1 2012-04-10 00:00:24 <gmaxwell> midnightmagic: you seemed to be happy to misrepresent drepper's position. ::shrugs:: On the balance I think it's probably a better tradeoff, I only commented because I think its just not right to describe this as a screw-the-idiots. ::shrugs::
  2 2012-04-10 00:00:35 <gmaxwell> In particular http://sources.redhat.com/ml/libc-alpha/2000-08/msg00061.html "Dammit, it is not safe"
  3 2012-04-10 00:03:08 <midnightmagic> i'm not making any representations about drepper's specific position. but thanks for putting words in my mouth again.
  4 2012-04-10 00:04:35 <luke-jr> ok, I really didn't mean to spark up a big heated debate when I asked about strl*&
  5 2012-04-10 00:04:58 <gmaxwell> midnightmagic: then what were you aruging with me for?
  6 2012-04-10 00:05:06 <sipa> i hope we can get rid of both strncpy and strlcpy, and move to C++ constructions instead...
  7 2012-04-10 00:05:12 <midnightmagic> I wasn't. you started arguing with me again.
  8 2012-04-10 00:05:33 <luke-jr> &
  9 2012-04-10 00:06:14 <gmaxwell> Because my response to you was "I think the position is a bit more complicated than that, it's that screwing up with stlcpy is still unsafe a false sense of security" a fine comment you could have ignored since it wasn't disagreeing with you, just adding information.. then you contradicted with "the position boils down to elitist attitudes in core devs"
 10 2012-04-10 00:06:26 <gmaxwell> but ::shrugs:: thats how I percieved it.
 11 2012-04-10 00:07:09 <luke-jr> nanotube: poke
 12 2012-04-10 00:07:33 <gmaxwell> midnightmagic: Sorry I left you feeling like I was picking an argument with you. t'wasn't intended.
 13 2012-04-10 00:07:51 <midnightmagic> that is not actually contradicting you. boiling down a complex issue to its core is acknowledging the complexity, while making an assertion that core devs in BSDs are elitist.
 14 2012-04-10 00:08:37 <midnightmagic> in other words, I'm actually agreeing with you, but trying to say that the core devs tend to be motivated to come up with lots of reasons why they don't want to make the user experience any easier. :-(
 15 2012-04-10 00:09:24 <midnightmagic> so, i'm just making a comment about core devs motivations is all, because i unfortunately have more direct experience with it than i'd like. :-(
 16 2012-04-10 00:14:24 <midnightmagic> http://sources.redhat.com/ml/libc-alpha/2000-08/msg00053.html "Beside, those who are using strcat or variants deserved to be punished."  <-- this is the kind of elitism I'm talking about when i say "18:37 < midnightmagic> the others often seem to be of the opinion that if you don't know what you're doing, you deserve what you get if you misuse strncpy/etc"
 17 2012-04-10 00:24:44 <midnightmagic> gah brutal, drepper works for goldman sachs now..?!
 18 2012-04-10 00:31:06 <coingenuity> midnightmagic: of course they're poaching bitcoin talent :)
 19 2012-04-10 00:38:10 <luke-jr> sipa: you broke everything :C
 20 2012-04-10 00:38:22 <luke-jr> sipa: starting with getalltransactions
 21 2012-04-10 00:55:59 <nanotube> luke-jr: poke back. :)
 22 2012-04-10 00:59:46 <sipa> luke-jr: right, i'll rebase them tomorrow
 23 2012-04-10 01:19:19 <da2ce7> sipa: how is your hd wallets bip comming along?
 24 2012-04-10 01:21:24 <kingkatari> anyone familer with the setup or config of poolserverj i just have a easy question
 25 2012-04-10 01:25:05 <gribble> New news from bitcoinrss: dooglus opened issue 1073 on bitcoin/bitcoin <https://github.com/bitcoin/bitcoin/issues/1073>
 26 2012-04-10 02:23:14 <gribble> shadders was last seen in #bitcoin-dev 13 weeks, 0 days, 2 hours, 12 minutes, and 52 seconds ago: <shadders> or that just java :p
 27 2012-04-10 02:23:14 <kingkatari> ;; seen shadders
 28 2012-04-10 02:29:22 <kingkatari> Hey all i have a miningpool tech question maybe you all can help
 29 2012-04-10 02:29:51 <kingkatari> listen.http.json-rpc.port=   referes to the port people connect their miner on right?
 30 2012-04-10 02:31:48 <kingkatari> listen.http.json-rpc.port=   referes to the port people connect their miner on right?
 31 2012-04-10 02:36:24 <kingkatari> sup bob
 32 2012-04-10 02:45:39 <kingkatari> is anyone even awake?
 33 2012-04-10 02:49:56 <kingkatari> SUp Slix
 34 2012-04-10 02:50:29 <Slix`> kingkatari: Hi.
 35 2012-04-10 02:50:44 <kingkatari> Hey Slix do you mine?
 36 2012-04-10 02:50:58 <Slix`> kingkatari: Nope.
 37 2012-04-10 02:51:03 <kingkatari> damn
 38 2012-04-10 02:51:27 <kingkatari> Sup Some
 39 2012-04-10 02:52:55 <kingkatari> I need to find someome that mines bitcoins to run a test for me
 40 2012-04-10 02:57:25 <doublec> try #bitcoin-mining (or something like that)
 41 2012-04-10 03:00:37 <kingkatari> been there it is a dead room no one awake
 42 2012-04-10 03:19:33 <kingkatari> sup some how are ya
 43 2012-04-10 03:40:48 <kingkatari> Sup NGL how are ya
 44 2012-04-10 03:41:20 <gribble> Error: 'help' is not a defined function.
 45 2012-04-10 03:41:20 <kingkatari> ;; calc help
 46 2012-04-10 03:41:28 <kingkatari> ;; help
 47 2012-04-10 03:41:29 <gribble> The bot responds when you start a line with the ! character. A good starting point for exploring the bot is the !facts command. You can also visit the bot's website for a list of help topics and documentation: http://gribble.sourceforge.net/
 48 2012-04-10 03:48:18 <kingkatari> !bc,totalbc
 49 2012-04-10 03:48:19 <gribble> 8752150.00000000
 50 2012-04-10 03:49:25 <gribble> The average time to generate a block at 400000 Khps, given current difficulty of 1626553.4813289 , is 28 weeks, 6 days, 3 hours, 23 minutes, and 5 seconds
 51 2012-04-10 03:49:25 <kingkatari> !bc,calc 400000
 52 2012-04-10 03:49:38 <kingkatari> !bc,calc 800000
 53 2012-04-10 03:49:39 <gribble> The average time to generate a block at 800000 Khps, given current difficulty of 1626553.4813289 , is 14 weeks, 3 days, 1 hour, 41 minutes, and 32 seconds
 54 2012-04-10 03:49:57 <kingkatari> !bc,calc 1000000
 55 2012-04-10 03:49:58 <gribble> The average time to generate a block at 1000000 Khps, given current difficulty of 1626553.4813289 , is 11 weeks, 3 days, 20 hours, 33 minutes, and 14 seconds
 56 2012-04-10 03:50:12 <kingkatari> !bc,calc 50000000
 57 2012-04-10 03:50:13 <gribble> The average time to generate a block at 50000000 Khps, given current difficulty of 1626553.4813289 , is 1 day, 14 hours, 48 minutes, and 39 seconds
 58 2012-04-10 03:51:42 <gribble> (bc,gend <an alias, 2 arguments>) -- Alias for "echo The expected generation output, at $1 Khps, given the supplied difficulty of $2, is [math calc 50*24*60*60 / (1/((2**224-1)/$2*$1*1000/2**256))] BTC per day and [math calc 50*60*60 / (1/((2**224-1)/$2*$1*1000/2**256))] BTC per hour.".
 59 2012-04-10 03:51:42 <kingkatari> !bc,gend 50000000
 60 2012-04-10 03:52:38 <kingkatari> !bc,stats
 61 2012-04-10 03:52:40 <gribble> Current Blocks: 175042 | Current Difficulty: 1626553.4813289 | Next Difficulty At Block: 175391 | Next Difficulty In: 349 blocks | Next Difficulty In About: 2 days, 14 hours, 25 minutes, and 56 seconds | Next Difficulty Estimate: 1569812.13645364 | Estimated Percent Change: -3.48844016052
 62 2012-04-10 03:53:06 <kingkatari> !bc,gend 50000000 1626553.4813289
 63 2012-04-10 03:53:07 <gribble> The expected generation output, at 50000000 Khps, given the supplied difficulty of 1626553.4813289, is 30.9190073407 BTC per day and 1.28829197253 BTC per hour.
 64 2012-04-10 03:53:44 <gribble> The expected generation output, at 500000000 Khps, given the supplied difficulty of 1626553.4813289, is 309.190073407 BTC per day and 12.8829197253 BTC per hour.
 65 2012-04-10 03:53:44 <kingkatari> !bc,gend 500000000 1626553.4813289
 66 2012-04-10 05:20:35 <Diapolo> sipa: Are you on? I like your patch and vote for merging it in before my util update as I would like to rebase based on your patch :).
 67 2012-04-10 08:50:05 <BlueMattBot> Project Bitcoin build #275: STILL FAILING in 1 min 23 sec: http://jenkins.bluematt.me/job/Bitcoin/275/
 68 2012-04-10 08:50:06 <BlueMattBot> mauron: Added comment lines in French & Turkish which basically mean "Bitcoin, virtual P2P cryptocurrency".
