1 2012-03-15 00:10:57 <doublec> is there a bitcoin released version with bip 30 support?
  2 2012-03-15 00:11:04 <BlueMatt> no, only 0.6
  3 2012-03-15 00:11:11 <BlueMatt> dont even know if its in rc1
  4 2012-03-15 00:11:14 <BlueMatt> dont think so
  5 2012-03-15 00:11:15 <doublec> so people who only use releases are stuck?
  6 2012-03-15 00:11:16 <BlueMatt> probably in rc2
  7 2012-03-15 00:11:25 <BlueMatt> only miners really need to upgrade
  8 2012-03-15 00:11:31 <BlueMatt> but if you mine and you only use releases, yea
  9 2012-03-15 00:11:45 <doublec> is there an announcement about needing to upgrade somewhere?
 10 2012-03-15 00:14:49 <BlueMatt> sipa: oh, thanks for actually looking at cblockstore, btw
 11 2012-03-15 00:15:10 <BlueMatt> or reviewing some of it, at least
 12 2012-03-15 00:15:35 <Graet> https://bitcointalk.org/index.php?topic=67738.0  and https://en.bitcoin.it/wiki/BIP_0030
 13 2012-03-15 00:15:56 <Graet> sipa also emailed most pools afaik
 14 2012-03-15 00:16:21 <Graet> i joined dev mailing list so i'd have more notice next time doublec
 15 2012-03-15 00:18:35 <doublec> no mention on bitcoin.org sadly
 16 2012-03-15 00:22:00 <BlueMatt> when your paste of bitcoin running in dr. memory (valgrind for windows) is so big you exceed pastebin's max file size, you know bitcoin has too many errors...
 17 2012-03-15 00:22:27 <Diablo-D3> or dr memory sucks
 18 2012-03-15 00:22:33 <denisx> or both
 19 2012-03-15 00:22:40 <BlueMatt> Diablo-D3: got a better alternative?
 20 2012-03-15 00:23:03 <Diablo-D3> cash.
 21 2012-03-15 00:23:16 <BlueMatt> link?
 22 2012-03-15 00:23:19 <BlueMatt> thats kinda hard to google...
 23 2012-03-15 00:29:16 <sipa> doublec: 0.5.3 is more or less released, and has bip30
 24 2012-03-15 00:31:10 <luke-jr> doublec: http://luke.dashjr.org/programs/bitcoin/files/bitcoin-0.5.3/
 25 2012-03-15 00:31:15 <luke-jr> sipa: did you push the tag btw?
 26 2012-03-15 00:31:31 <BlueMatt> is 0.5.3 == 0.5.3rc4?
 27 2012-03-15 00:31:34 <luke-jr> yes
 28 2012-03-15 00:31:45 <luke-jr> I used a hard link on my webserver <.<
 29 2012-03-15 00:31:52 <Diablo-D3> fail
 30 2012-03-15 00:54:24 <FaktioNN> If the value for bnProofOfWorkLimit was set to a very low number, it should create a very high target/low difficulty when compared to mainnet. Do such changes on a test chain require you to create a new genesis block, or is it possible to set the new worklimit with the same genesis block (would certainly make life easier)?
 31 2012-03-15 00:56:03 <theymos> You need a new genesis block unless you want to modify the difficulty adjustment code to make an exception for the genesis block.
 32 2012-03-15 00:56:29 <theymos> Actually, maybe not. I'm not sure whether Bitcoin actually checks the validity of the genesis block.
 33 2012-03-15 00:56:48 <BlueMatt> I think it may just compare to known hash and accept it
 34 2012-03-15 00:56:53 <BlueMatt> (its probably a checkpoint)
 35 2012-03-15 00:58:42 <FaktioNN> Yeah, it certainly seems that way. I guess the flipside for it though is generating the new genesis block will be very quick with an extremely high starting target value.
 36 2012-03-15 01:00:20 <forrestv> BlueMatt, if the majority of p2poolers are upgraded, all of the the unupgraded ones' shares will be orphaned. if a minority are upgraded, all of their shares will be orphaned
 37 2012-03-15 01:00:38 <BlueMatt> mmm, I was afraid of that
 38 2012-03-15 01:00:47 <forrestv> p2pool has a prev_block check that is similar to bitcoin's "discourage blocks" feature
 39 2012-03-15 01:01:02 <BlueMatt> bitcoin doesnt have a discourage blocks feature...
 40 2012-03-15 01:01:20 <shaptiul> gavin has proposed block shunning
 41 2012-03-15 01:01:32 <BlueMatt> proposed, not added
 42 2012-03-15 01:01:41 <luke-jr> forrestv: how does p2pool implement BIP 30? O.o
 43 2012-03-15 01:01:42 <forrestv> ah, okay, the proposed discourage blocks feature :P
 44 2012-03-15 01:02:13 <luke-jr> forrestv: it requires a full node implementation&
 45 2012-03-15 01:02:51 <forrestv> luke-jr, ? p2pool will never generate duplicate transactions, and as long as the underlying bitcoind doesn't accept blocks with dups, there is no problem
 46 2012-03-15 01:03:10 <luke-jr> forrestv: and what if a p2pool user *does*?
 47 2012-03-15 01:03:19 <forrestv> does what?
 48 2012-03-15 01:03:23 <luke-jr> generate dupe txns
 49 2012-03-15 01:03:25 <[Tycho]> Nice TX http://bitcoincharts.com/bitcoin/txlist/#184e372dc413ca4df478a9bc176723f310a9b90aa4b8042db32bb4471b505ee8
 50 2012-03-15 01:03:40 <forrestv> purposefully and maliciously, you mean?
 51 2012-03-15 01:03:43 <luke-jr> yes
 52 2012-03-15 01:03:45 <luke-jr> I suppose it's hard now, though
 53 2012-03-15 01:04:06 <forrestv> there are other attacks on p2pool that would have the same effect and would be much easier, such as just withholding block solutions
 54 2012-03-15 01:04:24 <luke-jr> true
 55 2012-03-15 01:05:29 <forrestv> as long as a supermajority of bitcoin's hashing power implements bip 30, bip 30 shouldn't affect p2pool
 56 2012-03-15 01:06:15 <BlueMatt> well, shit...the rpc bug in bitcoin-qt.exe...well it looks like its been there since the start.  Its very easily reproduceable on anything >=0.5.0, but I cant get it to die in 0.4.1
 57 2012-03-15 01:06:51 <theymos> What RPC bug?
 58 2012-03-15 01:07:05 <BlueMatt> the random rpc crashing bug where calling rpc functions kills bitcoin-qt.exe
 59 2012-03-15 01:07:11 <BlueMatt> only on win32 and only bitcoin-qt, not bitcoind
 60 2012-03-15 01:07:51 <luke-jr> O.o
 61 2012-03-15 01:07:51 <theymos> Oh, OK. Was wondering if I was possibly affected.
 62 2012-03-15 01:08:03 <forrestv> thanks for working on that, BlueMatt :) any idea what's going on there?
 63 2012-03-15 01:08:27 <BlueMatt> no, Ive been running it through valgrind/dr. memory for a while now and I really cat find anything...
 64 2012-03-15 01:09:24 <luke-jr> BlueMatt: stack trace?
 65 2012-03-15 01:09:31 <BlueMatt> nothing remotely useful
 66 2012-03-15 01:09:37 <luke-jr> :/
 67 2012-03-15 01:09:47 <BlueMatt> looks like it goes haywire before dying, overwriting stacks like mad
 68 2012-03-15 01:10:32 <BlueMatt> I mean you can see from some of the errors, iirc someone saw an error in like an rpc function from the net thread or something really random like that
 69 2012-03-15 01:10:32 <luke-jr> does WinGDB support stepping backward?
 70 2012-03-15 01:10:43 <BlueMatt> just stuff that should never, ever have appeared
 71 2012-03-15 01:11:08 <luke-jr> did you build with -O0 ?
 72 2012-03-15 01:11:13 <BlueMatt> yea
 73 2012-03-15 01:11:36 <forrestv> BlueMatt, yeah, that was mine
 74 2012-03-15 01:13:03 <etotheipi_> what do you think is the best to identify whether a wallet.dat file is on testnet, mainnet, namecoin, etc?
 75 2012-03-15 01:13:21 <BlueMatt> oh, I did find the issue with qmake stripping when i dont want it to, there was a -Wl,s hiding in the link (ld -s == --strip-all)
 76 2012-03-15 01:13:35 <etotheipi_> the only thing I can find is that account names store the actual Base58 addr, and I can get it from that
 77 2012-03-15 01:13:36 <BlueMatt> so if I find the time, releasing debug symbols with the next release should be easily possible
 78 2012-03-15 01:14:00 <etotheipi_> (assuming at least one address has a name/label)
 79 2012-03-15 01:14:20 <luke-jr> http://undo-software.com/undodb_requestnoncommercial.html
 80 2012-03-15 01:14:26 <luke-jr> in case you want to use non-free sw
 81 2012-03-15 01:14:41 <luke-jr> err
 82 2012-03-15 01:14:44 <luke-jr> Linux only? O.o
 83 2012-03-15 01:14:59 <BlueMatt> thats no good :(
 84 2012-03-15 01:15:08 <BlueMatt> well, actually maybe could run it through wine, but...
 85 2012-03-15 01:15:19 <luke-jr> BlueMatt: how easy is it to reproduce?
 86 2012-03-15 01:15:26 <BlueMatt> very for me
 87 2012-03-15 01:15:30 <luke-jr> how?
 88 2012-03-15 01:16:05 <BlueMatt> I run a while [ true ]; do bitcoind -rpcuser... walletlock; done and a getwork and a getmemory pool in three terminals and it dies in seconds
 89 2012-03-15 01:16:22 <BlueMatt> (on most builds)
 90 2012-03-15 01:16:32 <luke-jr> http://wiki.jswindle.com/index.php/Winedbg
 91 2012-03-15 01:16:37 <BlueMatt> sometimes debug builds act differently, etc...but they all eventually die
 92 2012-03-15 01:16:37 <luke-jr> maybe see if you can do it in that
 93 2012-03-15 01:16:52 <BlueMatt> yea, Ive messed a bit with that, nothing particularly useful afaict...
 94 2012-03-15 01:16:53 <BlueMatt> :(
 95 2012-03-15 01:17:07 <BlueMatt> the whole killing stacks before dying just kills debuggers...
 96 2012-03-15 01:17:41 <BlueMatt> and I dont have the patience to step through one instruction at a time with winedbg (its default)
 97 2012-03-15 01:17:58 <BlueMatt> Im considering whipping out ida in the next few days and really going at it, but only if I find the time
 98 2012-03-15 01:18:09 <BlueMatt> also, I have to go find a copy of ida again...
