luke-Jr> jgarzik: I think I noted them as abandoned
365 2012-11-26 19:53:29 <Luke-Jr> jgarzik: I think I noted them as abandoned
366 2012-11-26 19:53:33 <helo> will it relay such blocks?
367 2012-11-26 19:53:43 <Luke-Jr> helo: I'm not doing anything special with block relaying.
368 2012-11-26 19:54:18 <Luke-Jr> jgarzik: I do check to be sure they should still work before merging them, of course.
369 2012-11-26 19:54:40 <jgarzik> Luke-Jr: not that I can tell, but then I do have red-green colorblindness, and followed-link browser coloring screws things up too
370 2012-11-26 19:54:42 <jgarzik> Luke-Jr: not that I can tell, but then I do have red-green colorblindness, and followed-link browser coloring screws things up too
371 2012-11-26 19:54:51 <sipa> helo: only blocks accepted in the new best chain are relayed
372 2012-11-26 19:54:52 <sipa> helo: only blocks accepted in the new best chain are relayed
373 2012-11-26 19:55:09 <Luke-Jr> jgarzik: the abandoned ones are grey, and at least in my browser, the colours override the followed-link one :/
374 2012-11-26 19:55:48 <Luke-Jr> 1471, 1538, 1781, 1787
375 2012-11-26 19:55:49 <Luke-Jr> 1471, 1538, 1781, 1787
376 2012-11-26 19:56:04 <sipa> but are abandoned ones included?
377 2012-11-26 19:56:05 <sipa> but are abandoned ones included?
378 2012-11-26 19:56:12 <Luke-Jr> sipa: in next-test, not next
379 2012-11-26 19:56:13 <Luke-Jr> sipa: in next-test, not next
380 2012-11-26 19:56:14 <jgarzik> sipa: yes
381 2012-11-26 19:56:15 <jgarzik> sipa: yes
382 2012-11-26 19:56:18 <sipa> why?
383 2012-11-26 19:56:34 <Luke-Jr> sipa: doesn't hurt to test them
384 2012-11-26 19:56:36 <jgarzik> Luke-Jr: sipa's valgrinding also detected a problem with the header/body P2P split pull req
385 2012-11-26 19:57:10 <sipa> Luke-Jr: btw master may segfault when running for a long time and mining
386 2012-11-26 19:58:00 <Luke-Jr> sipa: good to know; I'll be sure to tolerate nodes being down I guess >_<
387 2012-11-26 19:58:58 <sipa> #2033 should fix that
388 2012-11-26 19:58:59 <sipa> #2033 should fix that
389 2012-11-26 19:59:12 <sipa> but only indirectly
390 2012-11-26 19:59:13 <sipa> but only indirectly
391 2012-11-26 20:01:06 <sipa> (the bug is lookup up amount/confirmations of outputs being spent in createnewblock, but apparently (and even on 0.7.x!) orphans can end up in the mempool
392 2012-11-26 20:01:22 <sipa> with non-existing outputs being queried
393 2012-11-26 20:01:23 <sipa> with non-existing outputs being queried
394 2012-11-26 20:02:42 <Luke-Jr> https://github.com/luke-jr/bitcoin/compare/0.8.0.eligius
395 2012-11-26 20:02:43 <Luke-Jr> https://github.com/luke-jr/bitcoin/compare/0.8.0.eligius
396 2012-11-26 20:03:18 <Luke-Jr> I knew orphans could land in mempool before ;)
397 2012-11-26 20:03:19 <Luke-Jr> I knew orphans could land in mempool before ;)
398 2012-11-26 20:04:52 <sipa> yes, i knew it could happen during reorgs
399 2012-11-26 20:04:53 <sipa> yes, i knew it could happen during reorgs
400 2012-11-26 20:05:45 <sipa> but not because of succesful conflicting transactions
401 2012-11-26 20:05:46 <sipa> but not because of succesful conflicting transactions
402 2012-11-26 20:16:39 <Luke-Jr> sipa: bug report: bitcoin_test is initializing .bitcoin and blocks db
403 2012-11-26 20:18:16 <Luke-Jr> do I need to open a github issue? :P
404 2012-11-26 20:18:17 <Luke-Jr> do I need to open a github issue? :P
405 2012-11-26 20:18:33 <sipa> hmm how did it prevent doing that in the past?
406 2012-11-26 20:18:55 <Luke-Jr> MockDB
407 2012-11-26 20:18:56 <Luke-Jr> MockDB
408 2012-11-26 20:18:59 <sipa> the database should be memenv
409 2012-11-26 20:19:00 <sipa> the database should be memenv
410 2012-11-26 20:19:08 <Luke-Jr> LevelDB bug then? :/
411 2012-11-26 20:19:09 <Luke-Jr> LevelDB bug then? :/
412 2012-11-26 20:19:15 <sipa> doubtful
413 2012-11-26 20:19:16 <Luke-Jr> it also makes blk00000.dat
414 2012-11-26 20:19:17 <Luke-Jr> it also makes blk00000.dat
415 2012-11-26 20:19:58 <Luke-Jr> sipa: bug #2: read-only blk00000.dat starts; no idea where it'd put new blocks O.o
416 2012-11-26 20:19:59 <Luke-Jr> sipa: bug #2: read-only blk00000.dat starts; no idea where it'd put new blocks O.o
417 2012-11-26 20:20:14 <sipa> ?