 69 2012-04-10 08:57:08 <gribble> New news from bitcoinrss: Diapolo opened issue 1074 on bitcoin/bitcoin <https://github.com/bitcoin/bitcoin/issues/1074>
 70 2012-04-10 09:21:57 <sipa> wumpus: i'm adding the date of the latest commit also to version.cpp
 71 2012-04-10 09:22:06 <sipa> any suggestion on how to format that in the about box
 72 2012-04-10 09:36:47 <t7> Diablo-D3: did you write that opencl code from scratch?
 73 2012-04-10 09:54:47 <Diablo-D3> t7: the short answer: yes. the long answer: not quite.
 74 2012-04-10 09:55:19 <t7> is it just unrolled loops
 75 2012-04-10 09:55:27 <sipa> longer answer: there's been tons of OpenCL kernels used for bitcoin mining, and many of them reused parts of or were inspired by previous ones
 76 2012-04-10 09:56:43 <Diablo-D3> t7: it started as a hack of someone elses
 77 2012-04-10 09:56:46 <Diablo-D3> which was hacked by someone else
 78 2012-04-10 09:56:50 <Diablo-D3> and then I rehacked that
 79 2012-04-10 09:56:55 <Diablo-D3> and then I hacked it some more
 80 2012-04-10 09:57:03 <Diablo-D3> and then I threw more hacks on it
 81 2012-04-10 09:57:18 <Diablo-D3> and then I wrote a perl script that is essentially an opencl compiler that outputs opencl
 82 2012-04-10 09:57:28 <sipa> the first public opencl kernel was written by m0mchil for poclbm, afaik
 83 2012-04-10 09:57:43 <t7> Diablo-D3: compiles from what?
 84 2012-04-10 09:58:42 <Diablo-D3> opencl.
 85 2012-04-10 09:58:45 <Diablo-D3> sipa: no
 86 2012-04-10 09:58:51 <Diablo-D3> m0's was a hack of someone else's
 87 2012-04-10 09:58:56 <Diablo-D3> and mine started out as a hack of his
 88 2012-04-10 10:00:35 <t7> Diablo-D3: like an optimizer ?
 89 2012-04-10 10:01:20 <Diablo-D3> no, like a full blown compiler.
 90 2012-04-10 10:02:12 <Diablo-D3> it contains the worlds shittiest opencl parser, written using line noise, and then it builds a full AST, and then proceeds to count op timing and shuffle instructions around for peak performance
 91 2012-04-10 10:24:26 <t7> lol u silly billy
 92 2012-04-10 10:33:13 <fimpfimp> I just sent away some BTC from bitcoind. When I do getbalance ACCOUNT, it's not updated. When i do getbalance with no argument, it is updated correctly. Bug?
 93 2012-04-10 10:33:34 <fimpfimp> "getbalance" yields 29, "getbalance ACCOUNT" yields 190. A single account cant hold more than the total, right?
 94 2012-04-10 10:33:49 <fimpfimp> now I returned the amount and "getbalance ACCOUNT" now shows the double amount
 95 2012-04-10 10:34:07 <sipa> how did you return it?
 96 2012-04-10 10:34:39 <fimpfimp> aha, so "" account has a balance of -160
 97 2012-04-10 10:34:59 <sipa> when sending money from an account, you must use sendfrom, and specify that account
 98 2012-04-10 10:35:02 <fimpfimp> I returned it to an address of the specified account
 99 2012-04-10 10:35:11 <sipa> otherwise it is taken from the "" account
100 2012-04-10 10:35:42 <sipa> so if you do a send-to-self from the "" account, to an address of ACCOUNT, yes, the balance of ACCOUNT will increase
101 2012-04-10 10:36:16 <sipa> also, they way unconfirmed transactions are counted differs between the account balance and the total balance calculations
102 2012-04-10 10:36:25 <sipa> so that could explain why those were different initially
103 2012-04-10 10:36:51 <fimpfimp> ok, thanks
104 2012-04-10 10:47:53 <freewil> whats the default rpc username?
105 2012-04-10 10:50:36 <sipa> none
106 2012-04-10 10:51:12 <freewil> empty string?
107 2012-04-10 10:59:48 <t7> has anyone implemented bitcoin-qt update notifications?
108 2012-04-10 11:00:19 <gmaxwell> No, and it would be a bad idea to do that until we have secure updates worked out.
109 2012-04-10 11:00:30 <t7> just check sf.net?
110 2012-04-10 11:00:43 <t7> https
111 2012-04-10 11:00:53 <gmaxwell> We don't want rapid deployment of new versions when anyone who can hack sourceforge or github could push a malicious update that most users would run.
112 2012-04-10 11:01:22 <t7> ah yeah i see
113 2012-04-10 11:04:24 <freewil> gmaxwell, isnt something being worked on with gitian
114 2012-04-10 11:04:41 <freewil> maybe something where at least 3 diff developers have to have a release signed
115 2012-04-10 11:05:33 <t7> what is a share in the context of pooled mining?
116 2012-04-10 11:05:46 <luke-jr> t7: a "block" at a lower difficulty
117 2012-04-10 11:05:56 <t7> ah
118 2012-04-10 11:05:59 <gmaxwell> freewil: Yes.
119 2012-04-10 11:06:13 <t7> so n - m zeros instead of n zeros
120 2012-04-10 11:06:38 <t7> do the pools decide what difficulty a share is?
121 2012-04-10 11:07:26 <luke-jr> yes
122 2012-04-10 11:21:23 <t7> p2p
123 2012-04-10 11:21:26 <t7> p2pool
124 2012-04-10 11:21:28 <t7> awesome
125 2012-04-10 11:24:30 <luke-jr> t7: p2pool is an awesome substitute for solo mining, but not for other pools :P
126 2012-04-10 11:25:56 <gmaxwell> (says the pool operator, :) )
127 2012-04-10 11:26:15 <luke-jr> (as if that matters)
128 2012-04-10 11:26:15 <t7> what happens if i submit a block without the 50 bitcoin tx ?
129 2012-04-10 11:26:27 <luke-jr> t7: you destroy it.
130 2012-04-10 11:26:30 <t7> will it still be accepted?
131 2012-04-10 11:26:40 <luke-jr> gmaxwell: also, that should be "says the *competing* pool operator" :P
132 2012-04-10 11:26:49 <luke-jr> otherwise you imply p2pool doesn't have one
133 2012-04-10 11:27:01 <luke-jr> t7: so long as there is a generation txn
134 2012-04-10 11:28:45 <gmaxwell> t7: the transaction has to be there.. but it can generate <50 btc.
135 2012-04-10 11:28:49 <gmaxwell> (incluing zero.
136 2012-04-10 11:28:50 <gmaxwell> )
137 2012-04-10 11:31:09 <luke-jr> or really, just verify blocks based on total outputs <= total inputs
138 2012-04-10 11:39:30 <sipa> cool, bitcoin just compiles when enabling c++11x extensions
139 2012-04-10 11:39:40 <luke-jr> sipa: I would expect so :p
140 2012-04-10 11:40:02 <sipa> it should even be a bit faster, potentially, without any code changes
141 2012-04-10 11:40:15 <sipa> but i can't measure that
142 2012-04-10 11:40:23 <Joric> very modern
143 2012-04-10 11:40:29 <sipa> ?
144 2012-04-10 11:40:42 <luke-jr> why faster?
145 2012-04-10 11:41:06 <sipa> rvalue references & move semantics
146 2012-04-10 11:42:01 <sipa> which are implemented for the STL library container when compiling with those extensions
147 2012-04-10 11:45:20 <luke-jr> 12fGdRstfff4KEky64kgBTauVU2Ertv4Lt is the new deepspam addr?
148 2012-04-10 11:45:42 <luke-jr> hmm no, looks like just another spammer
149 2012-04-10 11:47:13 <MasterChief> technical question
150 2012-04-10 11:47:22 <MasterChief> exactly how shitted up with crap is the blockchain?
151 2012-04-10 11:47:45 <sipa> define crap
152 2012-04-10 11:49:19 <MasterChief> shit that doesnt relate to transactions
153 2012-04-10 11:49:33 <MasterChief> useful transactions
154 2012-04-10 11:49:43 <luke-jr> sipa: man, you really broke everything with this lock change >_<
155 2012-04-10 11:50:00 <luke-jr> MasterChief: so Deepspam included?
156 2012-04-10 11:50:00 <sipa> luke-jr: yes
157 2012-04-10 11:50:10 <MasterChief> i dont know what that is
158 2012-04-10 11:50:18 <[Tycho]> Cool, we have a new flooder :)
159 2012-04-10 11:50:20 <luke-jr> MasterChief: Deepbit payouts
160 2012-04-10 11:50:38 <MasterChief> whats bad about that
161 2012-04-10 11:51:01 <sipa> not sure right now, but it was like 1/6th of the blockchain some time ago :)
162 2012-04-10 11:51:30 <[Tycho]> Not 1/6 of the blockchain, for sure.
163 2012-04-10 11:51:48 <sipa> maybe not anymore
164 2012-04-10 11:52:08 <[Tycho]> Someone counted the number of TXes per day, not the size.