 99 2012-03-15 01:19:02 <doublec> BlueMatt: so bitcoind crashes as well as bitcoin-qt?
100 2012-03-15 01:19:05 <BlueMatt> no
101 2012-03-15 01:19:09 <BlueMatt> only bitcoin-qt.exe
102 2012-03-15 01:19:43 <BlueMatt> its bug #820 and #640 (99% sure they are dups)
103 2012-03-15 01:29:48 <gribble> New news from bitcoinrss: TheBlueMatt opened pull request 938 on bitcoin/bitcoin <https://github.com/bitcoin/bitcoin/pull/938>
104 2012-03-15 01:32:46 <BlueMatt> forrestv: how do you usually reproduce the error?
105 2012-03-15 01:32:53 <BlueMatt> am I correct that p2pool will kill it quickly?
106 2012-03-15 01:35:29 <nanotube> what did i miss, you guys pinged something about 0.5.3?
107 2012-03-15 01:35:39 <forrestv> BlueMatt, i don't have much experience with it.. but it happened pretty quickly with p2pool in a VM when i last tried it
108 2012-03-15 01:35:59 <BlueMatt> you dont happen to have a list of all the rpc commands p2pool calls?
109 2012-03-15 01:36:14 <BlueMatt> or even better a list of calls it made on a particular run?
110 2012-03-15 01:37:31 <forrestv> during init: help, getinfo, validateaddress. once running: getwork, getblock, getmemorypool
111 2012-03-15 01:37:44 <BlueMatt> mmm, thanks
112 2012-03-15 01:38:06 <forrestv> +getaccountaddress during init
113 2012-03-15 01:38:12 <BlueMatt> ok
114 2012-03-15 01:38:41 <forrestv> i think it's safe to assume that it's not the ones during init, though, and it happens on builds without getblock
115 2012-03-15 01:38:55 <forrestv> so getwork/getmemorypool
116 2012-03-15 01:39:01 <BlueMatt> alright, thanks, Ill do some more reproduceing
117 2012-03-15 01:39:35 <forrestv> BlueMatt, a good way to reproduce it is to call them while bitcoin is downloading blocks
118 2012-03-15 01:40:09 <sipa> nanotube: yes, can you upload 0.5.3?
119 2012-03-15 01:42:15 <nanotube> sipa: sure, got me a link to the latest distribution files?
120 2012-03-15 01:43:23 <luke-jr> nanotube: http://luke.dashjr.org/programs/bitcoin/files/bitcoin-0.5.3/
121 2012-03-15 01:43:26 <luke-jr> nanotube: renames from rc4
122 2012-03-15 01:43:49 <nanotube> so no changes from the rc4 at all, other than a rename of the archives?
123 2012-03-15 01:44:23 <BlueMatt> hopefully the folder inside the archives got renamed too?
124 2012-03-15 01:44:24 <nanotube> (iow, if i just rename the files over on the sf FRS rather than reuploading, (and make a new shasums to match), everything will be peachy?
125 2012-03-15 01:44:34 <luke-jr> nanotube: right
126 2012-03-15 01:44:34 <nanotube> ah i guess that's a good point ;)
127 2012-03-15 01:44:43 <luke-jr> BlueMatt: oops, nope
128 2012-03-15 01:44:48 <luke-jr> I can do that I guess
129 2012-03-15 01:46:09 <nanotube> and i see we got the osx build too eh.
130 2012-03-15 01:47:18 <BlueMatt> before I forget ^
131 2012-03-15 01:47:30 <nanotube> heh
132 2012-03-15 01:47:44 <BlueMatt> usually takes a day or two before that happens...
133 2012-03-15 01:49:22 <nanotube> BlueMatt: hey btw, so i know how to verify gitian sigs for the win .zip, and the linux tgz. how would i do it for the .exe and the .dmg?
134 2012-03-15 01:49:43 <luke-jr> nanotube: afaik the dmg isn't gitian
135 2012-03-15 01:49:47 <BlueMatt> there is no such thing as gitian zips, only the contents...
136 2012-03-15 01:49:54 <BlueMatt> also, yea the mac one isnt done using gitian
137 2012-03-15 01:50:10 <BlueMatt> you have to extract the contents in the way that gitian produces them to build/out in gitian-builder
138 2012-03-15 01:50:20 <nanotube> BlueMatt: right, i mean, the zip and the tgz i can extract
139 2012-03-15 01:50:28 <BlueMatt> the win32-setup.exe is one file in that folder
140 2012-03-15 01:50:32 <BlueMatt> and is removed before making the zip
141 2012-03-15 01:50:32 <nanotube> but how do i verify the exe and dmg?
142 2012-03-15 01:50:38 <BlueMatt> you cant verify the dmg...
143 2012-03-15 01:50:47 <nanotube> ok, and exe? just shove it into build/out?
144 2012-03-15 01:50:50 <nanotube> and it'll work?
145 2012-03-15 01:50:58 <luke-jr> right
146 2012-03-15 01:51:06 <luke-jr> nanotube: exe and dmg are identical byte for byte, so you can upload them first :p
147 2012-03-15 01:51:23 <luke-jr> tar.gz upload done&
148 2012-03-15 01:51:25 <sipa> heh?
149 2012-03-15 01:51:33 <BlueMatt> exe shouldnt be identical though
150 2012-03-15 01:51:39 <BlueMatt> doesnt exe say version number in it?
151 2012-03-15 01:51:52 <luke-jr> BlueMatt: it doesn't say rc
152 2012-03-15 01:51:58 <BlueMatt> oh, ok
153 2012-03-15 01:51:59 <sipa> yes, but rcs do not have version number
154 2012-03-15 01:52:27 <BlueMatt> nanotube: file hierarchy (and hashes, but gitian checks that) should match that which is found in the bitcoin-build.assert (eg https://github.com/bitcoin/gitian.sigs/blob/master/0.5.3rc4-win32/bluematt/bitcoin-build.assert )
155 2012-03-15 01:52:34 <sipa> Starting from 0.6.0rc3, they do, by the way (somewhat)
156 2012-03-15 01:52:37 <BlueMatt> so win32-setup.exe is in top level
157 2012-03-15 01:52:55 <luke-jr> BlueMatt: it'd be nice if the gitian-win32 did the final zip btw :p
158 2012-03-15 01:53:05 <luke-jr> but then I guess it'd impede repacking like this
159 2012-03-15 01:53:08 <BlueMatt> luke-jr: there are reasons we dont, but its a longer story
160 2012-03-15 01:53:21 <BlueMatt> (gitian-downloader reasons)
161 2012-03-15 01:53:24 <sipa> meh, my build script does everything
162 2012-03-15 01:53:34 <BlueMatt> oh, shit, I need to make the setup an optional...
163 2012-03-15 01:54:07 <luke-jr> nanotube: http://luke.dashjr.org/programs/bitcoin/files/bitcoin-0.5.3/ updated correctly
164 2012-03-15 01:54:25 <BlueMatt> hah, Ive found the problem...now I just have to find the problem...
165 2012-03-15 01:55:26 <nanotube> hm, there appears to be no 0.5.3 in the gitian.sigs repo... ?
166 2012-03-15 01:55:33 <BlueMatt> its 0.5.3rc4
167 2012-03-15 01:55:50 <BlueMatt> since the files are the same, no on bothered rebuilding
168 2012-03-15 01:55:54 <nanotube> ah ic
169 2012-03-15 01:57:18 <BlueMatt> who knows a lot about C++ exceptions?
170 2012-03-15 01:57:26 <BlueMatt> (thats the problem with the win32 rpc)
171 2012-03-15 01:57:30 <nanotube> so the dmg, anyone care to confirm the checksum of of the dmg? 90690d3838fc232124dcbb53c915ebbb613b1bba0a8a3e1d30a888af8787dc49 ? (since it appears that that's the only way to check it?
172 2012-03-15 01:57:39 <BlueMatt> throw JSONRPCError is doing the funkyness
173 2012-03-15 01:57:44 <nanotube> luke-jr: thanks redownloading
174 2012-03-15 01:57:49 <BlueMatt> nanotube: have to ask gavin for that...
175 2012-03-15 01:59:04 <luke-jr> nanotube: Gavin personally uploaded the DMG to SourceForge as rc4; I just downloaded it from there
176 2012-03-15 01:59:36 <nanotube> ah ok.
177 2012-03-15 02:00:10 <BlueMatt> sipa: have you pushed the tag yet?
178 2012-03-15 02:00:37 <luke-jr> BlueMatt: you should make the debs gitian-built too
179 2012-03-15 02:00:40 <BlueMatt> luke-jr: why did you update the debian change log again???
180 2012-03-15 02:00:53 <BlueMatt> luke-jr: yea...thats a HUGE pita...
181 2012-03-15 02:01:11 <BlueMatt> because I have no control of the build process
182 2012-03-15 02:01:18 <BlueMatt> it has to be in a script that launchpad runs in a vm
183 2012-03-15 02:01:20 <luke-jr> o
184 2012-03-15 02:01:21 <BlueMatt> (I think(
185 2012-03-15 02:01:22 <BlueMatt> )
186 2012-03-15 02:01:36 <BlueMatt> I know it has to be run on their servers, Im pretty sure they are vms
187 2012-03-15 02:02:29 <luke-jr> lame
188 2012-03-15 02:02:40 <BlueMatt> kinda, but it forces sane build procedures
189 2012-03-15 02:02:40 <Someguy123> hey where's gavin? the faucet is missing its donation address
190 2012-03-15 02:02:44 <luke-jr> tcatm: ping
191 2012-03-15 02:02:46 <Someguy123> https://freebitcoins.appspot.com/
192 2012-03-15 02:02:59 <BlueMatt> arg, wtf does google think my ip is french (again)
193 2012-03-15 02:03:22 <doublec> BlueMatt: is the exception handled anywhere or does it not reach that point?
194 2012-03-15 02:03:40 <BlueMatt> doublec: havent done any research yet, I just finally narrowed it down to exceptions
195 2012-03-15 02:04:00 <BlueMatt> and now Im distracted doing ppa uploads...
196 2012-03-15 02:04:23 <BlueMatt> luke-jr: seriously, stop changing the debian changelog on releases...how many times have I said that?
197 2012-03-15 02:05:21 <BlueMatt> also, why is the 0.5.3 entry marked jan 10?
198 2012-03-15 02:10:53 <nanotube> BlueMatt: ok.... so here's the thing. i extracted the .zip into build/out/ (which got placed into 'bitcoin-0.5.3-win32' subdirectory.