418 2012-11-26 20:20:15 <sipa> ?
419 2012-11-26 20:20:44 <Luke-Jr> err
420 2012-11-26 20:20:49 <Luke-Jr> I meant non-readable
421 2012-11-26 20:20:50 <Luke-Jr> I meant non-readable
422 2012-11-26 20:20:51 <Luke-Jr> ???
423 2012-11-26 20:20:52 <Luke-Jr> ???
424 2012-11-26 20:21:05 <Luke-Jr> bug #3: read-only blk00000.dat still fails to open :/
425 2012-11-26 20:21:06 <Luke-Jr> bug #3: read-only blk00000.dat still fails to open :/
426 2012-11-26 20:21:26 <maaku> miner_tests.cpp -- anyone know who generated the blockinfo nonces?
427 2012-11-26 20:21:38 <sipa> is that master, Luke-Jr?
428 2012-11-26 20:22:02 <Luke-Jr> sipa: master-based, yes
429 2012-11-26 20:22:03 <Luke-Jr> sipa: master-based, yes
430 2012-11-26 20:22:10 <Luke-Jr> maaku: I did.
431 2012-11-26 20:22:17 <sipa> Luke-Jr: can you test whether the problem is also in master?
432 2012-11-26 20:22:18 <sipa> Luke-Jr: can you test whether the problem is also in master?
433 2012-11-26 20:23:47 <Luke-Jr> building???
434 2012-11-26 20:23:53 <sipa> Luke-Jr: i also don't understand your bug reports here
435 2012-11-26 20:23:54 <sipa> Luke-Jr: i also don't understand your bug reports here
436 2012-11-26 20:24:01 <Luke-Jr> :/
437 2012-11-26 20:24:34 <Luke-Jr> bug #4: master segfaults test_bitcoin at least in this case
438 2012-11-26 20:24:36 <Luke-Jr> unknown location(0): fatal error in "CreateNewBlock_validity": memory access violation at address: 0x00000018: no mapping at fault address
439 2012-11-26 20:24:37 <Luke-Jr> test/miner_tests.cpp(58): last checkpoint
440 2012-11-26 20:24:38 <Luke-Jr> test/miner_tests.cpp(58): last checkpoint
441 2012-11-26 20:25:51 <Luke-Jr> ok, I'll go into more detail on github as I confirm them on master
442 2012-11-26 20:25:52 <Luke-Jr> ok, I'll go into more detail on github as I confirm them on master
443 2012-11-26 20:27:38 <sipa> i assuke somehow running test overwrote part of your blocks files
444 2012-11-26 20:27:39 <sipa> i assuke somehow running test overwrote part of your blocks files
445 2012-11-26 20:27:52 <sipa> and that that broke normal ooeration
446 2012-11-26 20:27:57 <Luke-Jr> impossible ???
447 2012-11-26 20:27:58 <Luke-Jr> impossible ???
448 2012-11-26 20:28:13 <sipa> ?
449 2012-11-26 20:28:14 <Luke-Jr> I run test automatically at every build, and I'm doing something silly-useful here in the first place
450 2012-11-26 20:28:14 <sipa> ?
451 2012-11-26 20:28:15 <Luke-Jr> I run test automatically at every build, and I'm doing something silly-useful here in the first place
452 2012-11-26 20:29:48 <sipa> chmod -r what mode?
453 2012-11-26 20:29:49 <sipa> chmod -r what mode?
454 2012-11-26 20:30:13 <Luke-Jr> -r is the mode???
455 2012-11-26 20:30:14 <Luke-Jr> -r is the mode???
