1 2012-05-25 00:04:31 <jgarzik> anyways... cheers!  if I broke something, it will get fixed in two weeks :)
  2 2012-05-25 00:04:47 <luke-jr> &
  3 2012-05-25 00:05:58 <BlueMatt> croatia...fun
  4 2012-05-25 01:01:33 <graingert> so who's heard of http://standards.freedesktop.org/basedir-spec/basedir-spec-0.6.html
  5 2012-05-25 01:02:09 <gribble> New news from bitcoinrss: graingert opened issue 1390 on bitcoin/bitcoin <https://github.com/bitcoin/bitcoin/issues/1390>
  6 2012-05-25 01:02:23 <graingert> http://standards.freedesktop.org/basedir-spec/basedir-spec-latest.html
  7 2012-05-25 01:03:47 <graingert> it gives the proper locations of the wallet the config files and the block chain
  8 2012-05-25 01:11:28 <luke-jr> graingert: to be taken with a grain of salt; freedesktop.org is GNOME-centric
  9 2012-05-25 01:14:55 <graingert> it's mostly stuff for between applications
 10 2012-05-25 01:15:02 <graingert> *desktops
 11 2012-05-25 01:15:14 <graingert> eg: DnD is the drag-and-drop specification shared between GTK+ and Qt.
 12 2012-05-25 01:15:25 <graingert> "freedesktop.org is building a base platform for desktop software on Linux and UNIX.  The elements of this platform have become the backend for higher-level application-visible APIs such as Qt, GTK+, XUL, VCL, WINE, GNOME, and KDE.  The base platform is both software and specifications. "
 13 2012-05-25 01:20:27 <graingert> luke-jr: I really don't see any gnome centric stuff there
 14 2012-05-25 01:20:55 <luke-jr> graingert: Telepathy
 15 2012-05-25 01:21:49 <luke-jr> and when I brought it to KDE's attention, they were all "yeah we know, they're basically GNOME nowadays" -.-
 16 2012-05-25 01:22:50 <graingert> and in what way is telepathy GTK?
 17 2012-05-25 01:23:02 <graingert> and can I have a citation for that quote
 18 2012-05-25 01:24:02 <graingert> http://telepathy.freedesktop.org/doc/telepathy-qt/
 19 2012-05-25 01:24:16 <graingert> copyright Nokia Corporation
 20 2012-05-25 01:24:22 <graingert> not very gnome centric
 21 2012-05-25 01:24:57 <luke-jr> graingert: the main libraries are GNOME-stack
 22 2012-05-25 01:25:04 <luke-jr> even if they have Qt bindings
 23 2012-05-25 01:25:22 <luke-jr> furthermore, those Qt bindings require Qt built as part of a GNOME system to work
 24 2012-05-25 01:26:22 <graingert> http://www.developer.nokia.com/Community/Wiki/Libglib
 25 2012-05-25 01:26:33 <graingert> and libxml seem to be the only "GTK" deps
 26 2012-05-25 01:27:45 <graingert> actually you can use this http://packages.ubuntu.com/precise/libtelepathy-qt4-2
 27 2012-05-25 01:27:55 <graingert> and that uses only the qt4-xml libs
 28 2012-05-25 01:29:27 <freewil> anyone want to help me out by generating some orphaned txs on testnet?
 29 2012-05-25 01:29:37 <graingert> either way it's better than dumping all the stuff in ~/.bitcoin/
 30 2012-05-25 01:29:52 <graingert> that way we can say "clear your cache"
 31 2012-05-25 01:30:00 <graingert> or delete the config files
 32 2012-05-25 01:30:10 <graingert> and be safe from them deleating the wallet
 33 2012-05-25 01:35:05 <graingert> deleting*
 34 2012-05-25 01:35:12 <graingert> and maybe some day I'll learn to spell
 35 2012-05-25 01:37:42 <freewil> im glad you clarified that, ive been confused for the last 5 mins
 36 2012-05-25 04:50:58 <olp> What's the proper way to use bitcoin-qt with Tor? I disconnect my network, Start up bitcoin-qt, change the socks proxy settings to 127.0.0.1:9050, exit bitcoin-qt, start network, start bitcoin-qt, and I get "Invalid -proxy address" and bitcoin-qt closes.
 37 2012-05-25 04:52:08 <olp> but if I set the proxy settings with network connections on, then it works. But then there's the possibility that info is being leaked...
 38 2012-05-25 05:01:05 <freewil> olp, i dont think tor is officially supported yet
 39 2012-05-25 05:01:12 <freewil> https://github.com/bitcoin/bitcoin/pull/1174
 40 2012-05-25 05:15:41 <olp> I think I got it to work by editing .config/Bitcoin/Bitcoin-Qt.conf
 41 2012-05-25 05:15:59 <olp> Where else does bitcoin hide config files!
 42 2012-05-25 05:16:57 <Diablo-D3> olp: um
 43 2012-05-25 05:17:13 <Diablo-D3> setting it in the qt ui settings should work fine
 44 2012-05-25 05:17:42 <olp> Well feel free to test it out as I mentioned (on Ubuntu at least)
 45 2012-05-25 05:19:06 <Diablo-D3> well, it worked the last time I tried
 46 2012-05-25 05:19:09 <Diablo-D3> although that was like a year ago
 47 2012-05-25 05:19:16 <Diablo-D3> bitcoin is growing native tor support, though
 48 2012-05-25 05:19:21 <Diablo-D3> so I suggest you wait for that
 49 2012-05-25 05:19:31 <olp> ya the key point is that it doesnt work when you set it while network connections are off
 50 2012-05-25 05:19:50 <Diablo-D3> weird
 51 2012-05-25 05:19:54 <Diablo-D3> tell gmaxwell and sipa
 52 2012-05-25 05:19:58 <Diablo-D3> and gavin
 53 2012-05-25 05:20:03 <olp> its ok I can file a little report on github
 54 2012-05-25 07:27:59 <sipa> olp: where exactly do you set that?
 55 2012-05-25 07:28:48 <SomeoneWeird> what should i be using to hash passwords?
 56 2012-05-25 07:29:04 <sipa> for what purpose?
 57 2012-05-25 07:29:46 <SomeoneWeird> for storing in a db
 58 2012-05-25 07:29:48 <SomeoneWeird> ?
 59 2012-05-25 07:32:43 <sipa> just for authentication?
 60 2012-05-25 07:33:49 <SomeoneWeird> yeah, but what else could it be used for?
 61 2012-05-25 07:34:08 <sipa> to encrypt stuff, for example
 62 2012-05-25 07:34:43 <olp> sipa: Settings->Options->Connect through SOCKS4 proxy
 63 2012-05-25 07:35:02 <SomeoneWeird> ah right, just auth for a website
 64 2012-05-25 07:35:22 <olp> it shows Proxy IP: 0:0:0:0:0:0:0:0, Port: 0
 65 2012-05-25 07:35:55 <sipa> SomeoneWeird: iterated hmac-sha512, using custom salt, truncate to 256 bit
 66 2012-05-25 07:36:08 <sipa> that'd do :)
 67 2012-05-25 07:36:30 <sipa> olp: sounds correct
 68 2012-05-25 07:37:52 <olp> when I set it to 127.0.0.1:9050 and close, the file .config/Bitcoin/Bitcoin-Qt.conf shows 0:0:0:0:0:0:0:0 instead of 127.0.0.1, so I have to manually set it otherwise Bitcoin starts with an error and closes
 69 2012-05-25 07:38:25 <sipa> hmm, which version of the code is that, olp?