199 2012-03-15 02:11:12 <nanotube> then i copied in the setup.exe, and ran ./bin/gverify ....
200 2012-03-15 02:11:17 <BlueMatt> gotta move everything up a folder
201 2012-03-15 02:11:20 <sipa> BlueMatt, luke-jr: pusahed
202 2012-03-15 02:11:22 <BlueMatt> (it shouldnt be in any subdir)
203 2012-03-15 02:11:24 <nanotube> ah
204 2012-03-15 02:11:29 <nanotube> let's try
205 2012-03-15 02:13:15 <sipa> BlueMatt: is RPC broken on bitcoin-qt in general, or just on win32?
206 2012-03-15 02:13:26 <BlueMatt> win32
207 2012-03-15 02:14:01 <nanotube> BlueMatt: ok, so i moved everything up a dir, so that the -qt.exe and everything is in build/out
208 2012-03-15 02:14:11 <nanotube> i ran bin/gverify, and it said all a-ok
209 2012-03-15 02:14:35 <nanotube> then... i went and modified the files (echo "bla" >> bitcoin-0.5.3-win32-setup.exe)
210 2012-03-15 02:14:38 <nanotube> ran gverify again...
211 2012-03-15 02:14:42 <nanotube> and it said everything a-ok
212 2012-03-15 02:14:56 <nanotube> which leads me to suspect that something isn't right :)
213 2012-03-15 02:15:47 <nanotube> so i guess the question is, what exactly is gverify actually checking?
214 2012-03-15 02:17:11 <BlueMatt> hmm, it should be checking the files
215 2012-03-15 02:17:21 <BlueMatt> maybe its only checking the sigs for you...
216 2012-03-15 02:17:27 <BlueMatt> change one of the sigs and see what happens?
217 2012-03-15 02:18:05 <nanotube> well, to 'check the sig' one must look at both the signature and the file... you can't 'check the sig' without referring to the file that was signed...
218 2012-03-15 02:18:19 <BlueMatt> checks the sigs against each other I mean
219 2012-03-15 02:18:42 <BlueMatt> change one of the hashes in a sig and see
220 2012-03-15 02:18:45 <BlueMatt> and change a sig and see
221 2012-03-15 02:19:10 <nanotube> ok, you mean the bitcoin-build.assert file?
222 2012-03-15 02:19:14 <BlueMatt> yea
223 2012-03-15 02:20:03 <nanotube> ok, changed the .assert, it said 'bad signature'. good
224 2012-03-15 02:20:20 <nanotube> so it checks that your .assert.sig matches the .assert... but doesn't check the actual file hashes?
225 2012-03-15 02:20:25 <nanotube> that seems rather ... useless.
226 2012-03-15 02:20:44 <BlueMatt> hmm, thats odd I know I checked that it did a few days ago...
227 2012-03-15 02:21:15 <BlueMatt> one sec
228 2012-03-15 02:21:18 <nanotube> ok
229 2012-03-15 02:21:36 <sipa> devrandom: what gverify supposed to check exactly?
230 2012-03-15 02:22:14 <nanotube> well, at worst, i can just run sha256sum -c bitcoin-build.assert :)
231 2012-03-15 02:22:30 <nanotube> after i know that the sigs match...
232 2012-03-15 02:24:01 <nanotube> yea that works as a poor-man's solution :) until gverify starts verifying.
233 2012-03-15 02:24:02 <BlueMatt> well sort of, you have to modify the files abit, but yea
234 2012-03-15 02:24:09 <nanotube> nah, no need to modify
235 2012-03-15 02:24:12 <nanotube> just cd to build/out
236 2012-03-15 02:24:25 <BlueMatt> that file has like deb shasums in it too...
237 2012-03-15 02:24:26 <nanotube> then sha256sum -c ../path/to/the/assertfile
238 2012-03-15 02:24:31 <nanotube> yea, but i can ignore missing files
239 2012-03-15 02:24:42 <BlueMatt> and I thought it had headers...
240 2012-03-15 02:24:45 <BlueMatt> ok, well whatever
241 2012-03-15 02:24:48 <nanotube> at the end it tells me a summary, how many files matched, how many missing.
242 2012-03-15 02:24:50 <nanotube> yes it has headers
243 2012-03-15 02:25:06 <BlueMatt> oh, I guess shasum ignores crap it doesnt get
244 2012-03-15 02:25:06 <nanotube> but sha256sum is smart enough to ignore lines that don't look like <shasum> <filename>
245 2012-03-15 02:25:08 <BlueMatt> well then ok
246 2012-03-15 02:25:10 <nanotube> yep
247 2012-03-15 02:25:22 <nanotube> ok, i'll do that for now, until devrandom comes back and tells us about gverify :)
248 2012-03-15 02:25:29 <BlueMatt> ppa building, well uploaded...will build soon
249 2012-03-15 02:25:31 <nanotube> or modifies it to do it :)
250 2012-03-15 02:25:54 <BlueMatt> gitian-downloader will do the checks...but it may not be so friendly to hack the zips to downloader-friendly form
251 2012-03-15 02:26:54 <BlueMatt> arg, sometimes I really hate launchpad..."Will build in 2 hours"
252 2012-03-15 02:26:57 <BlueMatt> wtf? seriously?
253 2012-03-15 02:27:01 <BlueMatt> its not like you are that overloaded
254 2012-03-15 02:27:16 <BlueMatt> and the fact that its not a queue, its just a 2 hour delay tells me you are just being a dick about it...
255 2012-03-15 02:30:07 <nanotube> BlueMatt: so... just build your own debs and shove them up in an apt repo.
256 2012-03-15 02:30:18 <nanotube> (as a bonus, you control the build process, so you can make it deterministic :) )
257 2012-03-15 02:30:26 <BlueMatt> meh, but then I have to host it
258 2012-03-15 02:30:46 <BlueMatt> and you lose the simplicity of telling people to "sudo apt-add-repository ppa:bitcoin/bitcoin" and have to tell them a url and pgp key...
259 2012-03-15 02:33:13 <luke-jr> BlueMatt: I've changed it exactly 0 times since you said it first
260 2012-03-15 02:33:27 <BlueMatt> mmm, I didnt see the timestamp as jan6...
261 2012-03-15 02:33:31 <BlueMatt> but whatever
262 2012-03-15 02:34:08 <luke-jr> [23:05:21] <BlueMatt> also, why is the 0.5.3 entry marked jan 10? <-- where?
263 2012-03-15 02:34:17 <BlueMatt> in contrib/debian/changelog
264 2012-03-15 02:34:31 <nanotube> BlueMatt: you don't have to host it, just use sf.net frs. set up a project "bitcoin-ppa", and rsync your repo over to a dir on the frs. (i do that for another project i run)
265 2012-03-15 02:34:48 <luke-jr> BlueMatt: probably because that's when I added it
266 2012-03-15 02:35:02 <nanotube> yes, the instructions to add repository expand to two commands. "apt-add-repository deb repo line" and "apt-key add line"
267 2012-03-15 02:35:06 <BlueMatt> luke-jr: mmm, whatever
268 2012-03-15 02:35:08 <nanotube> but that's pretty minor.
269 2012-03-15 02:35:31 <BlueMatt> nanotube: ah, well whatever, its simple enough to do it on launchpad and as a bonus its already set up :)
270 2012-03-15 02:35:37 <nanotube> just fyi - if you do decide to go that way :)
271 2012-03-15 02:36:01 <nanotube> yea if it works currently, that's fine.
272 2012-03-15 02:36:14 <nanotube> just introduces an extra party to trust (namely, launchpad)
273 2012-03-15 02:36:28 <sipa> people need to trust their OS anyway
274 2012-03-15 02:36:41 <BlueMatt> launchpad == canonical == ubuntu...
275 2012-03-15 02:36:44 <luke-jr> nanotube: how goes upload?
276 2012-03-15 02:36:46 <sipa> and we do too... gitian builds are ubuntu
277 2012-03-15 02:37:32 <nanotube> sipa: BlueMatt: compromising a single repo on launchpad to build bogus bitcoin binaries is a lot less discoverable than compromising the whole ubuntu OS so that it builds bogus bitcoin binaries. :)
278 2012-03-15 02:37:44 <BlueMatt> true...
279 2012-03-15 02:38:14 <nanotube> nobody sees what launchpad does behind the scenes. everyone sees what his os does (in aggregate :) )
280 2012-03-15 02:38:16 <sipa> nanotube: it would be a very interesting and potentially profitable attack, still!
281 2012-03-15 02:38:50 <nanotube> haha true, but to do that you'd have to compromise GCC or some such, and source code to all the standard packages is available. compromising launchpad can be completely secret.
282 2012-03-15 02:39:04 <nanotube> luke-jr: yea working on it :) just got caught up with some conversations. hehe
283 2012-03-15 02:39:12 <nanotube> anyway, just for your consideration, BlueMatt :D
284 2012-03-15 02:39:19 <BlueMatt> compromising launchpad == compromising the build machines that actually build ubuntu packages...
285 2012-03-15 02:39:34 <BlueMatt> well ~=
286 2012-03-15 02:39:53 <BlueMatt> (AFAICT)
287 2012-03-15 02:39:53 <sipa> BlueMatt: what if we'd just agree with all devs to introduce unsuspicious code in complex commits that steal people's coins disguised as "corrupted wallets" ?
288 2012-03-15 02:40:10 <BlueMatt> sipa: lets do it
289 2012-03-15 02:40:10 <nanotube> BlueMatt: 'compromising' can also include "govt agency leaning on canonical". not so easy to do for the actual OS which includes source code.
290 2012-03-15 02:40:21 <nanotube> haha
291 2012-03-15 02:40:29 <sipa> BlueMatt: we'll split based on number of contributed lines of code
292 2012-03-15 02:40:51 <BlueMatt> nanotube: the launchpad bitcoin repo contains source code in the exact same way that ubuntu does for any other package, btw
293 2012-03-15 02:41:02 <BlueMatt> sipa: nice
294 2012-03-15 02:42:26 <nanotube> BlueMatt: even code to the launchpad website that does the deb building?
295 2012-03-15 02:43:23 <luke-jr> sipa: not a fair measurement. your commits are smaller, but more valuable per line IMO
296 2012-03-15 02:43:33 <BlueMatt> no, but that is open elsewhere, note that the code that launchpad uses to build is the same code (and same servers, or mirror servers) that does regular deb builds of eg gcc for ubuntu
297 2012-03-15 02:43:40 <sipa> luke-jr: don't worry, i got plenty ;)
298 2012-03-15 02:44:10 <BlueMatt> sipa: can we do the counting after cblockstore?