456 2012-11-26 20:30:26 <Luke-Jr> removes the read bit
457 2012-11-26 20:30:52 <sipa> oh, confused with -R then, nvm
458 2012-11-26 20:30:53 <sipa> oh, confused with -R then, nvm
459 2012-11-26 20:33:00 <Luke-Jr> https://github.com/bitcoin/bitcoin/issues/2039 is the big problem for me atm
460 2012-11-26 20:33:23 <Luke-Jr> as I want to hardlink blk0001.dat and blk0002.dat from my 0.6.0 client (running as a trusted user)
461 2012-11-26 20:33:24 <Luke-Jr> as I want to hardlink blk0001.dat and blk0002.dat from my 0.6.0 client (running as a trusted user)
462 2012-11-26 20:35:44 <sipa> i suppose it should just try opening the blockfile in ro mode when reindexing
463 2012-11-26 20:37:47 <Luke-Jr> it does :/
464 2012-11-26 20:38:10 <sipa> ok, i'll have a look tonifght
465 2012-11-26 20:38:11 <sipa> ok, i'll have a look tonifght
466 2012-11-26 20:39:33 <Luke-Jr> wtf
467 2012-11-26 20:39:58 <Luke-Jr> OpenDiskFile interprets fReadOnly as "truncate the file"
468 2012-11-26 20:39:59 <Luke-Jr> OpenDiskFile interprets fReadOnly as "truncate the file"
469 2012-11-26 20:40:02 <Luke-Jr> how does this even work? :/
470 2012-11-26 20:40:25 <Luke-Jr> oh, it only tries to truncate if opening it for r+w failed
471 2012-11-26 20:40:26 <Luke-Jr> oh, it only tries to truncate if opening it for r+w failed
472 2012-11-26 20:40:32 <Luke-Jr> this is insane???
473 2012-11-26 20:40:33 <Luke-Jr> this is insane???
474 2012-11-26 20:41:18 <Luke-Jr> committed by sipa Aug 13 @ 7:11 PM
475 2012-11-26 20:41:19 <Luke-Jr> committed by sipa Aug 13 @ 7:11 PM
476 2012-11-26 20:41:20 <Luke-Jr> O.o
477 2012-11-26 20:41:27 <Luke-Jr> 
ACTION was expecting a 2 AM commit there :P

478 2012-11-26 20:44:07 <korozion> someone was telling me a while ago about ati cards and the drivers only supporting 4 cards on a single system
479 2012-11-26 20:44:21 <sipa> hmm, that just seems to prevent creating a new file if fReadOnly
480 2012-11-26 20:44:33 <sipa> not prevent opening it for writing or truncating
481 2012-11-26 20:44:34 <sipa> not prevent opening it for writing or truncating
482 2012-11-26 20:44:40 <gmaxwell> korozion: they support 8 GPUs, thats 4 dual gpu cards.
483 2012-11-26 20:45:10 <korozion> OK cool, I just wanted to confirm
484 2012-11-26 20:45:11 <korozion> OK cool, I just wanted to confirm
485 2012-11-26 20:45:24 <korozion> I was working on mining on Linux, changed to Windows to see if I could solve some problems
486 2012-11-26 20:47:49 <korozion> I think I only managed to get three cards working on Linux :(
487 2012-11-26 20:50:20 <Luke-Jr> sipa: the difference between rb+ and wb+ is truncating
488 2012-11-26 20:50:21 <Luke-Jr> sipa: the difference between rb+ and wb+ is truncating
489 2012-11-26 20:51:03 <Luke-Jr> korozion: you can workaround that 8 limit with a good motherboard too ;)
490 2012-11-26 20:51:10 <korozion> oh?
491 2012-11-26 20:51:11 <Luke-Jr> korozion: I run all my GPU miners in KVM instances
492 2012-11-26 20:51:12 <Luke-Jr> korozion: I run all my GPU miners in KVM instances
493 2012-11-26 20:51:17 <korozion> ahhh nice idea
494 2012-11-26 20:51:18 <korozion> ahhh nice idea
495 2012-11-26 20:51:18 <Luke-Jr> with PCIe passthrough
496 2012-11-26 20:51:26 <Luke-Jr> my host OS is 100% open source
497 2012-11-26 20:51:53 <Luke-Jr> sipa: confirmed my pullreq fixes at least #2039
498 2012-11-26 20:51:54 <Luke-Jr> sipa: confirmed my pullreq fixes at least #2039
499 2012-11-26 20:52:28 <Luke-Jr> jgarzik: do you really need the debug.log for #2039 now that I've identified and fixed the problem? <.<
500 2012-11-26 20:52:29 <Luke-Jr> jgarzik: do you really need the debug.log for #2039 now that I've identified and fixed the problem? <.<
501 2012-11-26 20:53:21 <sipa> Luke-Jr: the reason was to avoid seeking to the end of the file, which a/a+ do
502 2012-11-26 20:53:29 <korozion> is 2000 mhash decent?
503 2012-11-26 20:53:30 <korozion> is 2000 mhash decent?
504 2012-11-26 20:53:36 <Luke-Jr> bah! crash after reindex
505 2012-11-26 20:53:37 <Luke-Jr> bah! crash after reindex
506 2012-11-26 20:53:43 <sipa> i wanted "open rw, create if necessary, position at beginnen of file"
507 2012-11-26 20:53:44 <sipa> i wanted "open rw, create if necessary, position at beginnen of file"
508 2012-11-26 20:53:51 <sipa> i think i just missed the readonly use case
509 2012-11-26 20:54:03 <Luke-Jr> sipa: rb+ positions at the beginning
510 2012-11-26 20:54:04 <Luke-Jr> sipa: rb+ positions at the beginning
511 2012-11-26 20:54:12 <sipa> yes, but doesn't create
512 2012-11-26 20:54:15 <Luke-Jr> no?