165 2012-04-10 11:52:46 <luke-jr> MasterChief: it's 25% of all transactions because [Tycho] does it wrong
166 2012-04-10 11:53:04 <MasterChief> wtf
167 2012-04-10 11:53:12 <MasterChief> fix your shit [Tycho]
168 2012-04-10 11:53:16 <MasterChief> inb4 no
169 2012-04-10 11:53:32 <[Tycho]> I'm doing it in a classic way.
170 2012-04-10 11:54:22 <sipa> luke-jr: how did you get that build.h not found error
171 2012-04-10 11:54:55 <luke-jr> sipa: I ran 'make' <.<
172 2012-04-10 11:55:27 <sipa> where, which commit, source tree, ...?
173 2012-04-10 11:55:39 <sipa> it should run share/genbuild.sh src/build.h
174 2012-04-10 11:55:46 <sipa> to create that file
175 2012-04-10 11:56:06 <luke-jr> master + merge from branch
176 2012-04-10 11:56:28 <luke-jr> I'm still going through all the next-test pullreqs to identify ones that break, gimme a few&
177 2012-04-10 11:57:13 <sipa> make for bitcoin-qt, or in the src/ subdir?
178 2012-04-10 11:57:36 <luke-jr> not sure which one, sorry.
179 2012-04-10 11:57:48 <luke-jr> almost done with pullreq testing tho
180 2012-04-10 11:57:58 <luke-jr> (my GNUmakefile runs both in parallel)
181 2012-04-10 11:58:10 <sipa> if you have an own makefile, you'll need to change that yes
182 2012-04-10 11:58:26 <luke-jr> my own makefile *just* chains to the in-tree ones
183 2012-04-10 11:58:30 <sipa> hmm
184 2012-04-10 11:58:41 <sipa> that shouldn't be a problem
185 2012-04-10 12:00:30 <DrHaribo> What happened to the idea of requiring the block height number in the coinbase? Will that happen?
186 2012-04-10 12:00:48 <sipa> i hope so, but it's not very urgent now
187 2012-04-10 12:01:43 <[Tycho]> What's the purpose of this ?
188 2012-04-10 12:01:56 <luke-jr> to break Deepbit.
189 2012-04-10 12:02:22 <gmaxwell> [Tycho]: to make it effectively impossible to produce a duplicate coinbase transaction. (which when spent can produce duplicate regular transactions)
190 2012-04-10 12:03:49 <luke-jr> sipa: looks like makefile.unix
191 2012-04-10 12:04:09 <luke-jr> hmm
192 2012-04-10 12:04:11 <sipa> luke-jr: bitcoind depends on build.h, which is FORCED, and should recreate build.h
193 2012-04-10 12:04:12 <luke-jr> not getting it now :/
194 2012-04-10 12:04:56 <sipa> sounds like some concurrency problem, maybe with the qt build also using that file
195 2012-04-10 12:05:06 <luke-jr> oh
196 2012-04-10 12:05:09 <sipa> anyway, if you can reproduce it, let me know
197 2012-04-10 12:05:12 <luke-jr> maybe -j4
198 2012-04-10 12:05:19 <sipa> nah, i run -j5 as well
199 2012-04-10 12:05:24 <luke-jr> yep
200 2012-04-10 12:05:43 <sipa> do you do a make clean sometime?
201 2012-04-10 12:05:45 <luke-jr> sipa: rm your old build.h first
202 2012-04-10 12:05:51 <sipa> why?
203 2012-04-10 12:05:59 <luke-jr> so the stale file isn't there
204 2012-04-10 12:06:08 <sipa> doesn't matter if it's there
205 2012-04-10 12:06:45 <luke-jr> in fact
206 2012-04-10 12:06:46 <freewil> can someone make a new repo for the testnet-in-a-box under github.com/bitcoin
207 2012-04-10 12:06:57 <luke-jr> Makefile (bitcoin-qt) doesn't run without a preexisting build.h
208 2012-04-10 12:07:08 <sipa> ah, yes, that is true
209 2012-04-10 12:07:39 <sipa> it should put the command for creating build.h in the generated makefile, instead of creating it immediately
210 2012-04-10 12:08:33 <sipa> any advise for doing that?
211 2012-04-10 12:10:57 <sipa> wumpus: ^ ?
212 2012-04-10 12:11:54 <luke-jr> http://colby.id.au/node/145
213 2012-04-10 12:13:29 <freewil> sipa, do you think we could get a new repo up on github for the testnet-in-a-box: http://sourceforge.net/projects/bitcoin/files/Bitcoin/testnet-in-a-box/
214 2012-04-10 12:15:11 <freewil> gmaxwell?
215 2012-04-10 12:16:54 <sipa> it's mostly binary files and config
216 2012-04-10 12:18:06 <sipa> luke-jr: ok that should do it; i'm first trying to get that -dirty flag off in the windows build
217 2012-04-10 12:18:26 <t7> sipa is it easyto cross compile for windows?
218 2012-04-10 12:18:51 <luke-jr> t7: only with Ubuntu
219 2012-04-10 12:18:53 <freewil> yeah, just saying it could be updated - new wallet, new inverted config settings (noirc -> irc)
220 2012-04-10 12:19:04 <t7> luke-jr: why?
221 2012-04-10 12:19:19 <luke-jr> t7: that's the only platform gitian supports
222 2012-04-10 12:19:57 <sipa> t7: we use a system called gitian; that uses a virtual machine image with an Ubuntu install in it, and runs some verifyable scripts to produce the binaries from a git checkout in there
223 2012-04-10 12:20:20 <sipa> t7: that way, each developer can get exactly the same binary and compare the results (and sign them)
224 2012-04-10 12:20:49 <sipa> you can do cross-compiles without gitian too, but i don't bother anymore :)
225 2012-04-10 12:21:31 <t7> to_internal not found in DebugShrinkFile :O
226 2012-04-10 12:21:40 <sipa> ?
227 2012-04-10 12:25:19 <zooko> https://intersango.com  is giving 503
228 2012-04-10 12:26:40 <t7> luke-jr: have you still got that moustache?
229 2012-04-10 12:27:00 <luke-jr> &
230 2012-04-10 12:27:08 <t7> on github :)
231 2012-04-10 12:28:17 <t7> bitcoind wont build for me now
232 2012-04-10 12:28:22 <t7> i think im missing something
233 2012-04-10 12:28:24 <sipa> what error?
234 2012-04-10 12:29:03 <t7> seems to be in boost :|
235 2012-04-10 12:29:18 <sipa> can you just paste the error somewhere?
236 2012-04-10 12:30:21 <t7> i cant pip the output from make
237 2012-04-10 12:30:26 <t7> it just writes to console
238 2012-04-10 12:30:44 <sipa> ever heard about copy-paste?
239 2012-04-10 12:31:02 <t7> from ssh?
240 2012-04-10 12:31:10 <sipa> from your console
241 2012-04-10 12:31:21 <t7> ah putty can do it
242 2012-04-10 12:32:08 <t7> http://hpaste.org/66761
243 2012-04-10 12:33:40 <sipa> what version of boost?
244 2012-04-10 12:33:53 <t7> arch latest i think
245 2012-04-10 12:34:59 <luke-jr> which is?
246 2012-04-10 12:35:26 <t7> 1.49.0-1.1
247 2012-04-10 12:35:34 <sipa> should be fine...
248 2012-04-10 12:36:08 <alephozee> phantomcircuit: intersango is down?
249 2012-04-10 12:36:28 <t7> it built a couple of weeks ago
250 2012-04-10 12:38:11 <t7> god damn c++
251 2012-04-10 12:38:38 <luke-jr> sipa: just checking: did you intentionally remove the db.cpp changes from loadblock?
252 2012-04-10 12:39:14 <t7> how do i checkout a past head ?
253 2012-04-10 12:39:16 <t7> i forget
254 2012-04-10 12:39:18 <sipa> luke-jr: which ones?
255 2012-04-10 12:39:30 <luke-jr> --    dbenv.txn_checkpoint(0, nMinutes, 0);
256 2012-04-10 12:39:31 <luke-jr> -+    if (!fStarting)
257 2012-04-10 12:39:33 <luke-jr> -+        dbenv.txn_checkpoint(0, nMinutes, 0);
258 2012-04-10 12:40:14 <sipa> fStarting is gone, right?
259 2012-04-10 12:40:34 <sipa> if that's your question: yes that patch you pasted above is intentional
260 2012-04-10 12:40:41 <luke-jr> k
261 2012-04-10 12:41:23 <sipa> it was an attempt to not do any db syncing stuff while loading blocks, hoping that that would speed things up, but the rest was a huge log file
262 2012-04-10 12:41:27 <sipa> files
263 2012-04-10 12:41:35 <sipa> and it didn't help speed things up anyway
264 2012-04-10 12:41:52 <t7> inb4 sqlite
265 2012-04-10 12:42:06 <sipa> inb4?
266 2012-04-10 12:42:13 <t7> can you commit parts of the db to disk with sqlite
267 2012-04-10 12:43:01 <sipa> we're not using sqlite
268 2012-04-10 12:43:18 <t7> yet
269 2012-04-10 12:43:48 <luke-jr> t7: we're talking about speeding things up, not slowing them down.