 70 2012-05-25 07:38:37 <olp> 0.6.2
 71 2012-05-25 07:38:53 <sipa> ok, shouldn't be
 72 2012-05-25 07:39:00 <sipa> i'll have a look
 73 2012-05-25 07:39:14 <olp> but this only happens when I disconnect my network connections before starting bitcoin-qt to set the proxy settings
 74 2012-05-25 07:39:54 <olp> Network Manager icon -> uncheck Enable Networking
 75 2012-05-25 07:40:22 <weex> SomeoneWeird: look at phpass
 76 2012-05-25 07:40:47 <SomeoneWeird> not using php weex
 77 2012-05-25 07:40:51 <SomeoneWeird> and thanks sipa :)
 78 2012-05-25 07:41:19 <weex> mkay
 79 2012-05-25 07:43:00 <sipa> olp: hmm, strange that it would fail to recognize 127.0.0.1 even when networking is off
 80 2012-05-25 07:43:48 <olp> its on a VM that Im trying
 81 2012-05-25 07:43:52 <sipa> olp: anyway, if you set proxy, even while networking is enabled, you should not leak anything
 82 2012-05-25 07:44:29 <sipa> you can also start the program with -proxy=127.0.0.1:9050
 83 2012-05-25 07:45:07 <olp> ya for me I can find workarounds, just for other people it should be ideally easier
 84 2012-05-25 07:46:08 <olp> and Im thinking maybe it broadcasts some wallet addresses when it starts up, thats why Id prefer it to be setup before connecting network
 85 2012-05-25 07:46:36 <sipa> it never broadcasts wallet information :)
 86 2012-05-25 07:47:12 <sipa> ah, it may broadcast unconfirmed transactions sometimes
 87 2012-05-25 07:48:53 <olp> anyway, no biggie, but if you do test it out please let me know if Im wrong and its just some of my system settings that messed it up
 88 2012-05-25 09:04:01 <gribble> New news from bitcoinrss: Prohest opened issue 1391 on bitcoin/bitcoin <https://github.com/bitcoin/bitcoin/issues/1391>
 89 2012-05-25 10:08:00 <luke-jr> hmm
 90 2012-05-25 10:08:10 <luke-jr> 0.5.x just sync'd from scratch without problems
 91 2012-05-25 10:08:35 <sipa> why wouldn't it?
 92 2012-05-25 10:18:45 <luke-jr> sipa: MtRed reported it was stuck on a P2SH block
 93 2012-05-25 10:18:56 <luke-jr> ERROR: ConnectInputs() : 968a692ab9 P2SH VerifySignature failed
 94 2012-05-25 10:19:35 <luke-jr> thing is, I don't see how that could result from any kind of corruption
 95 2012-05-25 10:19:47 <luke-jr> and when he upgraded to 0.6.x, it worked without detecting any errors
 96 2012-05-25 10:20:16 <luke-jr> (after which he went back to 0.5.x since his patches don't work with 0.6.x)
 97 2012-05-25 10:20:37 <BlueMatt> has he run -checkblocks?
 98 2012-05-25 10:20:47 <luke-jr> no
 99 2012-05-25 10:21:05 <sipa> i doubt -checkblocks will reveal much on 0.5.x
100 2012-05-25 10:21:06 <luke-jr> but if corruption were the problem, wouldn't 0.6.x have had the same issue?
101 2012-05-25 10:21:15 <sipa> yes
102 2012-05-25 10:21:53 <luke-jr> plus, the very fact that it got to the P2SH check stage means the scriptSig hash matched, and nothing from the old txn is used at that point
103 2012-05-25 10:55:42 <coinmaster> hi everyone, what is the max number of letters in the comment ?
104 2012-05-25 10:58:33 <gavinandresen> what comment?
105 2012-05-25 11:00:30 <BlueMatt> anyone know off-hand how lazy bdb's delete is?
106 2012-05-25 11:01:55 <coinmaster> gavinandersen: basicaly i wonder about how many text i can add to each transaction actualy.. cause i think about a mail that uses the bitcoin system in a way all letters would be saved in the blockchain
107 2012-05-25 11:02:24 <gmaxwell> coinmaster: comments are not part of transactions.
108 2012-05-25 11:02:26 <gavinandresen> oh, the comment in the send* RPC commands?  No limit.
109 2012-05-25 11:02:50 <gmaxwell> coinmaster: they're local only.
110 2012-05-25 11:02:57 <gavinandresen> but as gmaxwell says that information is stored only in your wallet
111 2012-05-25 11:03:16 <coinmaster> gmaxwell: ok understood is there a way to store it in the blockchain ?
112 2012-05-25 11:03:47 <sipa> coinmaster: there may be, but you shouldn't (the blockchain is only for information that is necessary for the world to validate the transactions)
113 2012-05-25 11:04:01 <gavinandresen> Sure, encode the information in transaction fees.
114 2012-05-25 11:04:18 <sipa> bleh...
115 2012-05-25 11:04:20 <gmaxwell> coinmaster: the system is relatively hostile to additional data, adding non-transaction data to the blockchain is parasitic and abusive. The users of bitcoin are all self interested in validating transaction data because its what proves that the system is non-inflationary, other things? not so much.
116 2012-05-25 11:04:23 <gavinandresen> :)
117 2012-05-25 11:04:46 <coinmaster> sipa: i think about to have a bitcoin alternative blockchain like namecoins for that
118 2012-05-25 11:06:39 <luke-jr> coinmaster: use email
119 2012-05-25 11:07:23 <coinmaster> luke-jr: i guess i know what you mean, but to store the mail encoded in a chain should be the purpose
120 2012-05-25 11:07:34 <luke-jr> coinmaster: why?
121 2012-05-25 11:08:13 <coinmaster> luke-jr: 100% public comunication but two times sha 256 encoded
122 2012-05-25 11:08:44 <luke-jr> coinmaster: what need is there for it to be public? or hashed?
123 2012-05-25 11:09:02 <luke-jr> much less reason to force every bitcoin user to deal with it
124 2012-05-25 11:10:06 <coinmaster> luke-jr: no there is no reason to force every bitcoin user to do that.. thats why such a system should be an alternative like namecoin is it furfills aswell a kind of encoded public record
125 2012-05-25 11:10:33 <luke-jr> I see no problem with the existing email system.
126 2012-05-25 11:10:46 <sipa> coinmaster: do you want to use it as a timestamping service?
127 2012-05-25 11:10:55 <sipa> (in which case, see chronobit)
128 2012-05-25 11:12:55 <coinmaster> sipa: chronobit could be a part of this mailchain
129 2012-05-25 11:13:13 <sipa> the questions remains why you want the data in a blockchain
130 2012-05-25 11:13:24 <sipa> and chronobit is not a separate blockchain
131 2012-05-25 11:16:04 <gmaxwell> could be part of?  The point of chronobit is that it provides strongly provable datestamps without adding any size to the blockchain.
132 2012-05-25 11:16:47 <gmaxwell> It remains to be seen if the chain PoW distributed system technique will be applicable outside of bitcoin, the evidence suggests that it may not be.
133 2012-05-25 11:17:44 <coinmaster> sipa: yes right, i see it as a way of secured but transparent comunication for example for a company.. and lets assume the size of the chain would not realy mather
134 2012-05-25 11:19:01 <luke-jr> coinmaster: explain how the blockchain makes communication any more secure or transparent
135 2012-05-25 11:19:11 <sipa> what is 'secured' about it, any more than TCP/IP does, if you add a MAC to it?
136 2012-05-25 11:19:55 <coinmaster> the point is to transparency.. that everoyne that minies the chain does allso have the encoded block
137 2012-05-25 11:21:15 <sipa> and why would anyone be interested in seeing a obfuscated version of other people's data?
138 2012-05-25 11:21:38 <sipa> much less spand hash power to... what purpose exactly?
139 2012-05-25 11:21:51 <coinmaster> sipa: you could ask the same question 10 years ago for bitcoin
140 2012-05-25 11:22:53 <gmaxwell> And it could have been trivially answered.
141 2012-05-25 11:22:53 <sipa> bitcoin's purpose is clear: store and transfer of value over the internet
142 2012-05-25 11:23:00 <sipa> i don't see what your purpose is
143 2012-05-25 11:23:28 <coinmaster> sipa: ok its secured transparency
144 2012-05-25 11:23:28 <sipa> maybe you are looking for something like freenet?
145 2012-05-25 11:23:44 <sipa> give me a concrete example, who would benefit from it?
146 2012-05-25 11:23:59 <coinmaster> sipa: every citizen
147 2012-05-25 11:24:13 <sipa> you're getting even more vague
148 2012-05-25 11:24:15 <coinmaster> sipa: it could be used for votes
149 2012-05-25 11:24:25 <gmaxwell> In any case, the creator of bitcoin sat down and built it rather than arguing about its merits on IRC. I suggest you go do the same. Even if you're unsuccessful you'll perhaps learn something in the process.