513 2012-11-26 20:54:27 <Luke-Jr> bah
514 2012-11-26 20:54:28 <Luke-Jr> bah
515 2012-11-26 20:54:35 <Luke-Jr> ok, will fix
516 2012-11-26 20:54:36 <Luke-Jr> ok, will fix
517 2012-11-26 20:54:38 <sipa> well, i didn't test that - fopen(3) mentions creating for w+ but not for r+
518 2012-11-26 20:55:23 <Luke-Jr> what should be the correct behaviour for "file doesn't exist, read only"?
519 2012-11-26 20:55:36 <sipa> return error
520 2012-11-26 20:55:42 <Luke-Jr> k
521 2012-11-26 20:55:43 <Luke-Jr> k
522 2012-11-26 20:56:09 <Luke-Jr> sipa: would you rather seek to the end, or have a race condition to truncate? :/
523 2012-11-26 20:56:22 <sipa> i suppose that the mess isn't worth it
524 2012-11-26 20:56:23 <sipa> i suppose that the mess isn't worth it
525 2012-11-26 20:56:33 <Luke-Jr> ?
526 2012-11-26 20:57:33 <Luke-Jr> sipa: which one? <.<
527 2012-11-26 20:57:34 <Luke-Jr> sipa: which one? <.<
528 2012-11-26 20:57:35 <sipa> seeking is easier, i think
529 2012-11-26 20:57:36 <sipa> seeking is easier, i think
530 2012-11-26 20:58:32 <Luke-Jr> hmm
531 2012-11-26 20:58:36 <Luke-Jr> seeking won't work.
532 2012-11-26 20:58:41 <Luke-Jr> ugh, stupid fopen
533 2012-11-26 20:58:41 <sipa> ?
534 2012-11-26 20:58:42 <Luke-Jr> ugh, stupid fopen
535 2012-11-26 20:58:43 <sipa> i tried to avoid an unneccessary(?) seek, but probably mode="ab" + seek back to beginnen is easier
536 2012-11-26 20:58:44 <sipa> i tried to avoid an unneccessary(?) seek, but probably mode="ab" + seek back to beginnen is easier
537 2012-11-26 20:58:52 <sipa> *beginnING
538 2012-11-26 20:59:03 <Luke-Jr> "The initial file position for reading is at the beginning of the file, but output is always appended to the end of the file."
539 2012-11-26 20:59:04 <Luke-Jr> "The initial file position for reading is at the beginning of the file, but output is always appended to the end of the file."
540 2012-11-26 20:59:11 <Luke-Jr> *maybe* that works for us?
541 2012-11-26 20:59:42 <sipa> i have no idea how that would work, so i prefer not to touch it
542 2012-11-26 20:59:50 <Luke-Jr> yeah, sounds like a mess on its own
543 2012-11-26 21:00:02 <sipa> but ab is perfect, no?
544 2012-11-26 21:00:03 <sipa> but ab is perfect, no?
545 2012-11-26 21:00:08 <sipa> apart from the seek to the end
546 2012-11-26 21:00:17 <Luke-Jr> can't read
547 2012-11-26 21:00:18 <Luke-Jr> can't read
548 2012-11-26 21:00:26 <sipa> oh, right
549 2012-11-26 21:00:27 <sipa> oh, right
550 2012-11-26 21:00:29 <sipa> grrr
551 2012-11-26 21:00:30 <sipa> grrr
552 2012-11-26 21:00:37 <sipa> can't we just use open(2) ?
553 2012-11-26 21:00:53 <sipa> who designed that fopen interface
554 2012-11-26 21:01:07 <sipa> he deserves using edlin for the rest of his life
555 2012-11-26 21:01:08 <sipa> he deserves using edlin for the rest of his life
556 2012-11-26 21:01:12 <Luke-Jr> if we use open(2), we're platform-dependent for fdopen
557 2012-11-26 21:01:17 <Luke-Jr> no kidding
558 2012-11-26 21:01:17 <sipa> of course
559 2012-11-26 21:01:18 <Luke-Jr> no kidding
560 2012-11-26 21:01:40 <jgarzik> Luke-Jr: huh?
561 2012-11-26 21:01:55 <sipa> we can't depend on open(2)
562 2012-11-26 21:01:56 <sipa> we can't depend on open(2)
563 2012-11-26 21:02:02 <jgarzik> Luke-Jr: if you use fdopen, you do not use open, and vice versa
564 2012-11-26 21:02:06 <jgarzik> Luke-Jr: why not?
565 2012-11-26 21:02:08 <jgarzik> er
566 2012-11-26 21:02:10 <jgarzik> sipa: why not?
567 2012-11-26 21:02:11 <jgarzik> sipa: why not?
568 2012-11-26 21:02:33 <sipa> jgarzik: does that even exist on windows?