299 2012-03-15 02:44:14 <luke-jr> lol
300 2012-03-15 02:44:17 <luke-jr> I was about to mention that
301 2012-03-15 02:44:41 <sipa> let's not encourage you to make it even more breaking :)
302 2012-03-15 02:44:50 <BlueMatt> arg winedbg sucks...
303 2012-03-15 02:44:53 <BlueMatt> heh
304 2012-03-15 02:45:00 <sipa> (breaking in the sense of breaking other patches)
305 2012-03-15 02:50:08 <BlueMatt> I take that back winedbg is GREAT, it doesnt debug the process you are running, it debugs something else, what I have nfc???
306 2012-03-15 02:50:16 <BlueMatt> I think it may be debugging itself...or something?
307 2012-03-15 02:50:37 <sipa> lol
308 2012-03-15 02:51:13 <BlueMatt> if you call the ni "execute next assembly instruction"  command like 4 times, it continues to run the entire program without stopping, wtf???
309 2012-03-15 02:51:55 <sipa> it seems to try to mimic windows?
310 2012-03-15 02:52:09 <BlueMatt> I have nfc what winedbg is doing
311 2012-03-15 02:52:28 <BlueMatt> mimicing the average code quality of windows apps maybe
312 2012-03-15 02:52:30 <sipa> cool i didn't know near-field communication could do that
313 2012-03-15 02:52:37 <forsetifox> Heh.
314 2012-03-15 02:52:47 <BlueMatt> gnight
315 2012-03-15 02:55:21 <BlueMatt> man winedbg, section Bugs-> "A lot."
316 2012-03-15 02:55:43 <BlueMatt> fuck you
317 2012-03-15 02:56:22 <luke-jr> lol
318 2012-03-15 02:56:53 <luke-jr> try with --gdb?
319 2012-03-15 02:57:14 <BlueMatt> tried that, insta-crash
320 2012-03-15 02:57:33 <BlueMatt> "Internal crash at..."
321 2012-03-15 03:01:05 <sipa> +
322 2012-03-15 03:03:40 <nanotube> ok, 0.5.3 is up.
323 2012-03-15 03:03:57 <BlueMatt> nice
324 2012-03-15 03:04:59 <luke-jr> tcatm: ping
325 2012-03-15 03:06:20 <forsetifox> luke-jr: Should I just keep using the version you gave me to test since I can't go back to 0.5.3 without making another wallet?
326 2012-03-15 03:06:46 <luke-jr> forsetifox: should be fine& I'd keep updating until 0.6.0 comes out tho
327 2012-03-15 03:06:48 <sipa> forsetifox: are you running 0.6.0rc3?
328 2012-03-15 03:06:58 <forsetifox> Let me check.
329 2012-03-15 03:07:04 <BlueMatt> forsetifox: meh, just run 0.6.0rc3
330 2012-03-15 03:07:24 <sipa> It's listed as 0.6.0.3 in the program
331 2012-03-15 03:07:40 <forsetifox> It says "0.6.0-beta".
332 2012-03-15 03:07:48 <sipa> In that case, no.
333 2012-03-15 03:07:52 <jjjrmy-m> Does the guy who made instawallet one here?
334 2012-03-15 03:08:01 <jjjrmy-m> *on
335 2012-03-15 03:08:08 <sipa> jjjrmy-m: He's been here occasionally.
336 2012-03-15 03:08:10 <sipa> jav
337 2012-03-15 03:52:04 <gribble> New news from bitcoinrss: TheBlueMatt reopened pull request 938 on bitcoin/bitcoin <https://github.com/bitcoin/bitcoin/pull/938>
338 2012-03-15 03:54:42 <devrandom> sipa: gverify checks that all sigs are valid and that they have compatible assertions - i.e same output files
339 2012-03-15 04:01:17 <devrandom> sipa: it doesn't check the actual files.  It's only meant for builders, which supposedly generated their own assertion and their own files.  It's not a facility for checking others' builds.
340 2012-03-15 04:01:23 <devrandom> nanotube: ^^
341 2012-03-15 04:02:21 <nanotube> devrandom: heh i see. well... maybe a feature request? when we run with some argument like '-c', make it check the shasums on all files in build/out?
342 2012-03-15 04:02:34 <devrandom> the downloader checks actual files.  If you need a builder-friendly utility, let me know and I'll look into it.
343 2012-03-15 04:02:53 <nanotube> and throw out big scary warnings if (a) any of the sums don't match, or (b) any of the files in build/out are not present in assertion list.
344 2012-03-15 04:03:01 <devrandom> nanotube: ok, that sounds straightforward.
345 2012-03-15 04:03:23 <nanotube> cool :)
346 2012-03-15 04:04:08 <devrandom> (but not tonight)
347 2012-03-15 04:04:59 <nanotube> yea no hurry
348 2012-03-15 04:05:17 <nanotube> i've already uploaded the 0.5.3 release today :)
349 2012-03-15 04:05:51 <nanotube> just stick it on your todo list with some reasonable priority :D and thanks for the great work on gitian ;)
350 2012-03-15 04:06:31 <devrandom> glad it's catching on :)
351 2012-03-15 04:07:31 <nanotube> :)
352 2012-03-15 04:08:01 <luke-jr> devrandom: if only it worked with a simple chroot&
353 2012-03-15 04:09:24 <devrandom> luke-jr: I was thinking of implementing lxc support.  would that help?
354 2012-03-15 04:09:36 <luke-jr> don't see any reason it should require LXC :P
355 2012-03-15 04:09:43 <luke-jr> it's just a simple compile after all
356 2012-03-15 04:09:49 <luke-jr> but sure
357 2012-03-15 04:10:43 <devrandom> some compilation scenarios might require root, and I'm worried about malicious software breaking out of chroot if they have root
358 2012-03-15 04:11:02 <devrandom> I'd like gitian to work as a general build environment for untrusted software
359 2012-03-15 04:12:22 <luke-jr> Gentoo does all its building as a non-root user
360 2012-03-15 04:12:44 <phantomcircuit> portage
361 2012-03-15 05:23:48 <[Tycho]> Who created a patch to prevent losing subcent change amounts as fees ?
362 2012-03-15 05:25:34 <chwergy> I don't know but I just found out about it
363 2012-03-15 05:27:13 <Diablo-D3> I AM LOSING MONEYS?!
364 2012-03-15 05:27:33 <Diablo-D3> WHAT SICK MAN MAKES ME LOSE MONEYS
365 2012-03-15 05:27:44 <forsetifox> You're losing monkeys?
366 2012-03-15 05:27:55 <Diablo-D3> actually, I am eating a sandvich
367 2012-03-15 05:28:31 <forsetifox> Did it break Deepbit, Tycho?
368 2012-03-15 05:28:32 <copumpkin> I love sandvichen
369 2012-03-15 05:28:35 <Diablo-D3> two slices of kayam balongna, two slices of white american, two sandvich long cut pickles, horshradish mustard
370 2012-03-15 05:28:47 <[Tycho]> forsetifox: no.
371 2012-03-15 05:29:09 <jm9000> Bologna and american? wtf?
372 2012-03-15 05:29:22 <forsetifox> Good. I've been with your pool for a while and I've never seen so low of rewards on it for the last 12 hours. =/
373 2012-03-15 05:29:58 <[Tycho]> forsetifox: bad luck happens sometimes.
374 2012-03-15 05:30:07 <[Tycho]> Should average out anyway.
375 2012-03-15 05:30:11 <forsetifox> It's never been this bad. O.O
376 2012-03-15 05:30:28 <[Tycho]> No, I think it was already.
377 2012-03-15 05:30:34 <jm9000> I see Kayam bologna is polish. Probably better than that American garbage.
378 2012-03-15 05:30:36 <tcatm> luke-jr: pong
379 2012-03-15 05:31:14 <[Tycho]> forsetifox: yesterday there was already 1 hour without any block in the chain, so bad luck happens with anything :)
380 2012-03-15 05:31:38 <forsetifox> There's a bunch on there. Like 3-4 of them in the last 12 hours.
381 2012-03-15 05:31:56 <[Tycho]> 3-4 of what ?
382 2012-03-15 05:32:09 <forsetifox> More than 1 hour blocks.
383 2012-03-15 05:32:30 <[Tycho]> Hmm, P2SH is at 45.5% atm... Would be funny to support it at 49.9% :)
384 2012-03-15 05:32:47 <[Tycho]> forsetifox: I'm talking about the entire network.
385 2012-03-15 05:32:50 <forsetifox> Yeah. I don't think it's ever been that high.
386 2012-03-15 05:32:55 <forsetifox> Oh.
387 2012-03-15 05:33:01 <forsetifox> Everyone is having bad luck?
388 2012-03-15 05:33:08 <[Tycho]> forsetifox: sometimes.
389 2012-03-15 05:33:29 <forsetifox> When stevie (was his name?) had the luck graphs the bad and good luck just cycled.
390 2012-03-15 05:33:52 <forsetifox> Looked like a very long sine wave on your pool and a normal sine wave on everyone elses.
391 2012-03-15 05:34:01 <[Tycho]> There was one very nice site with pool luck graphs, but it's gone :(
392 2012-03-15 05:34:14 <forsetifox> Yeah. Stevie I think was his name.
393 2012-03-15 05:34:20 <forsetifox> He disappeared.
394 2012-03-15 05:34:20 <[Tycho]> I don't think so.
395 2012-03-15 05:34:42 <forsetifox> Someone in btcguild probably knows who I'm talking about.
396 2012-03-15 05:34:46 <[Tycho]> Going down atm: http://bitcoin.sipa.be/speed-lin-10k.png
397 2012-03-15 05:34:59 <[Tycho]> The url was something like l0ss.net
398 2012-03-15 05:35:24 <forsetifox> Yeah.
399 2012-03-15 05:36:01 <chwergy> is there a limit to the size of the blockchain?
400 2012-03-15 05:36:20 <forsetifox> Don't think so.
401 2012-03-15 05:36:26 <chwergy> o.o
402 2012-03-15 05:36:27 <forsetifox> It's at over a gig now.
403 2012-03-15 05:36:40 <chwergy> thats quite the flaw
404 2012-03-15 05:37:04 <forsetifox> It's necessary for bitcoin to work. Some other systems have shrunk theirs.
405 2012-03-15 05:37:21 <chwergy> yeah it is... thats why it is a flaw.
406 2012-03-15 05:37:43 <forsetifox> Do you even know what the block chain does?
407 2012-03-15 05:38:23 <chwergy> Do you even know how much sense your idioms make?
408 2012-03-15 05:38:36 <forsetifox> Do you always ask a question with a question? =P
409 2012-03-15 05:38:50 <chwergy> Yes
410 2012-03-15 05:39:16 <forsetifox> You should only judge something once you have all the data.