150 2012-05-25 11:24:31 <coinmaster> sipa: for basic gov correspondence
151 2012-05-25 11:25:26 <coinmaster> gmxwell: i agree, im on that path
152 2012-05-25 11:25:42 <_Fireball> coinmaster, are you proposing a voting system based on the distributed database and transaction processing approach?
153 2012-05-25 11:25:58 <coinmaster> _Fireball: right
154 2012-05-25 11:26:01 <sipa> try using a DHT!
155 2012-05-25 11:26:07 <_Fireball> there was a thread in bitcointalk.org about it
156 2012-05-25 11:26:13 <_Fireball> with some ideas and concepts
157 2012-05-25 11:28:18 <coinmaster> _Fireball: thanks i try to find that
158 2012-05-25 11:43:38 <copumpkin> is it just me or does bitcoin-qt reject my attempts to focus it the first time around? is that a known issue?
159 2012-05-25 11:45:39 <Diapolo> I'm not sure what you mean ...
160 2012-05-25 11:46:00 <copumpkin> here's a screen recording of it
161 2012-05-25 11:47:14 <copumpkin> (just finishing it)
162 2012-05-25 11:50:46 <copumpkin> http://dl.dropbox.com/u/361503/Screen.m4v
163 2012-05-25 11:50:48 <copumpkin> does that work for you?
164 2012-05-25 11:51:15 <copumpkin> it happens consistently after I hide (cmd+h or the menu option I select in that video)
165 2012-05-25 11:51:28 <Diapolo> alright, yes, but I'm on Win ... this should be fixed with the current Github master branch
166 2012-05-25 11:51:31 <copumpkin> v0.6.2.2-beta
167 2012-05-25 11:51:33 <copumpkin> oh okay
168 2012-05-25 11:51:35 <Diapolo> but there is no official build, that fixes this
169 2012-05-25 11:51:44 <luke-jr> copumpkin: what verison of OSX?
170 2012-05-25 11:51:51 <copumpkin> 10.7.4
171 2012-05-25 11:51:59 <luke-jr> hmm
172 2012-05-25 11:52:16 <copumpkin> I'm in no rush to get it fixed on my end, just wanted to make sure people were aware of it
173 2012-05-25 12:05:03 <Eliel_> would encourage people to pay fees :)
174 2012-05-25 12:07:17 <luke-jr> but wouldn't encourage people to mine here :p
175 2012-05-25 12:07:52 <Eliel_> why not? if it increases the number of people paying a fee, there's bound to be more left for the miners too
176 2012-05-25 12:08:02 <Eliel_> depending on the exact percentages of course
177 2012-05-25 12:08:03 <luke-jr> hmm
178 2012-05-25 12:09:17 <drizztbsd> Eliel_: more fees => more probability to win the lottery :P
179 2012-05-25 12:09:49 <Eliel_> drizztbsd: yes, but it doesn't make sense to put enormous amounts into the lottery nevertheless.
180 2012-05-25 12:10:10 <Eliel_> I expect the winnigs from it would be pretty small.
181 2012-05-25 12:10:16 <BlueMatt> but the miner picks the winner, so why cant the miner pick themselves?
182 2012-05-25 12:10:25 <Eliel_> BlueMatt: no, the block hash picks the winner
183 2012-05-25 12:10:30 <BlueMatt> ah
184 2012-05-25 12:10:31 <Eliel_> and network rules enforce it
185 2012-05-25 12:11:03 <luke-jr> closest txid to the block hash flipped?
186 2012-05-25 12:11:11 <Eliel_> would work
187 2012-05-25 12:11:12 <luke-jr> network rules can't enforce it :p
188 2012-05-25 12:11:26 <BlueMatt> miners could make a rainbow table and pretty much always win there
189 2012-05-25 12:11:26 <luke-jr> Eliel_: actually, it wouldn't work
190 2012-05-25 12:11:37 <drizztbsd> why network rules cannot enforce it?
191 2012-05-25 12:11:44 <luke-jr> drizztbsd: because it'd fork bitcoin
192 2012-05-25 12:11:51 <drizztbsd> oh
193 2012-05-25 12:11:53 <drizztbsd> baad
194 2012-05-25 12:12:04 <Eliel_> BlueMatt: huh? rainbow table of block hashes? ... makes no sense. that would mean blockchain is useless if possible.
195 2012-05-25 12:12:06 <sipa> Eliel_: i disagree with the notion of automatically assuming payments to an input address will arrive
196 2012-05-25 12:12:07 <luke-jr> hmm
197 2012-05-25 12:12:42 <BlueMatt> rainbow table of txes that pay to the miner
198 2012-05-25 12:12:48 <luke-jr> who has old OSX?
199 2012-05-25 12:13:02 <drizztbsd> I have an old OSX (tiger ppc) (at home)
200 2012-05-25 12:13:15 <luke-jr> OK, old i386 OSX
201 2012-05-25 12:13:21 <drizztbsd> ;(
202 2012-05-25 12:13:29 <Eliel_> sipa: it works well enough and I'm pretty sure if this was adopted, they'd soon start arriving.
203 2012-05-25 12:13:29 <luke-jr> since bitcoind isn't likely to support PPC any time soon
204 2012-05-25 12:13:36 <drizztbsd> oh, big endian
205 2012-05-25 12:14:29 <sipa> Eliel_: still, the fact that a reference to a particular key is used in the blockchain is in no way a statement that you're willing to receive funds with it
206 2012-05-25 12:14:36 <Eliel_> BlueMatt: you mean the miner would flood the block with their own txs?
207 2012-05-25 12:14:57 <luke-jr> Eliel_'s system is inherently opt-in for the miner anyway
208 2012-05-25 12:15:07 <luke-jr> if they want to steal it, they can just as well not participate
209 2012-05-25 12:15:19 <BlueMatt> Eliel_: sure, why not?
210 2012-05-25 12:15:24 <Eliel_> luke-jr: true, it could work on a voluntary basis too.
211 2012-05-25 12:15:48 <BlueMatt> fill the block to max size with txes that are likely to put them closer to the block hash
212 2012-05-25 12:16:01 <BlueMatt> fill block to max, after other txes, that is
213 2012-05-25 12:17:06 <Eliel_> BlueMatt: true, no point making it mandatory if that can't be fixed.
214 2012-05-25 12:18:11 <Eliel_> sipa: I don't think that matters. If there's a good enough chance to receive money through this system, people will make transactions that work with it.
215 2012-05-25 12:18:43 <sipa> and i still agree with the idea :)
216 2012-05-25 12:19:23 <Eliel_> sipa: the idea?
217 2012-05-25 12:19:50 <sipa> the idea that you're going to automatically pay to an address used once in a spent txout
218 2012-05-25 12:19:57 <BlueMatt> its a cool idea, would take some work to make it work, but its cool
219 2012-05-25 12:21:05 <BlueMatt> plus converting the network would be kinda...impossible?
220 2012-05-25 12:21:13 <Eliel_> it doesn't need to be all miners for the idea to work :)
221 2012-05-25 12:21:15 <gmaxwell> Sounds like a coin burning recipe.
222 2012-05-25 12:21:47 <Eliel_> gmaxwell: with the current system, yes, but if the idea sticks, it'll stop happening.
223 2012-05-25 12:22:22 <Eliel_> just need to careful in how to design which input address to pick.
224 2012-05-25 12:22:32 <gmaxwell> Eliel_: ... er so you want to make one use address behavior impossible in order to have more gambling?
225 2012-05-25 12:22:35 <Eliel_> so it's possible no matter what kind of transaction
226 2012-05-25 12:22:36 <gavinandresen> I don't like it from a PR/marketing point of view.  It'd just contribute to the "bitcoin is a gambling/ponzi scheme" idea
227 2012-05-25 12:22:56 <Graet> that was my feeling gavinandresen
228 2012-05-25 12:23:28 <Eliel_> how about if it's not a network rule but something miners do voluntarily?