569 2012-11-26 21:02:37 <Luke-Jr> well, without rewriting all the stuff using OpenDataFile
570 2012-11-26 21:02:39 <jgarzik> sipa: yes
571 2012-11-26 21:03:06 <Luke-Jr> 
ACTION is tempted to look at what we did in 0.7 and earlier and just compromise on that solution\\

572 2012-11-26 21:03:07 <Luke-Jr> 
ACTION is tempted to look at what we did in 0.7 and earlier and just compromise on that solution\\

573 2012-11-26 21:03:28 <sipa> Luke-Jr: mode string being passed into OpenBlockFile
574 2012-11-26 21:03:30 <sipa> ugly!
575 2012-11-26 21:03:36 <Luke-Jr> sipa: yes, but what strings? :P
576 2012-11-26 21:04:02 <Luke-Jr> if "ab", then we might not need to worry about unreadableness
577 2012-11-26 21:04:03 <Luke-Jr> if "ab", then we might not need to worry about unreadableness
578 2012-11-26 21:04:23 <sipa> "ab", "rb", "rb+"
579 2012-11-26 21:04:34 <sipa> depends on call site
580 2012-11-26 21:04:38 <Luke-Jr> sigh
581 2012-11-26 21:04:39 <Luke-Jr> sigh
582 2012-11-26 21:05:53 <Luke-Jr> which site needed rb+?
583 2012-11-26 21:06:13 <sipa> jgarzik: do you know the exact interpretation of that explanation of a+ in fopen(3) ?
584 2012-11-26 21:06:14 <sipa> jgarzik: do you know the exact interpretation of that explanation of a+ in fopen(3) ?
585 2012-11-26 21:07:03 <Luke-Jr> jgarzik: you sure strerror's thread-unsafety won't be a problem?
586 2012-11-26 21:07:34 <Luke-Jr> sipa: I wouldn't want to depend on platform-specific behaviour that goes against docs :/
587 2012-11-26 21:07:35 <Luke-Jr> sipa: I wouldn't want to depend on platform-specific behaviour that goes against docs :/
588 2012-11-26 21:08:42 <Luke-Jr> 
ACTION is tempted to just try "a+" and let someone report if that causes any problems he doesn't notice off-the-bat???

589 2012-11-26 21:08:43 <Luke-Jr> 
ACTION is tempted to just try "a+" and let someone report if that causes any problems he doesn't notice off-the-bat???

590 2012-11-26 21:09:05 <sipa> a+ + seek-to-begin?
591 2012-11-26 21:09:06 <sipa> a+ + seek-to-begin?
592 2012-11-26 21:10:14 <Luke-Jr> sipa: right
593 2012-11-26 21:10:15 <Luke-Jr> sipa: right
594 2012-11-26 21:10:58 <Luke-Jr> in other news, my crash during -reindex is caused by SetBestChain trying to FlushBlockFile and that not checking if the OpenBlockFile failed??? can I just ignore failures there? <.<
595 2012-11-26 21:12:14 <Luke-Jr> 
ACTION wonders if fdatasync needs write mode

596 2012-11-26 21:12:15 <Luke-Jr> 
ACTION wonders if fdatasync needs write mode

597 2012-11-26 21:14:15 <jgarzik> sipa: see, e.g., easy examples in our own levdldb source:
598 2012-11-26 21:14:16 <jgarzik> sipa: see, e.g., easy examples in our own levdldb source:
599 2012-11-26 21:14:17 <jgarzik> #ifdef LEVELDB_PLATFORM_WINDOWS
600 2012-11-26 21:14:18 <jgarzik> #else
601 2012-11-26 21:14:42 <jgarzik> 
ACTION wouldn't waste any more time on FILE*

602 2012-11-26 21:14:43 <jgarzik> 
ACTION wouldn't waste any more time on FILE*

603 2012-11-26 21:15:40 <jgarzik> Luke-Jr: <shrug> then use sys_errlist
604 2012-11-26 21:15:45 <jgarzik> which is also on all platforms
605 2012-11-26 21:15:59 <jgarzik> google for "windows strerror" for useful MSDN links
606 2012-11-26 21:16:48 <jgarzik> Luke-Jr: SetBestChain should know that it does not need to flush a read-only operation
607 2012-11-26 21:16:57 <jgarzik> my -reindex drilled down to that level, for that purpose
608 2012-11-26 21:18:35 <Luke-Jr> jgarzik: I don't see any evidence it knows it's read-only at all.
609 2012-11-26 21:18:40 <Luke-Jr> its only argument is a CBlockIndex.
610 2012-11-26 21:18:41 <Luke-Jr> its only argument is a CBlockIndex.
611 2012-11-26 21:18:55 <Luke-Jr> although, the crash is only at the end of blk00000.dat it seems, so maybe that's related?
612 2012-11-26 21:18:56 <Luke-Jr> although, the crash is only at the end of blk00000.dat it seems, so maybe that's related?