270 2012-04-10 12:43:52 <luke-jr> (sqlite isn't known for speed)
271 2012-04-10 12:43:57 <freewil> redis
272 2012-04-10 12:44:08 <freewil> redis the entire blockchain
273 2012-04-10 12:44:08 <sipa> t7: i have not idea what you're asking, and i've no idea why you think about sqlite
274 2012-04-10 12:45:46 <freewil> (all in memory)
275 2012-04-10 12:45:48 <gmaxwell> I wonder if these people existed in the early 1900s. ... blacksmith working on a horseshoe, random guy walks up... "Hey, you're still using a hammer? Powerhammers have BarnSCALE."
276 2012-04-10 12:46:27 <freewil> lol gmaxwell have you seen this... http://www.xtranormal.com/watch/6995033/mongo-db-is-web-scale
277 2012-04-10 12:46:39 <luke-jr> freewil: redis doesn't seem to support persistent storage&?
278 2012-04-10 12:46:53 <sipa> t7: i also have no idea what 'inb4' means
279 2012-04-10 12:47:12 <t7> if i did git cheakout HEAD~100
280 2012-04-10 12:47:19 <luke-jr> freewil: also, you're insane if you think we should store the entire blockchain in memory.
281 2012-04-10 12:47:22 <freewil> luke-jr, i was kidding, but it is meant to hold all the data in memory, but it does do disk writes for persistence
282 2012-04-10 12:47:25 <t7> does git checkout HEAD get me back to the most recent commit?
283 2012-04-10 12:47:31 <t7> i dont think it has
284 2012-04-10 12:47:38 <luke-jr> t7: no
285 2012-04-10 12:47:43 <luke-jr> t7: HEAD is what you have checked out
286 2012-04-10 12:47:49 <luke-jr> t7: you want master
287 2012-04-10 12:48:07 <t7> am i missing a dependency or something
288 2012-04-10 12:48:13 <helo> yes
289 2012-04-10 12:48:27 <t7> apart from a brain
290 2012-04-10 12:49:40 <helo> unsure
291 2012-04-10 12:50:13 <luke-jr> XD
292 2012-04-10 12:52:01 <sipa> explain me this: in a clean git directory, do "git describe --dirty", it will not contain the string "-dirty". touch a checked-in file, do "git describe --dirty", it will contains the string "-dirty". Do git diff. Do git describe --dirty: dirty flag gone again
293 2012-04-10 12:55:55 <gmaxwell> sipa: 1.7.7.6 doesn't do that for me.
294 2012-04-10 12:56:14 <sipa> 1.7.5.4 does
295 2012-04-10 12:56:43 <gmaxwell> http://pastebin.mozilla.org/1562347
296 2012-04-10 12:57:57 <sipa> gmaxwell: your directory was dirty to begin with
297 2012-04-10 12:58:28 <t7> guys it wasnt your fault
298 2012-04-10 12:58:30 <t7> it was arch
299 2012-04-10 12:58:39 <t7> had some workaround #ifdef for something
300 2012-04-10 12:58:55 <t7> probably gonna break loads of other things now but atleast bitcoin builds
301 2012-04-10 12:59:11 <t7> (in boost)
302 2012-04-10 13:00:20 <luke-jr> sipa: possibly fixed by bb571486ae93d02746c4bcc8032bde306f6d399a
303 2012-04-10 13:00:50 <luke-jr> so fixed in git 1.7.7
304 2012-04-10 13:01:11 <gmaxwell> sipa: http://pastebin.mozilla.org/1562364
305 2012-04-10 13:02:10 <luke-jr> describe: Refresh the index when run with --dirty
306 2012-04-10 13:02:11 <luke-jr> When running git describe --dirty the index should be refreshed.  Previously the cached index would cause describe to think that the index was dirty when, in reality, it was just stale.
307 2012-04-10 13:03:57 <luke-jr> "Precise Pangolin" is the code name for Ubuntu 12.04, scheduled for release on 26 April 2012. <-- minimum Ubuntu to update gitian to, for this fix
308 2012-04-10 13:04:15 <luke-jr> or we can just have our gitian script run 'git diff' first?
309 2012-04-10 13:06:59 <sipa> luke-jr: yes, i'm just doing a git diff in advance
310 2012-04-10 13:12:51 <sipa> gmaxwell: with 'touch' i meant really just the unix command touch, not actually modify contents
311 2012-04-10 13:13:18 <sipa> anyway, seems 1.7.7 fixes it
312 2012-04-10 13:13:32 <gmaxwell> sipa: touch didn't make it dirty at all for me.
313 2012-04-10 13:13:48 <sipa> that's expected on 1.7.7 :)
314 2012-04-10 13:18:18 <t7> o god main.cpp is 3k LOC
315 2012-04-10 13:19:28 <Joric_> *stunning silence*
316 2012-04-10 13:21:14 <t7> could be split into messages.cpp and block.cpp just from first scan
317 2012-04-10 13:21:22 <t7> (or something)
318 2012-04-10 13:23:08 <sipa> t7: take a look at 0.3.19 or so
319 2012-04-10 13:23:18 <sipa> net.cpp and main.cpp were a lot bigger even
320 2012-04-10 13:23:19 <t7> the devs prefer monolithic files?
321 2012-04-10 13:23:31 <sipa> no, but we don't like changing too much at once
322 2012-04-10 13:27:47 <luke-jr> t7: BlueMatt and sipa have been working on splitting out a blockstore.cpp
323 2012-04-10 13:28:41 <t7> You should start on a haskell port before you spend too much effort on the cpp code :)
324 2012-04-10 13:29:00 <t7> i duno how to handle more data than ram
325 2012-04-10 13:29:11 <t7> in haskell
326 2012-04-10 13:29:28 <sipa> t7: and risk that it is just *slightly* incompatible with the origin code, in a very hidden way?
327 2012-04-10 13:30:16 <sipa> and suddenly everyone who has upgraded ends up in a forked block chain
328 2012-04-10 13:30:22 <sipa> (by the way: i'm a big fan of Haskell)
329 2012-04-10 13:30:23 <Joric> seems nobody wrote full-blown haskell bitcoin node yet - rpc interface, tops
330 2012-04-10 13:30:41 <sipa> Joric: yes, roconnor has a fully verifying node in haskell
331 2012-04-10 13:30:44 <sipa> purecoin
332 2012-04-10 13:31:59 <luke-jr> plus, who has GHC?
333 2012-04-10 13:32:08 <sipa> any distro?
334 2012-04-10 13:32:37 <t7> well each distro seems to have a different version
335 2012-04-10 13:32:43 <luke-jr> I mean installed.
336 2012-04-10 13:32:43 <t7> 6 7 or 7.4 on arch
337 2012-04-10 13:32:48 <t7> causes big issues for me
338 2012-04-10 13:33:12 <sipa> luke-jr: is that relevant?
339 2012-04-10 13:33:22 <sipa> our users don't have qt creator installed either
340 2012-04-10 13:34:24 <t7> sipa how would you handle a structure too big to fit in ram? use IO refs and chunks or something? would be messy :( but i guess it is a pain in cpp too
341 2012-04-10 13:34:44 <sipa> t7: there are database libraries in Haskell, no? ;)
342 2012-04-10 13:35:15 <t7> yeah but not everyone wants to run postgre
343 2012-04-10 13:35:35 <sipa> http://hackage.haskell.org/package/BerkeleyDB-0.8.5
344 2012-04-10 13:35:57 <copumpkin> zomg
345 2012-04-10 13:36:04 <copumpkin> 'tis haskell!
346 2012-04-10 13:40:55 <sipa> luke-jr: buildinfo should be fixed now
347 2012-04-10 13:41:33 <luke-jr> sipa: we don't need Qt Creator to build
348 2012-04-10 13:41:55 <vvfff> i transfered some money from my silkroad account to bmr the other day and it went missing. it got sent to another address. wtf happend?
349 2012-04-10 13:42:21 <luke-jr> vvfff: wrong channel
350 2012-04-10 13:43:12 <MasterChief> http://axgle.github.com/images/haskell.jpg
351 2012-04-10 13:43:26 <sipa> luke-jr: is your argument really "building such a program requires a tool that not everyone has" /
352 2012-04-10 13:43:51 <luke-jr> sipa: well, there's not many *good* arguments to avoid Haskell :P
353 2012-04-10 13:43:57 <sipa> haha
354 2012-04-10 13:49:52 <t7> can the client recognise when it has an orphan node etc?
355 2012-04-10 13:51:08 <luke-jr> what's an orphan node?
356 2012-04-10 13:51:32 <t7> block i man
357 2012-04-10 13:51:34 <t7> mean*
358 2012-04-10 13:51:43 <graingert> t7: do you mean an orphan block?
359 2012-04-10 13:51:48 <luke-jr> sipa: src/net.cpp:1709:44: error: no matching function for call to CNetAddr::CNetAddr(in6_addr&)
360 2012-04-10 13:51:55 <luke-jr> t7: if they could, it wouldn't be an orphan block
361 2012-04-10 13:52:05 <graingert> t7: it accepts the non orphan block
362 2012-04-10 13:52:38 <luke-jr> sipa: only for Bitcoin-Qt somehow
363 2012-04-10 13:52:49 <t7> luke-jr: yeah but after a while it would get x messages confirming an alternate chain, right?