229 2012-05-25 12:23:38 <gavinandresen> And the microcash people already tried something like that, and have abandoned it.
230 2012-05-25 12:23:39 <Eliel_> I think there's definitely incentive out there.
231 2012-05-25 12:24:15 <Eliel_> we already have satoshidice anyway.
232 2012-05-25 12:24:35 <gavinandresen> if all your friends jump off a bridge....
233 2012-05-25 12:25:12 <BlueMatt> sipa: is it testing any, or has it not gotten any?
234 2012-05-25 12:25:14 <JFK911> then i claim to have secret revised wills for them
235 2012-05-25 12:25:18 <BlueMatt> sipa: should you prime it with one or two?
236 2012-05-25 12:25:26 <Eliel_> so, you think having pools doing this would be worse than just having satoshidice?
237 2012-05-25 12:25:29 <gmaxwell> Eliel_: satoshidice gives back a majority of all the funds sent to it.
238 2012-05-25 12:25:49 <gmaxwell> So you can't hope to tap into the same motivational structure.
239 2012-05-25 12:25:52 <sipa> BlueMatt: since jgarzik's seed contains my vps's v6 address statically, and my crawler seeds from existing other seeds, it should immediately have one
240 2012-05-25 12:26:19 <BlueMatt> sipa: same with mine
241 2012-05-25 12:26:24 <Eliel_> well, lottery in Finland gives back only 40%, nearly everyone plays it regardless.
242 2012-05-25 12:26:31 <BlueMatt> (mine has a static see, that is)
243 2012-05-25 12:27:35 <gmaxwell> Eliel_: also miners will always have an incentive to just not participate.
244 2012-05-25 12:27:56 <gmaxwell> And heck, we can't even get all of them to upgrade when they're being orphaned as a result of it.
245 2012-05-25 12:28:03 <Eliel_> :D
246 2012-05-25 12:28:44 <luke-jr> gmaxwell: dunno, blockchain forks that cut out only up to 25% of miners are favourable IMO :P
247 2012-05-25 12:29:01 <luke-jr> as long as I'm not part of the 25%, I benefit from the difficulty drop
248 2012-05-25 12:29:07 <Eliel_> I mean, think about it. Deepbit, for example, could well do something like this for any fee-including transaction that was sent directly to it :P
249 2012-05-25 12:29:26 <Eliel_> they get blocks fast enough that waiting time is not too big.
250 2012-05-25 12:30:10 <gmaxwell> Who turned on the bluesky fountain today in any case?
251 2012-05-25 12:30:58 <sipa> ?
252 2012-05-25 12:32:40 <gmaxwell> we've got Eliel_ with the block chain casino, the guy earlier with the amorphous public good from putting unspecified things in unspecified blockchains.
253 2012-05-25 12:32:52 <Eliel_> :D
254 2012-05-25 12:33:29 <Eliel_> I've been playing with different ideas to persuade people to include fees in transactions lately :)
255 2012-05-25 12:35:46 <gmaxwell> Eliel_: If you want that, make the fee options in the client more comfortable to use... right now you can only set a fee per KB  but then might end up with a fee anywhere from 1x to 100x what you specified or so because size is opaque to users.
256 2012-05-25 12:36:20 <luke-jr> slider on the send transaction page that shows estimated time to confirm ;)
257 2012-05-25 12:36:21 <BlueMatt> make fees much clearer to users, and give miners better options to require fees
258 2012-05-25 12:36:32 <BlueMatt> ^ and make algorithms for miners that make that possible
259 2012-05-25 12:36:37 <luke-jr> BlueMatt: well, the latter is easily merged ;)
260 2012-05-25 12:36:52 <BlueMatt> though, thanks to dicespam, that is somewhat now possible anyway
261 2012-05-25 12:37:05 <luke-jr> git master, bitcoind --help segfaults <.<
262 2012-05-25 12:37:11 <gmaxwell> sipa: one reason you're not finding v6 nodes is that v6 on my network broke
263 2012-05-25 12:37:18 <luke-jr> wait not master, v0.6.2.2
264 2012-05-25 12:37:26 <gmaxwell> wha? really?
265 2012-05-25 12:38:10 <sipa> luke-jr: on your osx build, or in general?
266 2012-05-25 12:39:14 <luke-jr> sipa: on my Linux build
267 2012-05-25 12:39:18 <luke-jr> OSX gives a stranger error
268 2012-05-25 12:39:45 <helo> are 1-transaction blocks really a problem? if people's transactions are being ignored (i.e. insufficient incentive), isn't the solution to provide more incentive via higher fees?
269 2012-05-25 12:39:58 <luke-jr> OSX: terminate called throwing an exceptionAbort trap: 6
270 2012-05-25 12:40:13 <Eliel_> helo: that's a good point :)
271 2012-05-25 12:41:17 <gmaxwell> helo: they were a rather high percentage for a bit but they went down. I don't know why jeff bought it up again.
272 2012-05-25 12:41:24 <luke-jr> helo: 10% of Eligius's blocks are 1-transaction simply because of the time it takes to process the transactions :/
273 2012-05-25 12:41:52 <gmaxwell> luke-jr: don't write mining code in python?
274 2012-05-25 12:41:52 <luke-jr> so IMO punishment isn't the solution: making that faster is
275 2012-05-25 12:43:58 <luke-jr> gmaxwell: that might help
276 2012-05-25 12:45:54 <luke-jr> so it seems OSX 10.6 users are getting DB_RUNRECOVERY errors with my bitcoind
277 2012-05-25 12:46:13 <luke-jr> which is odd, since my cross-compiler is based on 10.5 Xcode
278 2012-05-25 12:46:26 <gmaxwell> Seriously, even if eligius had the whole network hashrate you're telling me that it takes you about 1.05360 minutes to process transactions after a block?
279 2012-05-25 12:46:29 <luke-jr> [14:45:54] <kinlo> luke-jr: and that was with an empty bitcoin data dir
280 2012-05-25 12:46:30 <luke-jr> [14:46:16] <kinlo> so I can't run database recovery - there is no database to recover
281 2012-05-25 12:46:38 <kinlo> pompom
282 2012-05-25 12:46:42 <kinlo> hilighter.
283 2012-05-25 12:46:42 <luke-jr> gmaxwell: perhaps most of the time is miners who don't longpoll
284 2012-05-25 12:47:12 <gmaxwell> luke-jr: it's f@#$#@ conman's @#$@#$@# stupid prev based nonsense I bet!
285 2012-05-25 12:47:24 <luke-jr> gmaxwell: I thoguht he got rid of that :/
286 2012-05-25 12:47:42 <gmaxwell> Doesn't matter, it's not like people upgrade.
287 2012-05-25 12:47:49 <luke-jr> >_<
288 2012-05-25 12:48:28 <luke-jr> anyhow, this 10.6 error bothers me more
289 2012-05-25 12:48:46 <luke-jr> why would they get DB_RUNRECOVERY on an empty dir? :/
290 2012-05-25 12:49:57 <gmaxwell> not looking to the dir they think it is?
291 2012-05-25 12:50:17 <kinlo> unable to initialize mutex: Invalid argument
292 2012-05-25 12:50:26 <kinlo> that's in db.log from that empty dir
293 2012-05-25 12:50:30 <kinlo> well, there is more
294 2012-05-25 12:50:36 <kinlo> but that's the first error
295 2012-05-25 12:51:00 <luke-jr> hmm
296 2012-05-25 12:51:10 <luke-jr> and the --help on OSX has a mutex error too
297 2012-05-25 12:51:14 <gmaxwell> sounds like the !private shm stuff is busted.
298 2012-05-25 12:51:51 <kinlo> http://pastebin.com/PtUJLdYc is the full db.log
299 2012-05-25 12:52:19 <luke-jr> (note this does work on Lion)
300 2012-05-25 12:52:51 <luke-jr> perhaps I should try an older version, 0.5.x?
301 2012-05-25 12:53:12 <kinlo> can I get build instructions?    I'd like to build it myself actually
302 2012-05-25 12:53:22 <luke-jr> kinlo: you have gitian setup already?