613 2012-11-26 21:20:13 <jgarzik> Luke-Jr: I did not say it -does- know.  I said it -should- know.  There is simply no block data being written, therefore it is not a "read-write" versus "read-only" situation; that is just a symptom of the file open mode.  The situation is "we can, and must write data" versus "we never touch that file at all"
614 2012-11-26 21:20:14 <jgarzik> Luke-Jr: I did not say it -does- know.  I said it -should- know.  There is simply no block data being written, therefore it is not a "read-write" versus "read-only" situation; that is just a symptom of the file open mode.  The situation is "we can, and must write data" versus "we never touch that file at all"
615 2012-11-26 21:20:45 <Luke-Jr> hmm
616 2012-11-26 21:21:01 <jgarzik> Luke-Jr: SetBestChain simply should never know, nor touch, the block file in the -reindex case
617 2012-11-26 21:21:02 <jgarzik> Luke-Jr: SetBestChain simply should never know, nor touch, the block file in the -reindex case
618 2012-11-26 21:21:26 <jgarzik> Luke-Jr: See my old implementation; if a file number and position is passed in, a FILE* is never accessed.
619 2012-11-26 21:21:44 <jgarzik> Granted, maybe sipa requires read-only access in order to write the undo file.  I cannot say.
620 2012-11-26 21:22:16 <Luke-Jr> 
ACTION uses fImporting <.<

621 2012-11-26 21:22:22 <Luke-Jr> it already special-cased fIsInitialDownload
622 2012-11-26 21:22:23 <Luke-Jr> it already special-cased fIsInitialDownload
623 2012-11-26 21:51:31 <Luke-Jr> interesting, my fork got stuck on 188183
624 2012-11-26 21:54:09 <sipa> define stuck
625 2012-11-26 21:54:10 <sipa> define stuck
626 2012-11-26 21:54:38 <Luke-Jr> sipa: it says 188184 is invalid
627 2012-11-26 21:54:39 <Luke-Jr> sipa: it says 188184 is invalid
628 2012-11-26 21:54:42 <sipa> :o
629 2012-11-26 21:54:43 <sipa> :o
630 2012-11-26 21:54:48 <Luke-Jr> going to bisect it of course
631 2012-11-26 21:54:49 <Luke-Jr> going to bisect it of course
632 2012-11-26 21:55:02 <sipa> only your fork, or master too?
633 2012-11-26 21:55:03 <sipa> only your fork, or master too?
634 2012-11-26 21:55:07 <Luke-Jr> pretty sure master works
635 2012-11-26 21:55:08 <Luke-Jr> pretty sure master works
636 2012-11-26 21:55:10 <Luke-Jr> I'll recheck
637 2012-11-26 21:55:11 <Luke-Jr> I'll recheck
638 2012-11-26 21:59:26 <Luke-Jr> sipa: hmm, possible lack of pblocktree->Sync or view.Flush would do that?
639 2012-11-26 21:59:27 <Luke-Jr> sipa: hmm, possible lack of pblocktree->Sync or view.Flush would do that?
640 2012-11-26 21:59:52 <Luke-Jr> ah crap, -printtoconsole stops debug.log too :<
641 2012-11-26 21:59:53 <Luke-Jr> ah crap, -printtoconsole stops debug.log too :<
642 2012-11-26 22:01:01 <sipa> Luke-Jr: whenever you do a pcoinsTip->Flush(), you must first do a pblocktree->Sync()
643 2012-11-26 22:01:16 <Luke-Jr> sipa: yes, but I was skipping both while importing
644 2012-11-26 22:01:49 <sipa> that shouldn't be a problem, as you'd just be doing everything in RAM
645 2012-11-26 22:01:50 <sipa> that shouldn't be a problem, as you'd just be doing everything in RAM
646 2012-11-26 22:02:18 <Luke-Jr> the entire block chain to date can be done in RAM?
647 2012-11-26 22:02:19 <Luke-Jr> the entire block chain to date can be done in RAM?