411 2012-03-15 05:39:19 <chwergy> But I'm not socrates, so I don't always answer with a question.
412 2012-03-15 05:39:33 <chwergy> well YOU don't even have all the data
413 2012-03-15 05:39:47 <chwergy> I know what the block chain does
414 2012-03-15 05:39:53 <chwergy> its not difficult to figure out
415 2012-03-15 05:40:34 <chwergy> just saying.... for widespread adoption... it's quite the hurdle
416 2012-03-15 05:40:59 <forsetifox> It's minorly inconvenient once. When you install the bitcoin client.
417 2012-03-15 05:41:02 <chwergy> though could have servers do all the heavy lifting with the block chain..
418 2012-03-15 05:41:27 <chwergy> but if it has no limit... it becomes an HDD size issue
419 2012-03-15 05:42:04 <forsetifox> And HD's don't double each year, right?
420 2012-03-15 05:42:18 <chwergy> theres a limit to that too.
421 2012-03-15 05:42:35 <chwergy> Correction: theres a limit to that
422 2012-03-15 05:42:48 <chwergy> basically its mirroring the oil crisis
423 2012-03-15 05:43:11 <chwergy> we are running a infinite system on a finite resource base
424 2012-03-15 05:43:29 <forsetifox> Oh Odin. You've been alive 20 years and think you know everything?
425 2012-03-15 05:43:40 <forsetifox> Guessing your age since you're in school still. =P
426 2012-03-15 05:44:14 <[Tycho]> chwergy: you are using not the right client for widespread adoption
427 2012-03-15 05:44:16 <Diablo-D3> I KNOW ALL
428 2012-03-15 05:44:22 <Diablo-D3> I HAVE BEEN AROUND SINCE DAWN OF TIME
429 2012-03-15 05:44:24 <forsetifox> Oh shush dark one.
430 2012-03-15 05:44:30 <Diablo-D3> GOD WAS LIKE, DUDE, TAKE MY BEER, WATCH THIS
431 2012-03-15 05:44:32 <chwergy> trying to understand this thing. If you don't have something to add; it doesn't give you a license to mock me.
432 2012-03-15 05:45:05 <chwergy> kwaitwhat tycho?
433 2012-03-15 05:45:11 <forsetifox> You compared the blockchain to the oil crisis. What else am I gonna think?
434 2012-03-15 05:45:24 <chwergy> erm. its a valid comparison
435 2012-03-15 05:46:02 <Habbie> it's not - harddisk space is not something you consume irrevocably
436 2012-03-15 05:46:10 <Habbie> although OS growth sometimes looks that way ;)
437 2012-03-15 05:46:11 <Diablo-D3> we import our bitcoins from foreign countries?!
438 2012-03-15 05:46:35 <Diablo-D3> FUCK YEAH SEAKING!
439 2012-03-15 05:46:54 <jm9000> I prefer Mudkips
440 2012-03-15 05:47:24 <Diablo-D3> I dont get pokemon
441 2012-03-15 05:47:29 <chwergy> I'm saying HDD wont keep doubling in size
442 2012-03-15 05:47:33 <Diablo-D3> it came to the US when I was like in 8th grade
443 2012-03-15 05:47:44 <Joric> what's the difference between a raven and a writting desk... erm... bitcoin blockchain and the oil crisis?
444 2012-03-15 05:47:45 <Diablo-D3> like 2 years ago I finally got around to playing one
445 2012-03-15 05:48:11 <Diablo-D3> I beat it in 2 days
446 2012-03-15 05:48:14 <Diablo-D3> I dont get it =/
447 2012-03-15 05:48:27 <jm9000> You gotta catch them all.
448 2012-03-15 05:48:28 <chwergy> you must be the master of all Diablo
449 2012-03-15 05:48:55 <jm9000> I hope Diablo 3 doesn't suck like 2 did.
450 2012-03-15 05:49:40 <Diablo-D3> Joric: both have quills dipped in ink
451 2012-03-15 05:49:43 <Diablo-D3> chwergy: nope
452 2012-03-15 05:49:46 <Diablo-D3> cant stand the game
453 2012-03-15 05:50:05 <Diablo-D3> my nick predates blizzard entertainment, co
454 2012-03-15 05:50:41 <chwergy> I suppose i should have said "Diablo, you must be..."
455 2012-03-15 05:51:02 <chwergy> anyone run an MC server?
456 2012-03-15 05:51:02 <[Tycho]> chwergy: simple users should use light bitcoin clients that don't need to store the entire blockchain. Full client is for serious business and cryptogeeks.
457 2012-03-15 05:51:14 <chwergy> oh :P
458 2012-03-15 05:51:19 <chwergy> thanks tycho
459 2012-03-15 05:51:39 <chwergy> finding the light bitcoin client is an issue for me i suppose..?
460 2012-03-15 05:51:43 <chwergy> where to getget?
461 2012-03-15 05:51:49 <[Tycho]> Bitcoin forum.
462 2012-03-15 05:51:53 <forsetifox> Bitspinner or something is one?
463 2012-03-15 05:52:03 <bitfoo> chwergy: give electrum a try - http://ecdsa.org/electrum/
464 2012-03-15 05:52:03 <chwergy> oh i use that one on android
465 2012-03-15 05:52:06 <[Tycho]> I think there are a lot already.
466 2012-03-15 05:52:14 <chwergy> thanks bitfoo
467 2012-03-15 05:53:09 <chwergy> though... how to manage wallet if i have a lite client on phone and one on computer?
468 2012-03-15 05:53:44 <forsetifox> Both of them would have the private key, I think?
469 2012-03-15 05:53:53 <forsetifox> These guys would know better than I.
470 2012-03-15 05:54:09 <chwergy> this is where it gets a little fuzzy for me
471 2012-03-15 05:54:18 <chwergy> i just picked this up last night
472 2012-03-15 05:54:28 <Graet> main wallet on pc, phone wallet just enough for every day use, lose phone, lose little
473 2012-03-15 05:54:42 <chwergy> i thought that may be the case
474 2012-03-15 05:54:52 <forsetifox> Graet is wise. Listen to him. =P
475 2012-03-15 05:55:02 <Graet> better than having all your funds on phone imo :)
476 2012-03-15 05:55:05 <chwergy> so how to access wallet.dat when deling with bitspinner?
477 2012-03-15 05:55:08 <Habbie> uh
478 2012-03-15 05:55:11 <Graet> ty forsetifox lol
479 2012-03-15 05:55:24 <Habbie> just make sure your pc has all keys and your phone has the few keys you want to have available for mobile usage
480 2012-03-15 05:55:37 <Graet> i have to go pick kids up from school, gl chwergy
481 2012-03-15 05:55:41 <Habbie> lose your phone, lose nothing; lose your phone, someone finds it before you move the funds for those keys, lose some
482 2012-03-15 05:55:44 <chwergy> thanks graet
483 2012-03-15 05:56:11 <chwergy> how to manage keys?
484 2012-03-15 05:56:19 <chwergy> oh
485 2012-03-15 05:56:28 <chwergy> unencrypt wallet.dat
486 2012-03-15 05:56:34 <chwergy> and look inside?
487 2012-03-15 05:58:22 <chwergy> it seems i cant unencrypt
488 2012-03-15 05:59:15 <[Tycho]> Use key export and import functions
489 2012-03-15 06:01:39 <chwergy> in electrum..
490 2012-03-15 07:35:42 <gribble> New news from bitcoinrss: ancow opened issue 939 on bitcoin/bitcoin <https://github.com/bitcoin/bitcoin/issues/939>
491 2012-03-15 08:38:25 <[Tycho]> Another hour without a block...
492 2012-03-15 10:28:09 <riush> i'm confused.. http://blockexplorer.com/tx/3a5e0977cc64e601490a761d83a4ea5be3cd03b0ffb73f5fe8be6507539be76c - are 0-value outputs really allowed? and more importantly, how can "1" be a valid script sig for that output??
493 2012-03-15 10:29:51 <[Tycho]> Wow.
494 2012-03-15 10:37:14 <lianj> and what did we make wrong here, http://bitcoin.interesthings.de/script/3a5e0977cc64e601490a761d83a4ea5be3cd03b0ffb73f5fe8be6507539be76c:0 its false because the last value of the stack is not 1
495 2012-03-15 10:38:26 <lianj> last line of our Script.run is @stack.pop == 1. should it be different?
496 2012-03-15 10:39:55 <Graet> http://blockchain.info/address/1AJaMSZDGaANx1wd9u83DTznCxkdrJiK8
497 2012-03-15 10:40:04 <Graet> interesting amount of fail there
498 2012-03-15 10:41:07 <[Tycho]> It's luke
499 2012-03-15 10:41:37 <[Tycho]> The first one in the list was sent by me
500 2012-03-15 10:45:13 <[Tycho]> lianj: do you know where the bitcoin script interpreter checks for "1" ?
501 2012-03-15 10:46:16 <lianj> not sure, i was talking about my one :D and wonder if its correct after all
502 2012-03-15 10:47:23 <[Tycho]> But this one is really good find - http://blockexplorer.com/tx/3a5e0977cc64e601490a761d83a4ea5be3cd03b0ffb73f5fe8be6507539be76c
503 2012-03-15 10:48:42 <[Tycho]> Both 0 input AND 0 output.
504 2012-03-15 10:49:02 <lianj> yes but why is it valid. its not valid in my runner http://bitcoin.interesthings.de/script/3a5e0977cc64e601490a761d83a4ea5be3cd03b0ffb73f5fe8be6507539be76c:0
505 2012-03-15 10:49:06 <lianj> haha, yea that too!
506 2012-03-15 10:49:19 <riush> yea, that's spending the 0-value output (without a fee ;)
507 2012-03-15 10:49:47 <[Tycho]> He can mine any TX without fees.
508 2012-03-15 10:50:43 <[Tycho]> lianj: may be because your parser checks for "1", not for 0/empty ?
509 2012-03-15 10:51:22 <[Tycho]> I really was thinking that 0-output is not allowed.
510 2012-03-15 10:51:29 <riush> well, it pushes 1 to the stack, then pushes the data from the output to the stack, then checks if the last item is true-ish, which random data isn't
511 2012-03-15 10:52:09 <[Tycho]> Why random data is not true ?
512 2012-03-15 10:53:13 <lianj> so youre saying it should check is last item on the stack is 0 or empty stack and only then return false?
513 2012-03-15 10:53:31 <[Tycho]> May be. Ask someone else.