303 2012-05-25 12:54:14 <kinlo> I have nothing, but I got xcode installed and related development tools
304 2012-05-25 12:54:35 <luke-jr> kinlo: I am building on Ubuntu.
305 2012-05-25 12:54:38 <kinlo> oic
306 2012-05-25 12:55:13 <kinlo> I'll have a look at it later then, to setup a full gitian setup first
307 2012-05-25 12:55:25 <luke-jr> gitian is Ubuntu-only FWIW
308 2012-05-25 12:55:41 <kinlo> no debian support?
309 2012-05-25 12:55:45 <kinlo> my build server is debian
310 2012-05-25 12:59:30 <sipa> BlueMatt: bah, just realized that the bootstrapper only added addresses to the "to-be-tried" set, and doesn't try them immediately
311 2012-05-25 12:59:47 <sipa> and that the seeds.txt only contains nodes that have been contacted succesfully at least once
312 2012-05-25 12:59:54 <BlueMatt> ah
313 2012-05-25 13:00:52 <luke-jr> hmm
314 2012-05-25 13:00:56 <luke-jr> maybe I should try without -O3
315 2012-05-25 13:01:01 <luke-jr> I figured it was safe since makefile.osx had it
316 2012-05-25 13:08:13 <sipa> BlueMatt: o/
317 2012-05-25 13:08:27 <sipa> [2a02:348:5e:5a29::1]:833312.97%3.41%1.15%0.17%0.04%18153560001 "/Satoshi:0.6.99/"
318 2012-05-25 13:10:01 <BlueMatt> sipa: nice!
319 2012-05-25 13:14:57 <sipa> BlueMatt: and it answers it in ANY requests too
320 2012-05-25 13:15:06 <sipa> (though with very low probability)
321 2012-05-25 13:21:35 <luke-jr> for reference, vrag got this from --help on 10.6:
322 2012-05-25 13:21:37 <luke-jr> [14:18:16] <vragnaroda> terminate called after throwing an instance of 'boost::interprocess::lock_exception' what():  boost::interprocess::lock_exception
323 2012-05-25 13:21:38 <luke-jr> [14:18:18] <vragnaroda> Abort trap
324 2012-05-25 13:23:24 <gavinandresen> luke-jr: try compiling with -DBOOST_SPIRIT_THREADSAFE
325 2012-05-25 13:23:51 <luke-jr> gavinandresen: for v0.6.2.2?
326 2012-05-25 13:24:58 <gavinandresen> luke-jr: shouldn't matter for 0.6.2.2, but the error sounds like the THREADSAFE issue we had with 0.7
327 2012-05-25 13:25:13 <luke-jr> hmm
328 2012-05-25 13:25:37 <luke-jr> ok, when I'm done rebuilding everything (checking for determinism) I'll try without -O3 and with -DBOOST_SPIRIT_THREADSAFE
329 2012-05-25 13:26:51 <luke-jr> gavinandresen: although, I suspect that won't help since bdb is also giving us mutex errors
330 2012-05-25 13:27:39 <gavinandresen> I don't know nuthin about cross-compiling....
331 2012-05-25 14:20:00 <luke-jr> gavinandresen: did you see the miner tests I added to #1246 ?
332 2012-05-25 14:20:35 <luke-jr> gavinandresen: it builds a ~110 blockchain fork on mainnet starting with the genesis block, to test CreateNewBlock with
333 2012-05-25 14:21:25 <luke-jr> might (or might not) be useful for testnet_builder
334 2012-05-25 14:21:56 <gavinandresen> nice
335 2012-05-25 14:33:36 <Glasswalker> Hey, in the data from a getwork. The header is made up of concatenated little endian numbers (the last 32bits of which is the nonce). Is that header then converted to a single big-endian by sha256? (because sha256 assumes big endian) Or are only the individual bytes swapped for big endian? (meaning the bytes themselves are in the same order)
336 2012-05-25 14:34:05 <sipa> sha256 takes as input bytes, and produces as output bytes
337 2012-05-25 14:34:30 <sipa> internally, it interprets these bytes as big-endian number, but that is irrelevant unless you are implementing mining code
338 2012-05-25 14:35:24 <sipa> bitcoin uses the output bytes of sha256 as a 256-bit number in some places; in that case, the conversion is little-endian
339 2012-05-25 14:48:37 <Glasswalker> sipa: And what if I AM implementing mining code? :)
340 2012-05-25 14:49:32 <Glasswalker> I know the output 256bit hash is converted as a single number (to detect if the hash meets difficulty)
341 2012-05-25 14:49:51 <Glasswalker> the concern is the hashing input must exactly match the rest of the networks "method" or it won't make valid hashes
342 2012-05-25 14:50:02 <luke-jr> Glasswalker: see Eloipool source? :p
343 2012-05-25 14:50:37 <Glasswalker> ugh... :) why is it whenever I ask that nobody can simply answer the question and always points me at somebody's (usually somewhat messy due to optimization) source code?
344 2012-05-25 14:50:38 <Glasswalker> :)
345 2012-05-25 14:53:35 <Glasswalker> not saying your code is messy in general, just that generally optimized bitcoin code is not very "clear" (as a direct result of the optimization)
346 2012-05-25 14:56:53 <luke-jr> https://gitorious.org/bitcoin/eloipool/blobs/master/eloipool.py#line136\n3600389
347 2012-05-25 14:56:54 <luke-jr> https://gitorious.org/bitcoin/eloipool/blobs/master/eloipool.py#line136
348 2012-05-25 14:59:24 <Glasswalker> ok that helps :) but that is just telling me what getwork is outputting to the mining client. I need to know when actually hashing (so example using PHP SHA256(SHA256(data))) is the sha256 function doing any byte flipping on the input data?
349 2012-05-25 14:59:43 <helo> php :(
350 2012-05-25 14:59:54 <Glasswalker> lol it was an example posted in the wiki
351 2012-05-25 15:00:01 <Glasswalker> I'm writing this in Verilog to be exact
352 2012-05-25 15:00:35 <luke-jr> Glasswalker: getwork's data already has the first step of SHA256 done on it
353 2012-05-25 15:00:41 <Glasswalker> I'm implementing my own SHA2 algorithm. I've duplicated the "official" sha transform for a single 512bit block. I'll be using a precalculated midstate. But in my bitcoin specific part which wraps the SHA2 core, I need to know what bits (and in what order) hold the nonce
354 2012-05-25 15:00:42 <luke-jr> Glasswalker: so you need to skip that step :P
355 2012-05-25 15:00:44 <Glasswalker> which I'll be incrimenting
356 2012-05-25 15:00:46 <luke-jr> or undo it
357 2012-05-25 15:00:56 <Glasswalker> right
358 2012-05-25 15:00:58 <Glasswalker> I'm skipping it
359 2012-05-25 15:01:05 <Glasswalker> and doing only a single stage using the midstate
360 2012-05-25 15:01:23 <Glasswalker> but I need to know how to incriment the nonce. :) that's all I'm looking for
361 2012-05-25 15:01:32 <luke-jr> it's the same regardless of endian
362 2012-05-25 15:01:47 <Glasswalker> not in Verilog, where I need to explicitly specify endianness of data
363 2012-05-25 15:01:49 <luke-jr> the last 32 bits of the 80 bytes
364 2012-05-25 15:02:17 <luke-jr> this isn't an endian issue
365 2012-05-25 15:02:21 <luke-jr> nonce is endian-free
366 2012-05-25 15:02:30 <luke-jr> you can do it in little endian or big endian, it doesn't matter
367 2012-05-25 15:02:44 <Glasswalker> right because you're incrimenting it. But I needed to know it's location inside the data struct ;)
368 2012-05-25 15:02:54 <Glasswalker> which your comment above answered ;)
369 2012-05-25 15:02:59 <luke-jr> right, my point is that has nothing to do with endians :p
370 2012-05-25 15:03:15 <luke-jr> https://en.bitcoin.it/wiki/Block_hashing_algorithm
371 2012-05-25 15:03:18 <Glasswalker> unless as I asked originally, the whole header is being flipped
372 2012-05-25 15:03:55 <Glasswalker> if the whole 512bit block is being treated as a single 512bit number, versus individual bytes, versus 32bit fields or whatever
373 2012-05-25 15:04:08 <Glasswalker> that can move where the nonce lives in the 512bit data block
374 2012-05-25 15:04:13 <Glasswalker> that's why I was asking about endianness
375 2012-05-25 15:04:26 <graingert> does bitcoinJ verify blocks?