648 2012-11-26 22:02:33 <sipa> sure
649 2012-11-26 22:03:34 <sipa> the coin database is 130 MB in stored form... i suppose maybe 4x as large unpacked as CCoins objects in RAM
650 2012-11-26 22:03:35 <sipa> the coin database is 130 MB in stored form... i suppose maybe 4x as large unpacked as CCoins objects in RAM
651 2012-11-26 22:13:00 <sipa> ;;bc,blocks
652 2012-11-26 22:13:01 <gribble> 209731
653 2012-11-26 22:14:03 <TD> almost there
654 2012-11-26 22:14:13 <Luke-Jr> hmm
655 2012-11-26 22:14:14 <Luke-Jr> hmm
656 2012-11-26 22:22:18 <Luke-Jr> master reindex is good
657 2012-11-26 22:23:07 <Luke-Jr> now to make sure eligius fails on my dev system
658 2012-11-26 22:23:08 <Luke-Jr> now to make sure eligius fails on my dev system
659 2012-11-26 22:28:26 <jgarzik> Interesting description of Google Spanner database: http://www.wired.com/wiredenterprise/2012/11/google-spanner-time/all/   Bitcoin relevance...  they have systems constantly trying to reach a consensus on time
660 2012-11-26 22:29:07 <sipa> bitcoin never reaches consensus
661 2012-11-26 22:30:15 <Luke-Jr> sipa: https://gist.github.com/4151354
662 2012-11-26 22:30:22 <Luke-Jr> but again, not on master
663 2012-11-26 22:30:48 <Luke-Jr> oh duh
664 2012-11-26 22:30:51 <Luke-Jr> I'm an idiot
665 2012-11-26 22:30:55 <Luke-Jr> that's a+
666 2012-11-26 22:30:56 <Luke-Jr> that's a+
667 2012-11-26 22:31:05 <Luke-Jr> maybe
668 2012-11-26 22:33:12 <gmaxwell> jgarzik: but all systems which are in one trust domain: boring.
669 2012-11-26 22:33:13 <gmaxwell> jgarzik: but all systems which are in one trust domain: boring.
670 2012-11-26 22:33:55 <sipa> and centralized
671 2012-11-26 22:34:20 <sipa> though not on a single machine
672 2012-11-26 22:34:21 <sipa> though not on a single machine
673 2012-11-26 22:34:25 <jgarzik> not really
674 2012-11-26 22:34:51 <jgarzik> they definitely have data center-wide failure / don't-trust handling
675 2012-11-26 22:35:16 <jgarzik> one data center is a trust domain, from the failure PoV
676 2012-11-26 22:35:56 <jgarzik> I wouldn't call it centralized from a time-keeping perspective
677 2012-11-26 22:36:11 <jgarzik> a global filesystem itself can be described as centralized of course
678 2012-11-26 22:36:29 <jgarzik> GPS's and data centers can and do go offline without impact
679 2012-11-26 22:37:14 <sipa> i should read the spanner design doc again sometime :)
680 2012-11-26 22:37:15 <sipa> i should read the spanner design doc again sometime :)
681 2012-11-26 22:46:34 <jgarzik> http://research.google.com/archive/spanner.html
682 2012-11-26 22:46:35 <jgarzik> http://research.google.com/archive/spanner.html
683 2012-11-26 22:48:59 <jgarzik> still using Paxos under the hood, a la the original Google Chubby   http://research.google.com/archive/chubby.html
684 2012-11-26 22:49:14 <i18n> whatever happened to that bitcoin gambling game where you could bet that the block will end in some hex characters, and once there is a block that does, you were sent the pot?
685 2012-11-26 22:49:15 <i18n> whatever happened to that bitcoin gambling game where you could bet that the block will end in some hex characters, and once there is a block that does, you were sent the pot?
686 2012-11-26 22:49:39 <WeLoveCP> So the client right now verifies all the blocks, why don't you push a cached state of the blocks up to a certain date and have it hashed and distribute it p2p to speed that process up?
687 2012-11-26 22:49:40 <WeLoveCP> So the client right now verifies all the blocks, why don't you push a cached state of the blocks up to a certain date and have it hashed and distribute it p2p to speed that process up?
688 2012-11-26 22:50:15 <sipa> WeLoveCP: if you're going to trust us to do the database indexing for you, why are you running a full node in the first place?
689 2012-11-26 22:50:38 <sipa> WeLoveCP: the reason is because you exactly don't want to trust anyone - if you do, you're better off running a lightweight node
690 2012-11-26 22:50:39 <WeLoveCP> 'trust'?  if it is hashed and verified by many people it's more like web of trust?
691 2012-11-26 22:50:40 <WeLoveCP> 'trust'?  if it is hashed and verified by many people it's more like web of trust?
692 2012-11-26 22:50:54 <WeLoveCP> then you just calculate blocks from that date to the future?
693 2012-11-26 22:50:58 <sipa> lightweight nodes get their trust from the entire bitcoin network
694 2012-11-26 22:51:12 <sipa> that's much safer and much more automatic
695 2012-11-26 22:51:46 <WeLoveCP> ok cool
696 2012-11-26 22:51:47 <WeLoveCP> ok cool
697 2012-11-26 22:51:53 <sipa> http://bitcoin.stackexchange.com/questions/5471/would-it-be-possible-to-provide-a-downloadable-blockchain-that-is-updated-and-ve
698 2012-11-26 22:51:54 <sipa> http://bitcoin.stackexchange.com/questions/5471/would-it-be-possible-to-provide-a-downloadable-blockchain-that-is-updated-and-ve
699 2012-11-26 22:57:28 <i18n> do any of you guys remember the bitcoin game someone made a long time ago in which you could bet that a block would end in some hex value, and are sent the pot if that happens?
700 2012-11-26 22:57:29 <i18n> do any of you guys remember the bitcoin game someone made a long time ago in which you could bet that a block would end in some hex value, and are sent the pot if that happens?