364 2012-04-10 13:53:00 <luke-jr> t7: then it reorgs
365 2012-04-10 13:53:08 <t7> i havnt even looked into the protocol
366 2012-04-10 13:53:13 <t7> i should read up
367 2012-04-10 13:54:25 <luke-jr> sipa: why is USE_IPV6 disabled by default for Bitcoin-Qt? :/
368 2012-04-10 14:01:49 <luke-jr> sipa: http://paste.pocoo.org/show/579057/
369 2012-04-10 14:29:12 <luke-jr> sipa: haha, you're a step ahead of me in bugfinding! :P
370 2012-04-10 14:32:38 <sipa> ?
371 2012-04-10 14:36:18 <luke-jr> sipa: the commented out copy_file in db
372 2012-04-10 14:36:27 <sipa> ha, indeed :)
373 2012-04-10 14:40:57 <t7> are you two professional software devs?
374 2012-04-10 14:48:53 <luke-jr> sipa:             const char* pszFile = GetDebugLogName().c_str();
375 2012-04-10 14:49:09 <luke-jr> sipa: is that valid? Won't the temporary from GetDebugLogName() free the C const data?
376 2012-04-10 14:54:43 <sipa> luke-jr: where do you see that?
377 2012-04-10 14:56:56 <luke-jr> #917, sorry I assumed it was your changes
378 2012-04-10 14:57:00 <luke-jr> will post on that issue
379 2012-04-10 14:58:14 <sipa> you do make me wonder
380 2012-04-10 14:58:52 <luke-jr> well, you were touching paths and stuff most recently :p
381 2012-04-10 14:59:02 <sipa> the file() function calls .string().c_str(), and returns that, but the result of string() may be a temporary
382 2012-04-10 15:00:48 <luke-jr> :o
383 2012-04-10 15:03:31 <sipa> it's only a temporary on windows though
384 2012-04-10 15:03:37 <luke-jr> ?
385 2012-04-10 15:03:38 <sipa> on other systems it returns a const reference
386 2012-04-10 15:04:59 <luke-jr> hmm
387 2012-04-10 15:05:05 <luke-jr> no guarantee of that I presume
388 2012-04-10 15:05:15 <sipa> no, i just checked the source
389 2012-04-10 15:06:04 <sipa> can we throw out boost::filesystem v2?
390 2012-04-10 15:06:06 <luke-jr> sipa: shall I write a fix, or are you started on it already?
391 2012-04-10 15:06:11 <luke-jr> sipa: I wouldn't yet.
392 2012-04-10 15:06:23 <sipa> luke-jr: how would you fix that?
393 2012-04-10 15:07:03 <sipa> change every usage of file() and dir() into a #if BOOST_FILESYSTEM_VERSION == 3 ... #else ... #endif?
394 2012-04-10 15:07:18 <sipa> with .native_string().c_str() for v2 and .string().c_str() for v3?
395 2012-04-10 15:07:57 <luke-jr> no, I'd not assume .string() sticks around
396 2012-04-10 15:08:07 <luke-jr> put it in the stack properly
397 2012-04-10 15:08:32 <sipa> yes it does
398 2012-04-10 15:08:38 <sipa> but string() doesn't work in v2
399 2012-04-10 15:08:49 <luke-jr> it's not guaranteed to.
400 2012-04-10 15:09:02 <sipa> no, it's the wrong type
401 2012-04-10 15:09:18 <luke-jr> ?
402 2012-04-10 15:09:46 <sipa> it returns something of type boost::filesystem::path::string_type
403 2012-04-10 15:09:54 <sipa> which on windows is a string of wchar_t
404 2012-04-10 15:10:25 <sipa> so .c_str() applied to that will not get you a const char*
405 2012-04-10 15:10:43 <sipa> while in v3, .string() always returns std::string
406 2012-04-10 15:11:00 <luke-jr> &
407 2012-04-10 15:12:11 <luke-jr> sipa: so I'm fixing it? :p
408 2012-04-10 15:12:18 <sipa> luke-jr: i don't see how you would fix it
409 2012-04-10 15:12:29 <sipa> without excessive macro lines
410 2012-04-10 15:12:30 <luke-jr> I'll show you the diff in a bit, then :p
411 2012-04-10 15:12:44 <sipa> ah, here's an idea:
412 2012-04-10 15:12:48 <luke-jr> anyhow, why should IPv6 be un-built by default?
413 2012-04-10 15:13:09 <sipa> #define path_string string
414 2012-04-10 15:13:11 <sipa> (on v3)
415 2012-04-10 15:13:18 <sipa> and #define path_string native_string
416 2012-04-10 15:13:21 <sipa> (on v2)
417 2012-04-10 15:13:28 <sipa> and then use .path_string().c_str() everywhere
418 2012-04-10 15:13:46 <luke-jr> that has the same bug O.o
419 2012-04-10 15:13:52 <sipa> no it doesn't
420 2012-04-10 15:14:27 <sipa> path_string may return a temporary, but it will only be freed after the current statement is complete
421 2012-04-10 15:14:39 <luke-jr> that's not what the docs say :p
422 2012-04-10 15:14:45 <sipa> ?
423 2012-04-10 15:15:03 <luke-jr> at least IBM's C++ docs say temporaries are scoped lexically inside expressions
424 2012-04-10 15:15:12 <luke-jr> so when c_str() returns, the std::string is destroyed
425 2012-04-10 15:15:14 <luke-jr> before printf
426 2012-04-10 15:15:24 <sipa> hmm really
427 2012-04-10 15:15:44 <sipa> in that case, .string().c_str() can't work anywhere
428 2012-04-10 15:16:33 <sipa> http://stackoverflow.com/questions/3041249/when-are-temporaries-created-as-part-of-a-function-call-destroyed
429 2012-04-10 15:16:36 <sipa> that sounds good
430 2012-04-10 15:16:47 <kingkatari> Hey Luke
431 2012-04-10 15:18:11 <luke-jr> sipa: I'd want to see what "full-expression (1.9)" is defined as
432 2012-04-10 15:18:38 <kingkatari> Well I am Up and Running now
433 2012-04-10 15:18:59 <luke-jr> sipa: also, we're not passing .string() as an argument, we're passing a const char* that c_str() returns
434 2012-04-10 15:19:37 <sipa> doesn't matter, the return value of string() is a temporary that will not be destroyed until the entire expression is evaluated
435 2012-04-10 15:21:02 <luke-jr> no, until the full-expression where it is
436 2012-04-10 15:21:08 <luke-jr> in this case, "foo.string().c_str()"
437 2012-04-10 15:21:58 <sipa> ?
438 2012-04-10 15:22:16 <sipa> the entire expression, not just that part
439 2012-04-10 15:22:25 <luke-jr> that is the full expression involving foo.string()'s temporary
440 2012-04-10 15:22:56 <sipa> oh, you're still talking about that problem in #917?
441 2012-04-10 15:23:06 <luke-jr> .string().c_str() is always a bug, I mean
442 2012-04-10 15:23:30 <sipa> printf("bla: %s\n", path.string().c_str()); is fine
443 2012-04-10 15:23:41 <sipa> the string will only be destroyed after the evaluation of printf
444 2012-04-10 15:23:51 <luke-jr> that is not guaranteed.
445 2012-04-10 15:23:55 <sipa> yes it is
446 2012-04-10 15:24:02 <sipa> the expression is printf("bla: %s\n", path.string().c_str())
447 2012-04-10 15:24:15 <luke-jr> path.string().c_str() is also an expression
448 2012-04-10 15:24:21 <sipa> not the full expression
449 2012-04-10 15:25:48 <luke-jr> in this case, yes
450 2012-04-10 15:26:46 <luke-jr> you're confusing full expression with full statement
451 2012-04-10 15:27:02 <sipa> there's no such thing as a full statement
452 2012-04-10 15:27:22 <sipa> a statement is either a full expression, or a return statement, or a break, or a continue, or a few other thing
453 2012-04-10 15:27:28 <luke-jr> "When a temporary object is created during the evaluation of a full-expression (an expression that is not a subexpression of another expression), it is destroyed as the last step in its evaluation that lexically contains the point where it was created."
454 2012-04-10 15:27:38 <sipa> exactly
455 2012-04-10 15:28:47 <luke-jr> the printf is lexically outside the point where it was created.
456 2012-04-10 15:29:32 <sipa> sigh
457 2012-04-10 15:29:42 <sipa> the printf is part of the full expresion
458 2012-04-10 15:30:13 <sipa> nothing is destroyed before the entire full expression is evaluated
459 2012-04-10 15:31:30 <luke-jr> it's unclear with just StackOverflow and IBM. too bad C++03 isn't public
460 2012-04-10 15:32:02 <sipa> well path.string().c_str() is unambiguously a subexpression of printf("bla: %s\n", path.string().c_str())
461 2012-04-10 15:32:37 <luke-jr> Microsoft says at the end of the "expression statement"
462 2012-04-10 15:33:02 <luke-jr> you're probably right.
463 2012-04-10 15:34:04 <jgarzik> sigh.  or put more simply, the compiler knows when the temporaries are needed, and not, and will Do The Right Thing.
464 2012-04-10 15:34:44 <sipa> jgarzik: yes, but const char *pszStr = path.string().c_str(); is wrong
465 2012-04-10 15:34:50 <luke-jr> jgarzik: well, the concern I had was that the temporary *isn't* needed at the time printf is called, just some data it created and returned a pointer to
466 2012-04-10 15:35:12 <luke-jr> (and frees when it's destroyed)
467 2012-04-10 15:35:18 <jgarzik> sipa: sure, because you're storing the temporary in a permanent pointer variable
468 2012-04-10 15:35:29 <sipa> exactly, and i believe we may be doing that
469 2012-04-10 15:35:30 <jgarzik> the temp still goes away as expected
470 2012-04-10 15:36:33 <luke-jr> sipa: it's not merged yet
471 2012-04-10 15:37:06 <luke-jr> sipa: anyhow, back to IPv6& why not built by default?