376 2012-05-25 15:04:35 <Glasswalker> so I'm likely just over complicating ;)
377 2012-05-25 15:04:54 <luke-jr> Glasswalker: SHA256 deals with 32-bit chunks
378 2012-05-25 15:05:10 <luke-jr> there is no 512-bit number in SHA256
379 2012-05-25 15:05:33 <Glasswalker> yes I know, but the data input is a 512bit block (which is then broken down, and then expanded into many 32bit words)
380 2012-05-25 15:06:29 <luke-jr> graingert: glib is part of GNOME stack
381 2012-05-25 15:06:38 <luke-jr> graingert: and I want nothing to do with it
382 2012-05-25 15:07:03 <Glasswalker> I've already written that part. But because I'm only implementing the single 512bit transform stage (not the full algorithm which pads, and breaks up the message into 512bit message blocks)
383 2012-05-25 15:07:28 <graingert> luke-jr: GLib was released as a separate library so other developers, those who  did not make use of the GUI-related portions of GTK+, could make use of  the non-GUI portions of the library without the overhead of depending on  a full-blown GUI library.
384 2012-05-25 15:07:41 <luke-jr> Glasswalker: the only thing that deals with 512-bit blocks is the padding, really, IIRC
385 2012-05-25 15:07:52 <Glasswalker> right
386 2012-05-25 15:07:52 <luke-jr> graingert: glib *is* the overhead
387 2012-05-25 15:07:58 <Glasswalker> but all default SHA code will pad
388 2012-05-25 15:07:58 <graingert> riiiigh
389 2012-05-25 15:07:59 <graingert> t
390 2012-05-25 15:08:10 <Glasswalker> none of them do only a single 512bit message block
391 2012-05-25 15:08:14 <graingert> either way you don't need it for telepathy qt
392 2012-05-25 15:08:17 <luke-jr> graingert: Qt Core is a separate library too, but it's still Qt
393 2012-05-25 15:08:20 <Glasswalker> the sha algorithm expects a 512bit message block as input
394 2012-05-25 15:08:20 <luke-jr> graingert: yes, you do
395 2012-05-25 15:08:32 <luke-jr> graingert: not only that, but it requires Qt be compiled as a slave to Glib
396 2012-05-25 15:09:45 <Glasswalker> so because I'm only dealing with a single sha message block (using a midstate from the first hash) I will be getting a 512bit data input which is the second message block (of the 1024bit pre-padded total header)
397 2012-05-25 15:10:56 <graingert> luke-jr: you only need it for telepathy-mission-control-5
398 2012-05-25 15:11:08 <graingert> and anything other than telepathy-sunshine
399 2012-05-25 15:11:14 <luke-jr> graingert: IIRC it's something to do with VoIP support
400 2012-05-25 15:11:24 <graingert> luke-jr: which is broken at best
401 2012-05-25 15:11:27 <luke-jr> &
402 2012-05-25 15:11:49 <graingert> but ignoring that you don't need telepathy-glib
403 2012-05-25 15:12:15 <graingert> someone needs to write a version of telepathy-mission-control without deps on glib
404 2012-05-25 15:12:21 <luke-jr> the only reason I care about Telepathy at all is for the phone support :p
405 2012-05-25 15:12:47 <luke-jr> there's already a real standard for mere IM
406 2012-05-25 15:13:33 <graingert> what do you mean as slave to Glib
407 2012-05-25 15:14:09 <luke-jr> graingert: using glib's main loop
408 2012-05-25 15:14:24 <luke-jr> USE=glib
409 2012-05-25 15:19:37 <Glasswalker> so the output of getwork is data, target, and hash1 correct? data is the last 16 bytes of the block header correct? (block header being 80bytes, and SHA256 only taking 512bit at a time, that leaves 16bytes of the header for the second chunk)
410 2012-05-25 15:19:44 <Glasswalker> just making sure I'm interpreting this correctly
411 2012-05-25 15:21:58 <Glasswalker> or is "data" the entire block header?
412 2012-05-25 15:22:07 <Glasswalker> looks like the entire block header on the satoshi client.
413 2012-05-25 15:24:01 <Glasswalker> so:
414 2012-05-25 15:24:22 <Glasswalker> miner does getwork. Gets back "data" value.
415 2012-05-25 15:25:12 <Glasswalker> takes data value as-is, rips out the first 512bits, hashes this, and that should equal hash1 (and provide a midstate)
416 2012-05-25 15:25:20 <luke-jr> Glasswalker: hash1 is gone
417 2012-05-25 15:25:25 <Glasswalker> right
418 2012-05-25 15:25:28 <Glasswalker> I'm just saying for reference
419 2012-05-25 15:25:46 <luke-jr> Glasswalker: data is the full block header, after being processed by SHA2's first few steps
420 2012-05-25 15:26:41 <Glasswalker> ?
421 2012-05-25 15:26:51 <Glasswalker> that I don't get...
422 2012-05-25 15:27:23 <Glasswalker> define "first few steps" according to the sha2 algorithm?
423 2012-05-25 15:30:12 <Glasswalker> In sha2 usually data comes in arbitrary size. It's got 1 bit appended, then padded with 0 up to modulo 512 minus 64bit for the original (unpadded) length.
424 2012-05-25 15:30:18 <Glasswalker> that is then split into 512bit chunks
425 2012-05-25 15:31:16 <Glasswalker> the first chunk the transform algorithm is primed with the K constants (64x 32bit words)
426 2012-05-25 15:32:26 <Glasswalker> it's transformed, and the 8 32bit words are output for the hash
427 2012-05-25 15:32:52 <Glasswalker> then the next chunk does the same, adding each of the 8 32bit words to the previous chunks output
428 2012-05-25 15:32:56 <Glasswalker> and so on until you have a final hash
429 2012-05-25 15:41:03 <Glasswalker> so when you say it's the full header but processed by the first few rounds of sha, what does that mean?
430 2012-05-25 15:41:10 <Glasswalker> is there documentation on that anywhere?
431 2012-05-25 15:41:27 <Glasswalker> an example, or breakdown of the format?
432 2012-05-25 15:42:51 <Glasswalker> ie: is it the first 256bit (hash1) concatenated with the last 16bytes of the header?
433 2012-05-25 16:51:36 <gribble> New news from bitcoinrss: gavinandresen opened pull request 1392 on bitcoin/bitcoin <https://github.com/bitcoin/bitcoin/pull/1392>
434 2012-05-25 17:27:51 <graingert> I totally read "new genesis" block before the title
435 2012-05-25 17:28:07 <graingert> nanotube: is it possible to pull out the title of the pull request from RSS?
436 2012-05-25 17:33:24 <GTRsdk> most likely is possible
437 2012-05-25 17:56:25 <luke-jr> gavinandresen: has anyone tested your Mac builds on 10.5 or 10.6 recently?
438 2012-05-25 17:56:44 <luke-jr> vragnaroda: ping
439 2012-05-25 17:56:50 <gavinandresen> luke-jr: I'm running 10.6
440 2012-05-25 17:56:52 <vragnaroda> pong
441 2012-05-25 17:57:01 <luke-jr> gavinandresen: oh, ok. hmm
442 2012-05-25 17:58:03 <splatster> I've noticed a small, pesky bug on OS X 10.7 with bitcoin-qt v0.6.2.2
443 2012-05-25 17:58:20 <luke-jr> splatster: does it have to do with showing the window?
444 2012-05-25 17:58:21 <graingert> you guys are going to need to be signed or something for mac builds?
445 2012-05-25 17:58:41 <luke-jr> graingert: that's what I'm working ont
446 2012-05-25 17:58:42 <luke-jr> on*
447 2012-05-25 17:58:46 <graingert> kewl
448 2012-05-25 17:58:50 <splatster> luke-jr: Yup, when you click the app icon the window disappears until you click it again.