514 2012-03-15 10:53:42 <lianj> :)
515 2012-03-15 10:55:33 <[Tycho]> Amazing ! Looks like 0-outputs are possible. This opens a lot of opportunities to me.
516 2012-03-15 10:55:56 <riush> hehe. putting data in the blockchain for free? ;)
517 2012-03-15 10:56:32 <[Tycho]> I was expecting it to be disallowed, but actually only LACK of outputs is prohibited (but would be even better).
518 2012-03-15 10:56:51 <[Tycho]> riush: well, no. I already can put any data there for free.
519 2012-03-15 10:57:18 <lianj> [Tycho]: if i make that change, my tests fail at this assertion  Script.from_string("1 OP_DROP 2").run.should == false
520 2012-03-15 10:57:33 <lianj> what do you guys think, should true or false be right :D
521 2012-03-15 10:58:48 <[Tycho]> lianj: where is that script ?
522 2012-03-15 10:59:23 <lianj> a made up one in my test suite
523 2012-03-15 11:00:19 <[Tycho]> lianj: try it against real bitcoind and find out.
524 2012-03-15 11:00:33 <[Tycho]> Even if bitcoind is wrong you should provide same result :)
525 2012-03-15 11:01:08 <lianj> hehe
526 2012-03-15 11:21:22 <[Tycho]> http://online.wsj.com/article/SB10001424053111904583204576542851688284590.html
527 2012-03-15 11:28:27 <TD> community currencies have a long history for local economic regeneration. see "the future of money" by bernard lietaer for some good discussions of this. brazil is also quite innovative, economically
528 2012-03-15 11:33:56 <jjjrmy-m> Does anyone have an Instawallet account that's been used that I can borrow?
529 2012-03-15 11:37:33 <t7> yes
530 2012-03-15 11:37:43 <t7> let me find the url
531 2012-03-15 12:02:02 <gribble> New news from bitcoinrss: Diapolo opened issue 940 on bitcoin/bitcoin <https://github.com/bitcoin/bitcoin/issues/940>
532 2012-03-15 12:05:15 <denisx> how is it going with BIP30?
533 2012-03-15 12:06:23 <[Tycho]> Deployed and running
534 2012-03-15 12:08:24 <denisx> any news on the 1 transaction miner?
535 2012-03-15 12:08:31 <[Tycho]> No.
536 2012-03-15 12:12:06 <ThomasV> denisx: maybe he wants higher fees
537 2012-03-15 12:13:51 <ThomasV> someone should try transactions with 1btc fees, see if they go in these blocks
538 2012-03-15 12:14:20 <[Tycho]> I doubt this.
539 2012-03-15 12:14:28 <ThomasV> why?
540 2012-03-15 12:14:51 <ThomasV> you think it's an attack?
541 2012-03-15 12:15:09 <[Tycho]> No. I think they don't include TXes for some technical reasons.
542 2012-03-15 12:15:14 <[Tycho]> May be to save time.
543 2012-03-15 12:15:36 <Graet> or effort
544 2012-03-15 12:17:10 <[Tycho]> Even stock bitcoind drops TXes if it has no time to process them.
545 2012-03-15 12:18:00 <sipa> ;;bc,eligius
546 2012-03-15 12:18:01 <gribble> 294205259.776
547 2012-03-15 12:18:48 <[Tycho]> For example this block was created by me: https://blockchain.info/block-index/194033/0000000000000a123cd65bf68764bfd89adea84b66ea250c206d9c94c582d795
548 2012-03-15 12:19:02 <sipa> wow, pool hash rates vary a lot quickly
549 2012-03-15 12:19:12 <sipa> [Tycho]: didn't you have 3900 a few days ago?
550 2012-03-15 12:19:30 <[Tycho]> 4000 is with GPUMAX on
551 2012-03-15 12:19:48 <[Tycho]> Normal speed is 3600, as stated in topic
552 2012-03-15 12:19:53 <sipa> ic
553 2012-03-15 12:20:12 <[Tycho]> But the luck is more than 30% worse now.
554 2012-03-15 12:20:42 <[Tycho]> sipa: did you saw this redeeming ? :) http://blockexplorer.com/tx/3a5e0977cc64e601490a761d83a4ea5be3cd03b0ffb73f5fe8be6507539be76c#i4652380
555 2012-03-15 12:20:50 <[Tycho]> Unusual script.
556 2012-03-15 12:21:05 <[Tycho]> And the first TX I see with both 0 input and output
557 2012-03-15 12:23:09 <sipa> [Tycho]: interesting
558 2012-03-15 12:27:29 <fiddur> What does the bitcoind rpc return if sendfrom is tried with too low transaction fee?    Looking at the code, I would guess -4 with the message "Error: This transaction requires a transaction fee of at least %s because of its amount, complexity, or use of recently received funds".  Catching the needed txfee in a string comparison though is not what I'd like to do...
559 2012-03-15 12:30:46 <sipa> fiddur: afaik it just adds a fee
560 2012-03-15 12:31:48 <fiddur> sipa: So I wont get here:  https://github.com/bitcoin/bitcoin/blob/master/src/wallet.cpp#L1197   ?
561 2012-03-15 12:32:18 <fiddur> ...or is that just if the balance is too low to add the fee perhaps...
562 2012-03-15 12:36:15 <etotheipi_> is there a way to get the satoshi client to finally just give up on a Tx?  I finally got my satoshi wallet migrated into Armory, rescanned, and spent... the Satoshi client even sees it in the blockchain
563 2012-03-15 12:36:35 <etotheipi_> but despite blocks in the blockchain conflicting with this mempool tx, it's still broadcasting...
564 2012-03-15 12:36:51 <sipa> etotheipi_: yes, it assumes its own sends are unquestionably valid
565 2012-03-15 12:37:06 <sipa> i've been meaning to fix that for a long time
566 2012-03-15 12:37:15 <etotheipi_> is there a way to clear its memory pool?
567 2012-03-15 12:37:26 <sipa> etotheipi_: restart it, but the memory pool is not the problem
568 2012-03-15 12:37:43 <sipa> fiddur: sendmany uses CreateTransaction, not SendMoney
569 2012-03-15 12:40:40 <fiddur> sipa: I'm talking about sendfrom, not sendmany.
570 2012-03-15 12:40:54 <sipa> fiddur: oh, i misread
571 2012-03-15 12:41:20 <fiddur> ...ever thought about adding e.g. doxygen docs to the code?  ..or constants for the JSONRPCError codes? :D
572 2012-03-15 12:41:35 <sipa> wumpus maintains a doxygen somewhere
573 2012-03-15 12:41:43 <sipa> and those constants are good idea
574 2012-03-15 12:42:02 <sipa> any, it will add the (minimum) fee automatically, unless there is not enough balance for that
575 2012-03-15 12:42:12 <fiddur> Ok.  Thanks.
576 2012-03-15 12:42:35 <sipa> etotheipi_: it comes from the wallet system, which works remarkably independently from the blockchain code
577 2012-03-15 13:08:27 <etotheipi_> so SIPA, is there a way to just get rid of those tx?
578 2012-03-15 13:08:55 <etotheipi_> they're really irritating me, and there's conflicting tx in the blockchain already
579 2012-03-15 13:09:38 <nanotube> gotta manually edit the wallet using some wallet tools and delete that tx
580 2012-03-15 13:10:30 <etotheipi_> ugh... well I did just write a wallet migration/extraction tool... so I'm digging through bsddb code anyway...
581 2012-03-15 13:12:20 <Joric> etotheipi_, it's complicated
582 2012-03-15 13:12:59 <etotheipi_> I bet...
583 2012-03-15 13:13:29 <Joric> etotheipi_, even opensourced, bsddb it's not reverse-engineering friendly )
584 2012-03-15 13:14:01 <etotheipi_> Joric, btw I got the full migration tool working based on your pywallet code
585 2012-03-15 13:14:05 <etotheipi_> I'm going to send you 5 BTC
586 2012-03-15 13:14:19 <Joric> omg 1JoricCBkW8C5m7QUZMwoRz9rBCM6ZSy96
587 2012-03-15 13:15:45 <etotheipi_> I don't think I would've had the patience to figure it out, without your code
588 2012-03-15 13:15:52 <etotheipi_> (the wallet-encryption stuff)
589 2012-03-15 13:16:03 <sipa> etotheipi_: well, if you'd just remove the tx entries
590 2012-03-15 13:16:36 <Joric_> etotheipi_, thank you very much i've seen it hit the network already
591 2012-03-15 13:16:49 <sipa> etotheipi_: or you could run my very outdated rejectedtx branch
592 2012-03-15 13:16:52 <Joric_> i owe you a lot of beer
593 2012-03-15 13:17:27 <etotheipi_> Joric_, no worries... this is the second or third time you've done something that explicitly helped out
594 2012-03-15 13:17:35 <etotheipi_> (I think the first was OSX compiling)
595 2012-03-15 13:17:47 <sipa> etotheipi_: https://github.com/sipa/bitcoin/commit/70a01202cf3facdb251214ea12fae5a149f3b823
596 2012-03-15 13:17:53 <sipa> but very likely not to merge anymore
597 2012-03-15 13:18:11 <etotheipi_> sipa, :(  I have a feeling I'd have to compile the Satoshi client to use that, right?
598 2012-03-15 13:18:34 <sipa> oh, certainly
599 2012-03-15 13:18:45 <sipa> but i'm quite sure it won't even compile anymore
600 2012-03-15 13:18:52 <sipa> it's also on my todo list to update it ;)
601 2012-03-15 13:19:01 <Joric_> etotheipi_, i'm trying to write faster indexing https://github.com/joric/pyblockchain/blob/master/cppblockchain.cpp
602 2012-03-15 13:19:58 <Joric> damn provider
603 2012-03-15 13:19:59 <sipa> bah, was stupid, apparently i was in the wrong directory, with a 0.3.19 bitcoind in it, starting it, i got database errors... tried everything and eventually deleted my blockchain... then realized
604 2012-03-15 13:20:09 <Joric> https://github.com/joric/pyblockchain/blob/master/cppblockchain.cpp  takes ~2 minutes to load entire blockchain / calculate balances without loading completely into ram
605 2012-03-15 13:20:43 <Joric> i'm going to load blocks into 10mb buffer to avoid extra seekg while calculating tx hashes
606 2012-03-15 13:20:47 <etotheipi_> Joric... what I do in Armory takes about 10-45s depending on your hardware, and about double that if I use mmap
607 2012-03-15 13:21:19 <etotheipi_> the mmap solution is a medium-term solution that I am going to implement which allows me to let the system decide how much RAM to use
608 2012-03-15 13:22:25 <Eliel> perhaps check the amount of RAM the system has on startup and choose between mmap and your current system based on whether there's enough memory?