472 2012-04-10 15:37:38 <sipa> i considered it experimental when i wrote it
473 2012-04-10 15:37:50 <sipa> but i guess we can enable it by default, as it should work on all platforms
474 2012-04-10 15:41:12 <Jezzz> "Assertion failed!  Expression: pindexNew->pprev == pindexBest"
475 2012-04-10 15:41:20 <Jezzz> ^ my client says that on run and force quits
476 2012-04-10 15:41:36 <luke-jr> Jezzz: version?
477 2012-04-10 15:41:40 <Jezzz> "File src/main.cpp | Line: 1462"
478 2012-04-10 15:41:52 <Jezzz> hmm
479 2012-04-10 15:41:54 <luke-jr> &
480 2012-04-10 15:42:18 <jgarzik> might need another set of DNS seeds
481 2012-04-10 15:42:40 <sipa> Jezzz: don't run 0.6.0rc4 please
482 2012-04-10 15:42:43 <Jezzz> 0.6.0 Beta
483 2012-04-10 15:42:59 <BlueMatt> jgarzik: dnsseed.bluematt.me.
484 2012-04-10 15:43:05 <jgarzik> BlueMatt: nice
485 2012-04-10 15:43:07 <BlueMatt> ok, its static, but atleast its there
486 2012-04-10 15:43:22 <BlueMatt> sipa: you do still have a seed at that address, right?
487 2012-04-10 15:43:28 <sipa> BlueMatt: no...
488 2012-04-10 15:43:33 <BlueMatt> oh...
489 2012-04-10 15:43:44 <BlueMatt> well, shit
490 2012-04-10 15:43:57 <jgarzik> heh
491 2012-04-10 15:43:57 <sipa> i will be soon
492 2012-04-10 15:44:07 <sipa> wait
493 2012-04-10 15:44:09 <sipa> no i do
494 2012-04-10 15:44:15 <sipa> but it's not listening on ipv6
495 2012-04-10 15:44:27 <Jezzz> luke-jr:  0.6.0 Beta
496 2012-04-10 15:44:39 <sipa> Jezzz: you're running rc4
497 2012-04-10 15:44:41 <sipa> upgrade
498 2012-04-10 15:44:47 <Jezzz> thx
499 2012-04-10 15:45:47 <jgarzik> sipa: any major IPv6 issues still to iron out, besides DNS seeds?
500 2012-04-10 15:46:35 <sipa> jgarzik: maybe a throttling of how many non-ipv4 addresses non-ipv6 nodes relay/store
501 2012-04-10 15:47:04 <sipa> right now, they'll just store and forward every routable address, but only connect to IPv4 ones
502 2012-04-10 15:47:14 <luke-jr> probably shouldn't store addresses you don't support?
503 2012-04-10 15:47:27 <sipa> that's a possibility
504 2012-04-10 15:47:33 <luke-jr> if the IPv4 are all evicted by IPv6, it leaves you bootstrappign again
505 2012-04-10 15:48:00 <sipa> that would be extremely hard
506 2012-04-10 15:48:06 <luke-jr> also, need to keep in mind it's not IPv4 vs IPv6 here, but IPv4 vs IPv6 vs Tor vs I2P
507 2012-04-10 15:48:24 <sipa> yes, it's broader than just ipv6
508 2012-04-10 15:48:25 <luke-jr> just because I support IPv4 and IPv6, doesn't mean I should relay every Tor/I2P address
509 2012-04-10 15:48:37 <jgarzik> sipa: I admit, I was considering an "only v6 nodes relay v6 addresses" rule
510 2012-04-10 15:48:43 <jgarzik> and ditto for v4
511 2012-04-10 15:48:52 <luke-jr> jgarzik: sounds like a receipe for a split IMO
512 2012-04-10 15:49:01 <sipa> that means you'd end up with completely separate networks basically
513 2012-04-10 15:49:11 <jgarzik> in practice, highly doubtful
514 2012-04-10 15:49:17 <sipa> i'd prefer at least some relaying of addresses in network you don't support
515 2012-04-10 15:49:20 <jgarzik> huge incentives to interconnect exist
516 2012-04-10 15:49:25 <luke-jr> in practice, everyone should have IPv6 by now
517 2012-04-10 15:49:37 <sipa> as luke said, it's not just ipv6
518 2012-04-10 15:49:49 <jgarzik> s/practice/theory/.  practice means its actually happening in the real world, not just your imagination.
519 2012-04-10 15:49:56 <luke-jr> jgarzik: yes, I meant in the real world
520 2012-04-10 15:50:05 <sipa> irrelevant
521 2012-04-10 15:50:13 <luke-jr> Linux users have no excuse to build without IPv6 support, and Windows sets up a Teredo tunnel by default
522 2012-04-10 15:50:55 <jgarzik> sipa: is the pull request up-to-date?
523 2012-04-10 15:51:05 <sipa> jgarzik: yes
524 2012-04-10 15:51:22 <sipa> i'd prefer at least some relaying of addresses in network you don't support; maybe throttled, maybe not storing them in their own addr.dat, maybe with reduced probability, but just to ease bootstrapping, it would be nice to have some addresses from all networks floating around
525 2012-04-10 15:52:35 <jgarzik> sipa: maybe addrman already addressed these issues
526 2012-04-10 15:52:54 <sipa> relaying is outside of addrman's scope
527 2012-04-10 15:53:02 <jgarzik> sipa: mainly we don't want flooding of the entire v6 internet to v4 nodes, who have no hope of validating those addresses
528 2012-04-10 15:53:31 <jgarzik> sipa: it's an easy DoS to flood non-working addresses, without some logic like addrman in place
529 2012-04-10 15:53:32 <sipa> what about the policy: relay, but don't store things from networks you don't support
530 2012-04-10 15:53:53 <jgarzik> sipa: I wouldn't even relay everything.
531 2012-04-10 15:54:05 <sipa> relaying is already probabilistic, afaik
532 2012-04-10 15:54:48 <luke-jr> jgarzik: how about relaying at most 1 foreign address per hour?
533 2012-04-10 15:55:04 <luke-jr> or 1 IPv6, 1 Tor, 1 I2P
534 2012-04-10 15:55:52 <sipa> i don't see how even 1 per minute could harm
535 2012-04-10 15:56:10 <Jezzz> upgrade fixed me.  thx gents.
536 2012-04-10 15:56:47 <luke-jr> 1 per foreign network per 5 mins? :P
537 2012-04-10 15:56:48 <luke-jr> hmm
538 2012-04-10 15:57:04 <luke-jr> rules like this might cause the *same* addresses relayed across IPv4-only networks
539 2012-04-10 15:57:28 <sipa> an address is never relayed twice to the same node
540 2012-04-10 16:33:34 <sipa> luke-jr, jgarzik: updated
541 2012-04-10 16:33:45 <jgarzik> sipa: thanks.  was just in the middle of reviewing v6.
542 2012-04-10 16:33:57 <sipa> it's just an extra commit added
543 2012-04-10 16:34:16 <jgarzik> sipa: -externalip seems unclear whether it should be only used by v4 nodes, may be used by either but "probably" wants to be used by v4 nodes, etc.
544 2012-04-10 16:34:38 <jgarzik> sipa: I lean towards "-externalip4" and only use the feature on v4, because it has only been demonstrated to be needed in v4-land
545 2012-04-10 16:34:59 <luke-jr> -externalnat ?
546 2012-04-10 16:35:27 <sipa> jgarzik: as soon as onioncat support done, you'll probably need -externalip=blalkfhkjf.onion
547 2012-04-10 16:35:42 <sipa> with exactly the same semantics
548 2012-04-10 16:37:01 <jgarzik> sipa: fair enough
549 2012-04-10 16:42:00 <sipa> one other issue is socks4/socks5
550 2012-04-10 16:42:16 <sipa> you probably want a -socks=<n> option to select one
551 2012-04-10 16:45:03 <jgarzik> sipa: any idea if socks4 still widely deployed?
552 2012-04-10 16:45:16 <jgarzik> maybe people only use that due to lack of socks5 support
553 2012-04-10 16:45:25 <jgarzik> (don't know either way)
554 2012-04-10 16:45:47 <sipa> i'll make socks5 default
555 2012-04-10 16:49:34 <sipa> maybe if -externalip=X is selected, and X is in network N, then no other autodetection of local addresses in N should be performed
556 2012-04-10 16:50:03 <jgarzik> sipa: that seems highly reasonable
557 2012-04-10 16:50:18 <phantomcircuit> socks4 is very common as a proxy in corporate environments
558 2012-04-10 16:50:31 <phantomcircuit> outside that it's only seen as parts of malware providing open proxies
559 2012-04-10 16:50:38 <sipa> now, it takes priority anyway, so it probably doesn't matter really
560 2012-04-10 16:50:54 <phantomcircuit> the patch i saw completely disabled socks4
561 2012-04-10 16:51:03 <sipa> phantomcircuit: it's changed now
562 2012-04-10 16:51:14 <phantomcircuit> try 5 then try 4?