449 2012-05-25 17:58:56 <luke-jr> splatster: known & fixed in master
450 2012-05-25 17:59:00 <GTRsdk> For installation in a vanilla ML setup?
451 2012-05-25 17:59:11 <luke-jr> splatster: was the bug not there in 0.6.1 or 0.6.2 btw?
452 2012-05-25 17:59:19 <luke-jr> GTRsdk: ?
453 2012-05-25 17:59:37 <splatster> Nope, first time I've seen it was going from 0.6 to 0.6.2.2
454 2012-05-25 18:00:11 <luke-jr> oh, was probably introduced in 0.6.1 then
455 2012-05-25 18:00:29 <splatster> Also, the Command+M hotkey doesn't work to minimize the window.
456 2012-05-25 18:00:45 <GTRsdk> luke-jr: Mountain Lion which has GateKeeper, which only allows certain apps to run by default... is that what you're getting the signing for?
457 2012-05-25 18:01:11 <luke-jr> GTRsdk: & no
458 2012-05-25 18:01:26 <splatster> I've got a paid Apple Mac dev program license, if someone needs me to sign something for Mountain Lion I can.
459 2012-05-25 18:01:30 <GTRsdk> oh okay
460 2012-05-25 18:02:08 <luke-jr> splatster: does it support detached signatures?
461 2012-05-25 18:02:23 <splatster> Don't think so, no.
462 2012-05-25 18:02:28 <luke-jr> that is, a separate signature file
463 2012-05-25 18:02:33 <luke-jr> from the .app
464 2012-05-25 18:02:46 <luke-jr> or heck, even a separate signature file *in* the .app
465 2012-05-25 18:02:57 <splatster> The signature is generated at build time.
466 2012-05-25 18:03:01 <luke-jr> splatster: how?
467 2012-05-25 18:03:18 <splatster> Not sure how it signs, but if you change the app the sig will be invalid.
468 2012-05-25 18:03:37 <luke-jr> splatster: can you make an example "signed" and "unsigned" pair?
469 2012-05-25 18:03:44 <luke-jr> identical except for the signature?
470 2012-05-25 18:04:08 <splatster> The app doesn't -need- to be signed to work in OS X 10.8
471 2012-05-25 18:04:54 <luke-jr> vragnaroda: can you test http://luke.dashjr.org/tmp/code/bin-osx-bitcoind-v0.6.2.2.bz2
472 2012-05-25 18:05:07 <splatster> luke-jr: I can.
473 2012-05-25 18:05:11 <luke-jr> splatster: it'd be nice to have, though, if it enforces by default
474 2012-05-25 18:05:17 <luke-jr> splatster: you have what OSX ver?
475 2012-05-25 18:05:24 <splatster> 10.7
476 2012-05-25 18:06:03 <luke-jr> ok, but I already can test on 10.7 myself :P
477 2012-05-25 18:06:07 <splatster> ok
478 2012-05-25 18:06:24 <splatster> luke-jr: Are you on OS X right now?
479 2012-05-25 18:06:48 <luke-jr> splatster: I have a work laptop with OS X
480 2012-05-25 18:06:54 <luke-jr> splatster: I am building the binaries on Ubuntu
481 2012-05-25 18:07:17 <luke-jr> gavinandresen: what do you think about making -h work? :P
482 2012-05-25 18:07:58 <splatster> luke-jr: Go into Finder and right click any Apple or App Store app and go to show package contents, inside there you will see a folder called "_CodeSignature" which is what you're after
483 2012-05-25 18:08:18 <luke-jr> splatster: OK, so we need to find a way to add that to an existing app
484 2012-05-25 18:08:33 <luke-jr> splatster: can you research that? maybe Xcode makes logs of the command to sign?
485 2012-05-25 18:08:46 <vragnaroda> terminate called after throwing an instance of 'boost::interprocess::lock_exception' what():  boost::interprocess::lock_exception
486 2012-05-25 18:08:48 <splatster> Yup, lemme go see
487 2012-05-25 18:08:49 <vragnaroda> Abort trap
488 2012-05-25 18:09:00 <vragnaroda> (From the usage again)
489 2012-05-25 18:09:05 <luke-jr> vragnaroda: how about without --help?
490 2012-05-25 18:09:16 <luke-jr> I didn't touch boost yet, was hoping bdb works now
491 2012-05-25 18:09:23 <vragnaroda> -daemon seems to be working.
492 2012-05-25 18:09:27 <luke-jr> hmm
493 2012-05-25 18:09:32 <luke-jr> FWIW, I use -printtoconsole for testing
494 2012-05-25 18:09:55 <luke-jr> so it seems 10.6 doesn't support pthreads or something
495 2012-05-25 18:10:18 <splatster> luke-jr: Nothing about the code signing jumps out at me from the build logs.
496 2012-05-25 18:12:59 <vragnaroda> luke-jr: $ bitcoind -rpcuser=bitcoinrpc -rpcpassword=3NrBmJpMG65FgMezs3FZEfnWezxZDWieWVCtV3yQCLsS getblockcount
497 2012-05-25 18:13:03 <vragnaroda> 18518
498 2012-05-25 18:13:07 <luke-jr> splatster: maybe check into http://stackoverflow.com/questions/6896029/re-sign-ipa-iphone
499 2012-05-25 18:13:09 <vragnaroda> Agh
500 2012-05-25 18:13:14 <luke-jr> vragnaroda: fail
501 2012-05-25 18:13:16 <vragnaroda> Eh, whatever, it seems to be working.
502 2012-05-25 18:13:24 <luke-jr> vragnaroda: you posted your password&
503 2012-05-25 18:14:05 <vragnaroda> luke-jr: A password on a system that doesn't have a useful wallet.  (But, no, I didn't mean to copypaste that.)
504 2012-05-25 18:15:21 <luke-jr> vragnaroda: even on Linux and 10.7, --help segfaults, so maybe that's not a 10.6-specific problem
505 2012-05-25 18:15:28 <vragnaroda> luke-jr: But, anyhoo, it's running, seemingly without problems.
506 2012-05-25 18:15:30 <vragnaroda> Oh, lol
507 2012-05-25 18:17:31 <vragnaroda> (and the RPC password has been changed now, anyway :p)
508 2012-05-25 18:18:03 <vragnaroda> ((I can't believe I did that (even though it was pretty much useless).))
509 2012-05-25 18:38:29 <luke-jr> ca0816152d91c929b51b2f644fcb9a69797a49fa is the first bad commit
510 2012-05-25 18:38:37 <luke-jr> sipa: ^ why v0.6.2.2 segfaults on --help :P
511 2012-05-25 18:44:16 <luke-jr> hmm, master is fine tho
512 2012-05-25 18:44:28 <luke-jr> guess I need to dig to figure out which backport to prioritize XD
513 2012-05-25 18:52:55 <Joric> ThomasV, i'm about to release an electrum plugin for both browsers ff: http://goo.gl/stgmY chrome: http://goo.gl/1Y7GW
514 2012-05-25 18:53:19 <ThomasV> cool
515 2012-05-25 18:54:10 <ThomasV> how does it work?
516 2012-05-25 18:54:18 <Joric> just kidding hell knows how much time it will really take
517 2012-05-25 18:54:36 <Joric> i'm planning only http so far
518 2012-05-25 18:54:42 <ThomasV> heh, you got me :)
519 2012-05-25 18:55:22 <Joric> but really i learned a lot about browser extensions )
520 2012-05-25 18:56:17 <ThomasV> what did you learn?
521 2012-05-25 18:57:28 <Joric> oh there's a lot of details
522 2012-05-25 18:59:25 <Joric> found out ff plugins don't need restart now if you use addon-sdk (aka jetpack)
523 2012-05-25 19:00:11 <Joric> it's all xmlhttprequest / post at 8081 for now not sure i'll ever get to tcp sockets
524 2012-05-25 19:00:32 <Joric> how many servers currently support http?
525 2012-05-25 19:00:40 <Diablo-D3> Joric: a lot of plugins are using that now
526 2012-05-25 19:01:25 <ThomasV> Joric: all of them
527 2012-05-25 19:01:47 <ThomasV> but I guess I should add supports for websockets
528 2012-05-25 19:02:15 <ThomasV> it would probably be faster
529 2012-05-25 19:03:28 <Joric> ThomasV, did you get to proper json yet?