609 2012-03-15 13:22:38 <etotheipi_> Eliel, mmap basically already does that
610 2012-03-15 13:22:57 <etotheipi_> on my system with 16 GB of RAM, using the mmap solution runs almost as smoothly as the current full-RAM
611 2012-03-15 13:23:14 <etotheipi_> it just takes a little longer to do the initial scan (but it will rescan for new address balances in less than 1s after that)
612 2012-03-15 13:23:38 <Eliel> but didn't you just say mmap made it slower?
613 2012-03-15 13:23:56 <sipa> it can be
614 2012-03-15 13:24:07 <etotheipi_> it's a little bit slower
615 2012-03-15 13:24:31 <etotheipi_> on a system with a lot of RAM, it's not really noticeable... it maybe goes from 20s load time to 30s load time
616 2012-03-15 13:24:53 <etotheipi_> maybe not even that much... 25s?
617 2012-03-15 13:25:12 <Joric> i was thinking about exporting to sqlite https://github.com/joric/pyblockchain/blob/master/sqlite.py but it takes about 30 minutes )
618 2012-03-15 13:25:18 <etotheipi_> where mmap makes a difference is on systems with 1 GB of RAM
619 2012-03-15 13:25:24 <Eliel> is that one time thing or every startup? It's way too slow if it happens every time.
620 2012-03-15 13:25:36 <Joric> and will probably weight a lot more than the blockchain itself
621 2012-03-15 13:25:37 <etotheipi_> Eliel, atm it's every startup
622 2012-03-15 13:26:01 <etotheipi_> that's why it's a "medium-term" solution... I need to mitigate the fact that the blockchain is soon not going to fit into *anyone's* RAM
623 2012-03-15 13:26:12 <Eliel> yes, true.
624 2012-03-15 13:26:15 <etotheipi_> then I'll work on something more robust and long-term
625 2012-03-15 13:26:26 <etotheipi_> or integrate libcoin or something
626 2012-03-15 13:27:03 <etotheipi_> but given how much I love data structures (it's a bizarre hobby), I might just rework it myself
627 2012-03-15 13:27:24 <sipa> etotheipi_: what license is armory?
628 2012-03-15 13:27:43 <etotheipi_> I really need a good PATRICIA tree impl in C++
629 2012-03-15 13:27:55 <etotheipi_> sipa, ATM it's AGPL3
630 2012-03-15 13:27:59 <sipa> ic
631 2012-03-15 13:28:20 <etotheipi_> with dual-licensing negotiable
632 2012-03-15 13:28:57 <etotheipi_> I may loosen it up later... but it's not easy to go back from less-restrictive to more-restrictive if I change my mind later
633 2012-03-15 13:29:04 <sipa> true\n3134766
634 2012-03-15 13:29:26 <sipa> etotheipi_: have you seen stanford's online crypto course?
635 2012-03-15 13:29:45 <etotheipi_> sipa, I've heard about it from Raccoon's spamming :)
636 2012-03-15 13:29:49 <sipa> haha
637 2012-03-15 13:29:55 <Joric> PATRICIA tree omg i just used std::map (a self-balancing rb-tree if i remember right)
638 2012-03-15 13:30:13 <etotheipi_> Joric, you know PATRICIA trees?
639 2012-03-15 13:30:16 <etotheipi_> so few people do!
640 2012-03-15 13:30:39 <sipa> it's a binary trie with 'compacted' branches, right?
641 2012-03-15 13:30:40 <Joric> i do know
642 2012-03-15 13:30:43 <Joric> now
643 2012-03-15 13:30:53 <etotheipi_> I remember staying up all night with a buddy, about 16 hours straight, trying to figure out the "insert" function for a PATRICIA tree
644 2012-03-15 13:31:05 <etotheipi_> (it was for a data-structures homework)
645 2012-03-15 13:31:30 <sipa> etotheipi_: the crypto couse is nice though; i'm already through the first week's lectures and exercises
646 2012-03-15 13:31:33 <sipa> *course
647 2012-03-15 13:31:51 <etotheipi_> sipa, maybe I should get over my pride and do it...
648 2012-03-15 13:32:11 <etotheipi_> I have taken a grad-level course on crypto... but it never hurts to take more and risk learning something new :)
649 2012-03-15 13:32:12 <sipa> (i don't expect to learn too much practically, but i like seeing some formalism behind crypto)
650 2012-03-15 13:33:04 <sipa> i never had a course on crypto at the university; makes me wonder why i never tried to do it as an optional one though
651 2012-03-15 13:33:28 <etotheipi_> I was able to fit it into my Applied Math master's
652 2012-03-15 13:33:32 <etotheipi_> and totally worth it
653 2012-03-15 13:33:46 <etotheipi_> I got to do my final project on WEP
654 2012-03-15 13:34:05 <etotheipi_> that was fun:  WEP is a disaster, cryptographically...
655 2012-03-15 13:34:08 <sipa> the first programming exercise was: here are some messages encrypted with the same one time pad, decrypt it
656 2012-03-15 13:35:48 <etotheipi_> I found it really helpful to use python+sage when we started doing RSA and ECC
657 2012-03-15 13:36:21 <etotheipi_> I'm going to try to find that final project one WEP -- it's comical how many problems there are with it
658 2012-03-15 13:36:23 <sipa> If you've learned that much already, maybe the stanford course will have little to offer.
659 2012-03-15 13:37:27 <sipa> etotheipi_: actually, the lecturer has already mentioned more than "WEP is a really nice thing, there are so many things wrong with it i can often refer to it as an example how not to do things"
660 2012-03-15 13:37:46 <sipa> *once
661 2012-03-15 13:37:53 <Joric> 'in 2008 the GPL family was 70 percent of licenses. As of December of 2011, it was 57 percent.' http://www.cio.com/article/698663/How_Open_Source_Licenses_Affect_Your_Business_and_Your_Developers
662 2012-03-15 13:39:19 <etotheipi_> sipa, yeah that's exactly how I feel about WEP... it's a perfect sandbox for playing with all the things that you can do wrong
663 2012-03-15 13:40:03 <etotheipi_> my favorite attack was the fact that is uses just a simple XOR mask of the packets, and you know the structure, so you know where the IP address is in the packet
664 2012-03-15 13:40:57 <etotheipi_> so you can intercept the encrypted packets, flip some bits in the IP field, and have the router send it to one of your own IP addresses (it will decrypt it to send it externally)
665 2012-03-15 13:43:30 <etotheipi_> and the WEP protocol doesn't require any consecutive IVs or anything... so you can replay a packet as many times as you want
666 2012-03-15 13:48:12 <etotheipi_> that's a fun first problem set
667 2012-03-15 13:50:01 <Joric> i like the westheimer's rule 'To estimate the time it takes to do a task: estimate the time you think it should take, multiply by two and change the unit of measure to the next highest unit. Thus, we allocate two days for a one hour task.'
668 2012-03-15 13:56:07 <luke-jr> tcatm: 0.5.3 is ready
669 2012-03-15 13:56:24 <luke-jr> [Tycho]: I did that a long time ago
670 2012-03-15 13:58:01 <tcatm> luke-jr: There's a NACK from BlueMatt. Can you resolve that first (either get a consensus or remove that commit)?
671 2012-03-15 13:58:23 <luke-jr> lianj: top stack just has to be non-zero to succeed
672 2012-03-15 13:58:42 <luke-jr> Graet: that address is my "automatically try to spend everything" script
673 2012-03-15 13:59:55 <luke-jr> Joric: the total amount of GPL has still gone up, though.
674 2012-03-15 14:00:48 <luke-jr> sipa: ping
675 2012-03-15 14:01:37 <luke-jr> tcatm: I don't care strongly either way. I'll just remove it, unless sipa says it's fine :p
676 2012-03-15 14:02:03 <luke-jr> tcatm: Matt's NACK is based on a false premise, though - by his rule, we'd need to remove Satoshi :p
677 2012-03-15 14:03:32 <gavinandresen> Satoshi can have write access to the github repo any time he asks.
678 2012-03-15 14:03:51 <luke-jr> lol
679 2012-03-15 14:04:20 <luke-jr> gavinandresen: you're welcome to answer instead of sipa, didn't realize you were here/active :p
680 2012-03-15 14:04:34 <gavinandresen> I agree with Matt
681 2012-03-15 14:04:57 <luke-jr> tcatm: removed the commit
682 2012-03-15 14:05:55 <tcatm> merged.
683 2012-03-15 14:07:50 <lianj> luke-jr: empty stack is true too?
684 2012-03-15 14:07:58 <luke-jr> lianj: no, empty stack is false
685 2012-03-15 14:08:08 <Joric> stack is innocent
686 2012-03-15 14:08:15 <lianj> ok thanks
687 2012-03-15 14:08:21 <luke-jr> lianj: I think strictly speaking, you can get multibyte falses too
688 2012-03-15 14:10:27 <luke-jr> gavinandresen: tcatm: on a bit more thought, I do disagree with Matt simply on the basis that bitcoin.org should not prefer a specific client :p
689 2012-03-15 14:10:47 <luke-jr> but I guess that status quo takes more work to change than just the author list
690 2012-03-15 14:12:29 <tcatm> I don't like hosting the rendered files statically on github anymore :/ There service hooks seem to be broken again and thus bitcoin.org is not updating.
691 2012-03-15 14:12:42 <luke-jr> (and doesn't necessarily put me on the list, since Spesmilo isn't really maintained anymore)
692 2012-03-15 14:12:52 <luke-jr> tcatm: I tried the manual hook
693 2012-03-15 14:16:09 <luke-jr> tcatm: (it didn't work)
694 2012-03-15 14:19:16 <tcatm> https://github.com/bitcoin/bitcoin.github.com/blob/master/index.html has updated just fine.
695 2012-03-15 14:27:57 <Joric> stallman approves http://bitcoinmedia.com/stallman-likes-bitcoin/
696 2012-03-15 14:32:32 <gavinandresen> uh oh
697 2012-03-15 14:34:20 <helo> "His statement that bitcoin is not currently anonymous by default but can be made anonymous is totally correct." <--- not what stallman said
698 2012-03-15 14:34:35 <helo> "Bitcoin, I believe, can be used anonymously although its not inherently anonymous" <-- what he actually said
699 2012-03-15 14:36:10 <[Tycho]> lianj: looks like I was right, the bitcoin wiki says "A transaction is valid if nothing in the combined script triggers failure and the top stack item is true (non-zero)"
700 2012-03-15 14:45:00 <sipa> luke-jr: what was it about?