563 2012-04-10 16:51:40 <sipa> no, -socks=4 is you want socks4
564 2012-04-10 16:52:19 <phantomcircuit> well for some socks proxies you can use either 4 or 5
565 2012-04-10 16:52:30 <phantomcircuit> so it might be worth it to try 5 then try 4 as a fallback
566 2012-04-10 16:53:59 <jgarzik> phantomcircuit: when configuring their proxy, won't the user typically know which ahead of time?
567 2012-04-10 16:54:11 <jgarzik> (configuring their proxy support inside bitcoin, I mean)
568 2012-04-10 17:04:05 <phantomcircuit> jgarzik, probably but there is little reason not to have a -socks option which tries one then the other
569 2012-04-10 17:41:38 <jgarzik> phantomcircuit: apologize for the repeated question (over time), but...
570 2012-04-10 17:41:49 <jgarzik> phantomcircuit: does your python stuff validate the full block chain including scripts etc.?
571 2012-04-10 17:41:55 <phantomcircuit> not even close
572 2012-04-10 17:42:04 <jgarzik> phantomcircuit: know any py code that does?
573 2012-04-10 17:42:04 <phantomcircuit> there's actually a bug in the block header ordering
574 2012-04-10 17:42:14 <phantomcircuit> so it doesn't even do that correctly
575 2012-04-10 17:45:38 <helo> <3 python
576 2012-04-10 17:47:30 <jgarzik> helo: it's nice as long as you don't need multi-threading
577 2012-04-10 17:47:50 <helo> i'm surprised you aren't using twisted for the networking stuff... i'm sure you're sick of hearing that though :)
578 2012-04-10 17:48:38 <jgarzik> helo: async networking doesn't fix the problem of stalling for a while doing CPU-intensive tasks
579 2012-04-10 17:48:58 <jgarzik> it can fork, so one may muddle thorough with heavyweight threading
580 2012-04-10 17:49:21 <jgarzik> but then you waste more CPU copying data around
581 2012-04-10 17:49:24 <helo> oh, i made the twisted comment regarding phantomcircuit's bitcoin-alt python implementation
582 2012-04-10 17:49:27 <jgarzik> or doing ugly shm'ing
583 2012-04-10 17:57:46 <helo> experience has shown me the python approach to multi-threading is usually valid: don't do it.
584 2012-04-10 18:02:43 <lianj> helo: don't do it on a multicore machine ^^ wahh
585 2012-04-10 18:05:26 <pjorrit> stick to one python per trouser
586 2012-04-10 18:14:08 <gribble> New news from bitcoinrss: laanwj opened pull request 1075 on bitcoin/bitcoin <https://github.com/bitcoin/bitcoin/pull/1075>
587 2012-04-10 18:20:05 <BlueMattBot> Project Bitcoin build #276: STILL FAILING in 1 min 21 sec: http://jenkins.bluematt.me/job/Bitcoin/276/
588 2012-04-10 18:20:06 <BlueMattBot> laanwj: Show error message instead of exception crash when unable to bind RPC port
589 2012-04-10 18:23:16 <phantomcircuit> helo, i absolutely detest twisted
590 2012-04-10 18:23:23 <phantomcircuit> the basic concept is solid
591 2012-04-10 18:23:29 <phantomcircuit> the implementation is absolute garbage
592 2012-04-10 18:23:43 <sipa> gavinandresen: further remarks on #1054?
593 2012-04-10 18:24:11 <gavinandresen> lemme look
594 2012-04-10 18:28:30 <luke-jr> phantomcircuit: seen Eloipool code btw?
595 2012-04-10 18:29:45 <sipa> luke-jr: had any more problems with building buildinfo?
596 2012-04-10 18:29:54 <luke-jr> sipa: not afaik
597 2012-04-10 18:33:59 <BlueMattBot> Project Bitcoin build #277: STILL FAILING in 15 sec: http://jenkins.bluematt.me/job/Bitcoin/277/
598 2012-04-10 18:39:15 <luke-jr> [15:55:50] <superjames> the fuck is [category] => immature
599 2012-04-10 18:39:17 <luke-jr> [15:55:55] <superjames> with no address
600 2012-04-10 18:39:18 <luke-jr> [15:56:00] <superjames> broke my code
601 2012-04-10 18:39:20 <luke-jr> lol
602 2012-04-10 18:39:56 <superjames> thought it was a bad joke
603 2012-04-10 18:40:17 <luke-jr> superjames: I thought it was pretty funny >_<
604 2012-04-10 18:40:44 <luke-jr> superjames: it used to be invisible until 120 confirms, then it suddenly appeared in your history :P
605 2012-04-10 18:41:18 <phantomcircuit> luke-jr, i have not seen this before
606 2012-04-10 18:41:46 <luke-jr> phantomcircuit: https://gitorious.org/bitcoin/eloipool
607 2012-04-10 18:45:04 <MasterChief> guys
608 2012-04-10 18:45:11 <MasterChief> strange question
609 2012-04-10 18:45:34 <MasterChief> does anyone know if neodymium iron boton magnets outgas anything nasty
610 2012-04-10 18:46:05 <MasterChief> because ive linked my sore throat to when i got one to put on my desl
611 2012-04-10 18:46:08 <MasterChief> desk
612 2012-04-10 18:46:28 <MasterChief> so science guys?
613 2012-04-10 18:47:36 <MasterChief> fuck what channel is this, the wrong one
614 2012-04-10 18:47:43 <Joric> yes, neodymium oxide is quite poisonous
615 2012-04-10 18:48:06 <MasterChief> well shit
616 2012-04-10 18:48:25 <MasterChief> these magnets are plated though
617 2012-04-10 18:48:33 <upb> neodymium iron boton magnets is what brought about the holocaust
618 2012-04-10 18:49:04 <MasterChief> they brought a holocaust on my fingernail
619 2012-04-10 18:49:52 <Joric> hazardous in case of skin contact (irritant), of ingestion, of inhalation
620 2012-04-10 18:51:05 <MasterChief> ok i wil put it on the other side of the room for now and go to sleep
621 2012-04-10 18:54:31 <gmaxwell> I've never suffered any obvious ill effects from handling oxydized neodymium magnets, except perhaps an interest in bitcoin
622 2012-04-10 18:54:47 <copumpkin> gmaxwell: you seen the guy who crushed his finger in the giant magnets?
623 2012-04-10 18:55:14 <BlueMatt> I've never suffered any obvious ill effects from handling oxydized neodymium magnets, except perhaps when I lost my wallet because my hdd got erased
624 2012-04-10 18:55:18 <BlueMatt> :)
625 2012-04-10 18:56:21 <gmaxwell> copumpkin: No but I've been injured just enough by them to not want to see.
626 2012-04-10 18:56:42 <copumpkin> it pulled the skin off his fingertip in a pleasant way
627 2012-04-10 18:58:56 <Joric> are they magneting through the finger?
628 2012-04-10 18:59:26 <Joric> never had those things
629 2012-04-10 19:00:05 <MasterChief> theres a vis on youtube of a guy holding up about 100lb of barbells through a pine dinner table with a large nib magnet
630 2012-04-10 19:00:39 <MasterChief> they are an existential threat to anyone who goes near them
631 2012-04-10 19:02:16 <Keefe> how does Armory detect whether the Satoshi client is running?
632 2012-04-10 19:10:13 <phantomcircuit> Keefe, probably checks for port 8333/8332
633 2012-04-10 19:14:19 <Keefe> 8332 is open, yet Armory says Satoshi isn't available
634 2012-04-10 19:14:35 <Diapolo> Are there any Windows users out there that are able to compile Bitcoin-Qt from a Git repo and have compiled and working boost libs? I need someone to try an experimental optimisation for initial block-chain processing (reduce file fragmentation) in the near future.
635 2012-04-10 19:15:19 <BlueMatt> jenkins can do it
636 2012-04-10 19:15:38 <BlueMatt> jenkins.bluematt.me
637 2012-04-10 19:20:40 <Diapolo> What would be the best way to reach him?
638 2012-04-10 19:21:02 <BlueMatt> its a website
639 2012-04-10 19:21:18 <BlueMatt> if you point it to a git repo, it will build it, run the test suite and sanity-test the build
640 2012-04-10 19:21:43 <BlueMatt> give me a git repo when you want a build, and Ill whip one up for you
641 2012-04-10 19:22:14 <Diapolo> okay I understand ^^ thought it was a person :D
642 2012-04-10 19:23:02 <Diapolo> well it's not really a build problem, but a test for which I need a real person, who can check how fragmented the blockchain file is and if it's processed somewhat quicker
643 2012-04-10 19:23:23 <BlueMatt> oh, I thought you needed a build so you could test yourself
644 2012-04-10 19:24:40 <Diapolo> but to understand that build, if succeeded could be used by a person i.e. download zip, extract and start bitcoin-Qt?
645 2012-04-10 19:24:48 <Diapolo> then it would help me
646 2012-04-10 19:25:13 <BlueMatt> yea
647 2012-04-10 19:26:12 <Diapolo> I'll talk to you, when I have some experimental branch ready :) thanks.
648 2012-04-10 19:37:50 <sipa> Diablo-D3: you can't run it yourself?
649 2012-04-10 19:37:54 <sipa> ow
650 2012-04-10 19:38:05 <sipa> Diapolo's gone
651 2012-04-10 19:57:02 <djoot> hey luke-jr , what's a good font for showing tonal?