701 2012-11-26 22:57:46 <i18n> i think it was one of the very first games...
702 2012-11-26 22:58:15 <i18n> and it's interesting because it's not just a random number, because it used the block chain
703 2012-11-26 23:01:13 <Luke-Jr> jgarzik: it would be nice if bitcoind had some way to resume-index from its last known block, if files were appended or it crashed during IBreindex
704 2012-11-26 23:01:14 <Luke-Jr> jgarzik: it would be nice if bitcoind had some way to resume-index from its last known block, if files were appended or it crashed during IBreindex
705 2012-11-26 23:01:42 <jgarzik> Luke-Jr: reindex is a rare one-time event, so not worth the effort
706 2012-11-26 23:01:51 <Luke-Jr> aw
707 2012-11-26 23:01:53 <jgarzik> new users >= 0.8 will never use that code at all
708 2012-11-26 23:01:53 <sipa> Luke-Jr: it *does* that
709 2012-11-26 23:01:54 <jgarzik> new users >= 0.8 will never use that code at all
710 2012-11-26 23:01:58 <Luke-Jr> sipa: it does? :o
711 2012-11-26 23:02:01 <sipa> it resumes from a reindex
712 2012-11-26 23:02:26 <sipa> it stores a flag in the blk database that it is reindexing, and removes it when finished
713 2012-11-26 23:02:27 <sipa> it stores a flag in the blk database that it is reindexing, and removes it when finished
714 2012-11-26 23:02:47 <sipa> if the flag is present at startup, it starts over again, skipping over parts of the files that are already indexes
715 2012-11-26 23:05:09 <Luke-Jr> nice
716 2012-11-26 23:06:22 <i18n> nobody remembers it? wow. that makes me wish i didn't get bored of bitcoin two years ago.
717 2012-11-26 23:06:23 <i18n> nobody remembers it? wow. that makes me wish i didn't get bored of bitcoin two years ago.
718 2012-11-26 23:07:40 <Luke-Jr> i18n: there's still at least one ongoing
719 2012-11-26 23:07:41 <Luke-Jr> i18n: there's still at least one ongoing
720 2012-11-26 23:07:42 <i18n> by get bored of i mean lose interest
721 2012-11-26 23:07:43 <i18n> by get bored of i mean lose interest
722 2012-11-26 23:07:47 <Luke-Jr> except the "pot" is ASIC hardware
723 2012-11-26 23:07:50 <i18n> Luke-Jr: really? do you know what it's called?
724 2012-11-26 23:07:53 <Luke-Jr> no
725 2012-11-26 23:07:54 <i18n> oh
726 2012-11-26 23:07:57 <Luke-Jr> luceo (otc) runs it
727 2012-11-26 23:08:11 <i18n> 
ACTION considers creating one

728 2012-11-26 23:08:12 <i18n> 
ACTION considers creating one

729 2012-11-26 23:11:17 <Luke-Jr> bitcoind: main.cpp:1543: bool CBlock::DisconnectBlock(CBlockIndex*, CCoinsViewCache&): Assertion `blockUndo.vtxundo.size() + 1 == vtx.size()' failed.
730 2012-11-26 23:11:18 <Luke-Jr> sipa: O.o???
731 2012-11-26 23:11:19 <Luke-Jr> sipa: O.o???
732 2012-11-26 23:12:34 <sipa> Luke-Jr: that shouldn't be an assert, but it does mean the undo data and the block data are out of sync
733 2012-11-26 23:13:09 <sipa> what did you do to get that?
734 2012-11-26 23:19:05 <Luke-Jr> sipa: reindexed from a read-only blk file
735 2012-11-26 23:19:22 <Luke-Jr> tired of fighting it, so giving up and copying the block files now
736 2012-11-26 23:19:23 <Luke-Jr> tired of fighting it, so giving up and copying the block files now
737 2012-11-26 23:19:26 <Luke-Jr> will let you know if it still happens
738 2012-11-26 23:19:27 <Luke-Jr> will let you know if it still happens
739 2012-11-26 23:19:38 <Luke-Jr> (giving up = after I finish this pullreq)
740 2012-11-26 23:20:15 <sipa> right, having read-only linked files is certainly that sounds nice to support, but i don't expect it to work out of the box now
741 2012-11-26 23:26:47 <Luke-Jr> there, pullreq fixed (though I'm ignoring the truncate race for now)
742 2012-11-26 23:28:29 <sipa> i wonder how the old code prevented writing the genesis block to the block files?
743 2012-11-26 23:28:34 <Luke-Jr> sipa: with copied files, it assert fails still. trying again with -reindex
744 2012-11-26 23:28:57 <sipa> in test i mean
745 2012-11-26 23:28:58 <sipa> in test i mean
746 2012-11-26 23:44:56 <Luke-Jr> sipa: assert still fails with -reindex, trying on master now :/