701 2012-03-15 14:46:41 <luke-jr> sipa: whether or not to add me to the developer list on bitcoin.org
702 2012-03-15 14:47:51 <sipa> ic
703 2012-03-15 14:50:35 <bitcoin-otc-uk> Hi iam pleased to announce the launch of #bitcoin-otc-uk . Please take the time to have a look Thankyou.
704 2012-03-15 14:50:57 <jrmithdobbs> please take the time to go fuck yourself
705 2012-03-15 14:51:13 <luke-jr> bitcoin-otc-uk: sorry, we already have a channel for that
706 2012-03-15 14:51:17 <luke-jr> it's called #bitcoin-otc
707 2012-03-15 14:51:53 <sipa> jrmithdobbs: is that necessary? (i agree it's not very on-topic here, though)
708 2012-03-15 14:52:18 <jrmithdobbs> sipa: sorry, pissy mood (shit goin down at work) and just glanced up and saw spam, it was not necessary or appropriate.
709 2012-03-15 14:52:26 <jrmithdobbs> ;p
710 2012-03-15 14:52:38 <gavinandresen> drama in the otc world would be a good discussion for #bitcoin, imho
711 2012-03-15 14:52:41 <luke-jr> nor*
712 2012-03-15 14:53:00 <luke-jr> gavinandresen: do you have time to audit my BIP 16 backport by any chance?
713 2012-03-15 14:53:16 <bitcoin-otc-uk> big charges exchangeing. for paypal in uk from US
714 2012-03-15 14:53:28 <luke-jr> bitcoin-otc-uk: so?
715 2012-03-15 14:53:32 <jrmithdobbs> bitcoin-otc-uk: go. away.
716 2012-03-15 14:53:47 <gavinandresen> luke-jr: 0.5.something backport?  maybe-- although I'd much rather people just upgraded to 0.6
717 2012-03-15 14:54:29 <sipa> Did BlueMatt get any progress on locating the win32-qt-rpc bug last night?
718 2012-03-15 14:54:37 <luke-jr> http://luke.dashjr.org/programs/bitcoin/w/bitcoind/luke-jr.git/commitdiff/b0dc119b75d5f0d19c9aef1047f6083f5b3de638
719 2012-03-15 14:57:12 <lianj> luke-jr: [Tycho]: yea, thanks again. fixed it in my runner
720 2012-03-15 15:00:08 <gavinandresen> luke-jr: if the goal is minimal support, then a much smaller patch is possible.  You just need the fully-verify BIP16 code from script.cpp, but treat BIP16 transactions as non-standard (don't put them in your memory pool, don't mine them)
721 2012-03-15 15:00:40 <luke-jr> gavinandresen: hmm, good point
722 2012-03-15 15:01:03 <luke-jr> (and the sigop counting)
723 2012-03-15 15:01:32 <gavinandresen> If you assume nobody is going to waste time mining blocks with out-of-bounds BIP16 sigop counts then you could get away with not counting them....
724 2012-03-15 15:01:59 <gavinandresen> (you would risk building on top of a bad block, though)
725 2012-03-15 15:02:02 <sipa> that's the same as assuming nobody is going to mine invalid BIP16 transactions at all, no?
726 2012-03-15 15:02:26 <sipa> (i.e. not backporting anything at all)
727 2012-03-15 15:04:03 <gavinandresen> sipa:  I suppose the danger would be lazy miners who don't upgrade and have very-expensive BIP16-spending transactions in their memory pool... they might create a block that exceeds the max sigops number and not know it....
728 2012-03-15 15:05:16 <gavinandresen> So, yeah, the backport really needs to count BIP16 sigops, because otherwise an attacker could leverage a lazy miner to create blocks that the backported code might build on.
729 2012-03-15 15:05:25 <sipa> Exactly.
730 2012-03-15 15:05:58 <gavinandresen> (not an easy/cheap attack, since very-expensive-BIP16 transactions will have high fees....)
731 2012-03-15 15:06:21 <sipa> Is there a 0.5.3 release announcement on the forums?
732 2012-03-15 15:07:15 <sipa> (I'd like to do a BIP30-related post in the mining subforum, but preferably there is a release announcement to refer to)
733 2012-03-15 15:09:38 <Joric> does blockchain contain orphaned blocks atm?
734 2012-03-15 15:10:03 <luke-jr> sipa: wiaitng for bitcoin.org to update
735 2012-03-15 15:10:31 <etotheipi_> Joric, you blockchain file may include invalid blocks... but someone redownloading the chain won't see them
736 2012-03-15 15:10:31 <sipa> luke-jr: it is updated
737 2012-03-15 15:10:35 <sipa> (you need to refresh)
738 2012-03-15 15:10:48 <sipa> Joric: yes
739 2012-03-15 15:11:16 <sipa> etotheipi_: depends what kind of invalidness
740 2012-03-15 15:11:26 <sipa> (only unconnectable blocks, afaik)
741 2012-03-15 15:11:28 <Joric> i mean while i'm parsing it should i build merkle tree or something? it's not enough to scan it consequently?
742 2012-03-15 15:11:46 <sipa> Joric: not a merkle tree, just a tree
743 2012-03-15 15:13:37 <etotheipi_> Joric, the way I do it is that I read all blocks (headers) into a map indexed by their hash... then when I'm done I go through the map (in arbitrary order) and walk down the chain (using prevHash to get to the prev block)
744 2012-03-15 15:14:07 <etotheipi_> then I stop once I either hit a hash I don't know, or the genesis block
745 2012-03-15 15:14:45 <etotheipi_> it's a bit more complicated than that... but that's the gist of how I organize the chain and detect orphans
746 2012-03-15 15:15:55 <luke-jr> sipa: https://bitcointalk.org/index.php?topic=68895.0
747 2012-03-15 15:16:19 <luke-jr> tcatm: ping, update forum version thing
748 2012-03-15 15:16:40 <etotheipi_> Joric, here's the way I do it:  https://github.com/etotheipi/BitcoinArmory/blob/master/cppForSwig/BlockUtils.cpp#L2393
749 2012-03-15 15:16:56 <etotheipi_> it takes less than 1/20 of a second to organize the chain once all the blocks are indexed in the map
750 2012-03-15 15:17:30 <etotheipi_> that method also detects the longest chain
751 2012-03-15 15:17:42 <etotheipi_> and marks any invalid branches
752 2012-03-15 15:27:26 <BlueMatt> wumpus: you dont happen to have an old build of bitcoin-qt.exe sitting around that was built on windows?
753 2012-03-15 15:28:07 <andytoshi> maybe it's time to haul Knuth out and read the chapters on searching/sorting
754 2012-03-15 15:28:09 <andytoshi> with tape drives and their ilk :P
755 2012-03-15 15:29:17 <andytoshi> etotheipi_ and Joric were discussing blockchain organization
756 2012-03-15 15:29:25 <BlueMatt> ah
757 2012-03-15 15:29:50 <BlueMatt> sipa: "Did BlueMatt get any progress on locating the win32-qt-rpc bug last night?" afaict we are (somehow) building without exception support
758 2012-03-15 15:29:57 <BlueMatt> still trying to figure out how/why
759 2012-03-15 15:30:54 <etotheipi_> andytoshi, searching and sorting are my specialty!  although I could be doing better if I can find an acceptable PATRICIA tree implementation
760 2012-03-15 15:31:45 <andytoshi> cool! i thought you were a physicist..
761 2012-03-15 15:32:07 <etotheipi_> andytoshi, I am... but strangely, my favorite class in undergrad was a brutal data-structures class
762 2012-03-15 15:32:55 <etotheipi_> I guess I do so much programming/problem-solving, that the optimizations from using the right data structures really made me happy
763 2012-03-15 15:34:57 <etotheipi_> having to use binary search trees to organize tx data seems *sooo* inefficient
764 2012-03-15 15:35:08 <etotheipi_> and it doesn't allow me to do prefix searching...
765 2012-03-15 15:35:47 <etotheipi_> :)
766 2012-03-15 15:35:47 <sipa> search for the prefix's previous node, then iterate?
767 2012-03-15 15:39:33 <sipa> etotheipi_: it's not optimally efficient, but at least it's O(log n)
768 2012-03-15 15:40:14 <etotheipi_> so how do you do it in a binary search tree?  all my keys are 32 bytes, and I have a 5-byte prefix... I want to get all nodes that start with that prefix
769 2012-03-15 15:40:43 <etotheipi_> first I have to make sure my operator< and == are written to accommodate this searching
770 2012-03-15 15:41:05 <sipa> you can search for the first node that sorts after your prefix, right?
771 2012-03-15 15:41:39 <sipa> and then do depth-first walking from there, until you hit a node that sorts after the prefix
772 2012-03-15 15:43:45 <etotheipi_> so I guess I could make it:  prefix0:  true, I can find the first item that is greater than <prefix>000000 and then do a sorted extraction of that branch up to <prefix>FFFFFFF
773 2012-03-15 15:44:04 <etotheipi_> * err... skip the first line before "true"
774 2012-03-15 15:44:04 <sipa> yes, indeed
775 2012-03-15 15:44:41 <etotheipi_> the question is how to get that actually implemented using std::maps
776 2012-03-15 15:44:48 <sipa> not only of that branch, right, you may need to go up after exploring the entire subbranch
777 2012-03-15 15:45:01 <sipa> but if you only want the first match, you'll have it in one step
778 2012-03-15 15:45:20 <sipa> (either it's immediately after <prefix>fffff, or it's not and you have a match)
779 2012-03-15 15:45:53 <etotheipi_> that's straightforward if I wrote the map/tree class myself
780 2012-03-15 15:46:09 <etotheipi_> I have to go check the documentation to see if I can do that with std::map
781 2012-03-15 15:46:38 <etotheipi_> I think std::map is going to give me a yes/no whether "txHash" exists... no function for "get me the closest element..."
782 2012-03-15 15:47:01 <sipa> etotheipi_: there is, upper_bound
783 2012-03-15 15:47:06 <etotheipi_> oh, you're right
784 2012-03-15 15:47:17 <sipa> and then use ++ on the returned iterator
785 2012-03-15 15:47:18 <etotheipi_> just found "lower_bound" and "upper_bound"
786 2012-03-15 15:47:40 <etotheipi_> so ++ will increment the iterator in sort-order?
787 2012-03-15 15:47:44 <sipa> yes
788 2012-03-15 15:47:51 <sipa> it does a DFS
789 2012-03-15 15:48:04 <etotheipi_> (sorry, these are details of STL libraries that I was never sure about)
790 2012-03-15 15:48:07 <sipa> afaik
791 2012-03-15 15:48:31 <sipa> etotheipi_: well, i never wrote any serious C++ before I started with bitcoin