652 2012-04-10 20:10:26 <luke-jr> djoot: for IRC, or in general?
653 2012-04-10 20:19:48 <pjorrit> holy tonal.. wtf
654 2012-04-10 20:21:35 <Diablo-D3> [05:37:50] <sipa> Diablo-D3: you can't run it yourself?
655 2012-04-10 20:21:36 <Diablo-D3> tabfail
656 2012-04-10 20:23:45 <sipa> yeah
657 2012-04-10 20:45:15 <Diablo-D3> http://githubbub.com/
658 2012-04-10 20:49:42 <djoot> well, in general luke-jr , but yeah, irc in particular :)
659 2012-04-10 20:52:11 <pjorrit> cool artsy effect, pretty useless though Diablo-D3  ;D
660 2012-04-10 21:02:07 <luke-jr> djoot: I prefer a font called Console that was distributed with KDE 3
661 2012-04-10 21:02:27 <luke-jr> I have a mirror here http://luke.dashjr.org/education/tonal/glyphs/fonts/Console/
662 2012-04-10 21:02:52 <djoot> thanks
663 2012-04-10 21:10:10 <graingert> http://proggyfonts.com/
664 2012-04-10 21:10:25 <etotheipi_> ProggyCleanTTSZ FTW!
665 2012-04-10 21:10:47 <graingert> lol
666 2012-04-10 21:10:51 <graingert> that's totally my font
667 2012-04-10 21:10:55 <graingert> xD
668 2012-04-10 21:10:58 <etotheipi_> lol
669 2012-04-10 21:11:34 <etotheipi_> they are very nice fonts for programming...
670 2012-04-10 21:13:57 <graingert> unfortuneatly if I ever get a retina screen
671 2012-04-10 21:14:03 <graingert> they will be unvisiable
672 2012-04-10 21:15:05 <graingert> luke-jr: do you have a link to the official source of that font?
673 2012-04-10 21:15:27 <luke-jr> graingert: no, I imagine it's in the KDE 3 history somewhere :/
674 2012-04-10 21:15:34 <luke-jr> KDE 4 didn't bother to keep it around
675 2012-04-10 21:15:39 <graingert> Diablo-D3: ECLIPSE!
676 2012-04-10 21:15:48 <graingert> Diablo-D3: look at githubbug
677 2012-04-10 21:15:51 <graingert> hubbub*
678 2012-04-10 21:15:54 <graingert> it's just eclipse
679 2012-04-10 21:16:14 <graingert> they have eclipsed all other item
680 2012-04-10 21:17:16 <Diablo-D3> must be merge day
681 2012-04-10 21:17:24 <Diablo-D3> btw, I actually hate proggyfont
682 2012-04-10 21:17:28 <Diablo-D3> they look great
683 2012-04-10 21:17:31 <Diablo-D3> they really really do
684 2012-04-10 21:17:38 <Diablo-D3> but even the biggest variant isnt big enough
685 2012-04-10 21:18:20 <Diablo-D3> I used "Fixed 11" which maps to 9x15.pcf on my machine
686 2012-04-10 21:18:32 <Diablo-D3> (or I just specify 9x15 directly for urxvt)
687 2012-04-10 21:21:47 <djoot> thanks for the fonts luke
688 2012-04-10 21:23:00 <djoot> now #bitcoin-watch just looks cryptic :)
689 2012-04-10 21:23:57 <Diablo-D3> trust me, the fixed bitmap fonts that come with X are so much better
690 2012-04-10 21:25:15 <luke-jr> XD
691 2012-04-10 21:25:22 <luke-jr> Diablo-D3: Console *is* a fixed bitmap font.
692 2012-04-10 21:26:07 <Diablo-D3> luke-jr: Console is a vague font name, which one do you mean?
693 2012-04-10 21:31:19 <luke-jr> Diablo-D3: the one I have mirrored on my site.
694 2012-04-10 21:34:16 <etotheipi_> Diablo-D3: you just need to put your monitors on articulating mounts and put it like 6 in from your face... then the proggy fonts are easier to read
695 2012-04-10 21:34:38 <graingert> get a low rez LED panel
696 2012-04-10 21:35:15 <etotheipi_> yeah, get a 24" monitor with 1280x1024 native res
697 2012-04-10 21:36:08 <graingert> no 800*600
698 2012-04-10 21:36:56 <Diablo-D3> etotheipi_: I have a 85 dpi monitor (1920x1200 at 26")
699 2012-04-10 21:37:03 <Diablo-D3> its about a foot from my face
700 2012-04-10 21:37:32 <etotheipi_> well I was kidding after all, but I do highly recommend the monitor mounts ... totally worth it
701 2012-04-10 21:37:49 <etotheipi_> I have two 24", 1920x1200 levitating in front of me
702 2012-04-10 21:38:35 <etotheipi_> and a 24"x18" whiteboard on the desk below them (where the monitor stands would normally be
703 2012-04-10 21:39:16 <Diablo-D3> my 26" has an expensive stand
704 2012-04-10 21:39:19 <Diablo-D3> its really nice
705 2012-04-10 21:39:38 <graingert> I want to get one of those 6 way stands
706 2012-04-10 21:39:46 <Diablo-D3> and I suspect I could get vesa mounts and hang three monitors off of this
707 2012-04-10 21:40:07 <etotheipi_> the ones I got just clip/tighten to the back edge of the desk
708 2012-04-10 21:40:30 <etotheipi_> http://www.amazon.com/LX-Desk-Mount-LCD-Arm/dp/B00358RIRC/ref=sr_1_7?ie=UTF8&qid=1334101220&sr=8-7
709 2012-04-10 21:41:04 <etotheipi_> it's *really nice* to have all that desk space back
710 2012-04-10 21:41:55 <Diablo-D3> I have a desk so big Im surprised the navy hasnt pressed it into service as an aircraft carrier
711 2012-04-10 21:42:15 <etotheipi_> although, I suppose that wall mounting would be even better
712 2012-04-10 21:43:47 <etotheipi_> on the other hand, I've found that no matter how much desk space you have... you always find a way to use all of it
713 2012-04-10 21:48:11 <etotheipi_> as long as you have so much desk space, you might turn some of it into a whiteboard:  http://www.amazon.com/Adhesive-Erase-Material-10-Feet-Long/dp/B0033STJ08
714 2012-04-10 22:33:59 <BlueMattBot> Project Bitcoin build #278: STILL FAILING in 16 sec: http://jenkins.bluematt.me/job/Bitcoin/278/
715 2012-04-10 22:34:00 <BlueMattBot> pieter.wuille: Force regeneration of build.h in qmake builds
716 2012-04-10 22:34:07 <BlueMatt> fuck off BlueMattBot
717 2012-04-10 22:36:17 <luke-jr> lol
718 2012-04-10 23:16:31 <TuxBlackEdo> i got some beef to pick with you guys
719 2012-04-10 23:16:38 <TuxBlackEdo> hear me out
720 2012-04-10 23:16:57 <TuxBlackEdo> I look at my "about bitcoin" thing to figure out what my version # is
721 2012-04-10 23:16:59 <TuxBlackEdo> and it says
722 2012-04-10 23:17:07 <TuxBlackEdo> 0.6.0.6-beta
723 2012-04-10 23:17:20 <TuxBlackEdo> so i got to http://bitcoin.org/ to check if i got the latest version
724 2012-04-10 23:17:22 <TuxBlackEdo> and it says
725 2012-04-10 23:17:30 <Diablo-D3> thats 0.6.0 final
726 2012-04-10 23:17:30 <TuxBlackEdo> Latest version: 0.6.0
727 2012-04-10 23:17:36 <Diablo-D3> they habitually fuck version strings up
728 2012-04-10 23:17:42 <BlueMatt> 0.6.0.X == 0.6.0rcX
729 2012-04-10 23:17:45 <BlueMatt> rc6 == final
730 2012-04-10 23:17:47 <TuxBlackEdo> oh.. well there's my answer
731 2012-04-10 23:17:52 <luke-jr> mine says
732 2012-04-10 23:17:58 <Diablo-D3> bluematt: bitcoin devs == FUCKING LAZY
733 2012-04-10 23:18:02 <luke-jr> Bitcoin version v0.6.0-163-g8a77024-beta
734 2012-04-10 23:18:17 <BlueMatt> Diablo-D3: the0.6 version numbers are...maybe not perfect
735 2012-04-10 23:18:49 <luke-jr> sipa: is this a bug in your buildinfo branch? shouldn't it be 0.6.0.6 at least?
736 2012-04-10 23:19:04 <sipa> luke-jr: no, it uses git tags
737 2012-04-10 23:19:15 <sipa> and the git tag is v0.6.0
738 2012-04-10 23:19:19 <luke-jr> sipa: how does it work for alerts?
739 2012-04-10 23:19:27 <sipa> independent
740 2012-04-10 23:19:37 <luke-jr> so it's still 60006+?
741 2012-04-10 23:20:04 <luke-jr> or wait, alerts use the protocol version still, don't that?
742 2012-04-10 23:20:05 <luke-jr> they*
743 2012-04-10 23:22:07 <sipa> as it uses git tags, the version reported will actually include the rcN part for release candidates
744 2012-04-10 23:22:11 <sipa> which is probably a good thing
745 2012-04-10 23:23:21 <luke-jr> I think it's worth it.