530 2012-05-25 19:03:48 <ThomasV> for what?
531 2012-05-25 19:04:09 <Joric> JSON.parse just says invalid character '
532 2012-05-25 19:04:12 <Joric> no kidding
533 2012-05-25 19:04:20 <ThomasV> for the wallet file?
534 2012-05-25 19:04:33 <Joric> yeah i'm using a beautified version for now
535 2012-05-25 19:04:57 <ThomasV> I did not change its format
536 2012-05-25 19:05:12 <ThomasV> you said you were going to use ff's storage
537 2012-05-25 19:05:58 <luke-jr> ThomasV: ' isn't a valid quoting character in JSON
538 2012-05-25 19:06:55 <ThomasV> luke-jr: no it isn't
539 2012-05-25 19:07:12 <Joric> yeah but i'm not sure about backward compatibility
540 2012-05-25 19:07:54 <ThomasV> how so?
541 2012-05-25 19:08:26 <Joric> didn't check if ast.literal_eval is able to parse a proper json
542 2012-05-25 19:09:40 <ThomasV> that's a minor issue. I can use json.loads if necessary
543 2012-05-25 19:11:12 <Joric> i decided to mimic the original ui will use the same i18n strings
544 2012-05-25 19:11:55 <ThomasV> so your screenshot is your current ui?
545 2012-05-25 19:12:19 <Joric> yes it's all real, semi-working extentions )
546 2012-05-25 19:12:32 <ThomasV> cool
547 2012-05-25 19:13:45 <Joric> *extensions, one of them is, actually, an add-on )
548 2012-05-25 19:14:25 <Joric> it's all twitter bootstrap got used to it heard some ppl already hate it
549 2012-05-25 19:15:17 <Joric> there are three parts, popup window + options window + context menu
550 2012-05-25 19:15:53 <Joric> could potentially get more auditory than satoshi's client ;)
551 2012-05-25 19:16:28 <graingert> ThomasV: don't use WebSockets when EventSource will do better
552 2012-05-25 19:17:00 <ThomasV> graingert: what is that?
553 2012-05-25 19:17:03 <graingert> http://www.w3.org/TR/eventsource/
554 2012-05-25 19:17:13 <graingert> To enable servers to push data to Web pages over HTTP or using   dedicated server-push protocols, this specification introduces the   EventSource interface.
555 2012-05-25 19:17:55 <graingert> http://www.html5rocks.com/en/tutorials/eventsource/basics/
556 2012-05-25 19:18:20 <ThomasV> ic
557 2012-05-25 19:19:05 <ThomasV> .. moar work for me :_/
558 2012-05-25 19:19:14 <graingert> less than WebSockets
559 2012-05-25 19:19:56 <ThomasV> really? cool
560 2012-05-25 19:20:06 <ThomasV> websockets are easy, iirc
561 2012-05-25 19:20:48 <Joric> http is fine i'm mostly terrified by tcp sockets didn't really plan those in js
562 2012-05-25 19:21:19 <Joric> chrome uses 'experemental' api for those, it's not even enabled by default
563 2012-05-25 19:22:36 <graingert> also WebSockets basically needs SSL
564 2012-05-25 19:27:35 <graingert> ThomasV: websockets are for games
565 2012-05-25 19:27:43 <ThomasV> ok
566 2012-05-25 19:27:54 <graingert> where you really need two way coms
567 2012-05-25 19:28:16 <graingert> also you'll need an http server built for handling lots of long connections using event source
568 2012-05-25 19:39:27 <luke-jr> vragnaroda: found the segfault issue, and it does seem likely that it's related to your other error on --help too
569 2012-05-25 19:39:39 <vragnaroda> Interesting.
570 2012-05-25 19:39:41 <luke-jr> vragnaroda: no other issues?
571 2012-05-25 19:40:42 <vragnaroda> luke-jr: Seems to be working fine right now.  It's still running.
572 2012-05-25 19:41:10 <luke-jr> great
573 2012-05-25 19:41:15 <luke-jr> now to see if I can get Qt to build <.<
574 2012-05-25 21:19:57 <splatster> .isdown blockexplorer.com
575 2012-05-25 21:20:07 <pebbles> splatster: http://blockexplorer.com Is Down
576 2012-05-25 21:28:11 <Joric> why this tx is still not relayed ? https://blockchain.info/address/1JA3eXMdTwoZBvoK4goqbTF54mFSV8C3h2
577 2012-05-25 21:28:15 <Joric> any ideas?
578 2012-05-25 21:32:45 <gmaxwell> Joric: probably because one of its ancestors is not related.
579 2012-05-25 21:32:49 <gmaxwell> re relayed.
580 2012-05-25 21:41:01 <luke-jr> )%*#(%#
581 2012-05-25 21:41:09 <luke-jr> Qt doesn't have official 32-bit builds anymore
582 2012-05-25 21:41:18 <luke-jr> and building it for Mac requires ObjC compiler -.-
583 2012-05-25 22:15:10 <luke-jr> all dependencies except Qt now gitian-built for OS X
584 2012-05-25 22:18:57 <GTRsdk> Are you using MacPorts to get the dependencies built?
585 2012-05-25 22:19:07 <GTRsdk> or manually downloading and compiling?
586 2012-05-25 22:19:35 <devrandom> luke-jr: nice
587 2012-05-25 22:21:45 <graingert> Joric: this address is very busy https://blockchain.info/address/19owJPyuY3AYwPPFcXgR65DWfA6FgLg1ij
588 2012-05-25 22:23:59 <denisx> damn apple kicked my pool stats app out off the appstore
589 2012-05-25 22:24:54 <Joric> denisx, 'Apps must comply with all legal requirements in any location where they are made available to users.' ??
590 2012-05-25 22:25:10 <denisx> Joric: yeah, something like that
591 2012-05-25 22:27:15 <gmaxwell> nevermind the fact that there is no sane way to say that a pool stats app is any more unlawful anywhere than anything else in their appstore... but this is why you shouldn't waste your resources developing for that platform.
592 2012-05-25 22:28:39 <Joric> that is what apple wrote about blockchain.info app i got a slightly different reply
593 2012-05-25 22:29:06 <GTRsdk> get it into a Cydia repo
594 2012-05-25 22:30:35 <Joric> one of my games was banned from appstore after about two months
595 2012-05-25 22:34:56 <denisx> I would say the day in summer 2008 when the appstore opened was peak apple
596 2012-05-25 22:42:29 <luke-jr> GTRsdk: manually building. does MacPorts support Ubuntu?
597 2012-05-25 22:42:52 <luke-jr> gmaxwell: PM?
598 2012-05-25 22:43:43 <sipa> luke-jr: Ubuntu? i thought it was for OSX?
599 2012-05-25 22:43:58 <luke-jr> sipa: yes, but Gitian runs on Ubuntu
600 2012-05-25 22:44:10 <luke-jr> sipa: as does all my OSX gitian stuff
601 2012-05-25 22:44:41 <sipa> i don't understand your question; why do you want macports?
602 2012-05-25 22:51:19 <luke-jr> sipa: I don't think I do. GTRsdk was suggesting it.
603 2012-05-25 22:51:27 <GTRsdk> Joric: do you have them in a Cydia repo?
604 2012-05-25 22:51:48 <GTRsdk> luke-jr: just out of curiosity... how are you building OS X apps on Ubuntu?
605 2012-05-25 22:52:12 <luke-jr> GTRsdk: cross-compiler
606 2012-05-25 22:53:14 <GTRsdk> Did you get the compiler from XCode?
607 2012-05-25 22:53:15 <Joric> GTRsdk, do ppl buy cydia apps?
608 2012-05-25 22:53:46 <GTRsdk> Joric: if you charge money for it, yes. If you don't, then no, but if you have a donate link, maybe. If not, then no.
609 2012-05-25 22:54:28 <luke-jr> GTRsdk: Xcode 3.1.2
610 2012-05-25 22:55:08 <GTRsdk> Cool. I didn't know that was possible.
611 2012-05-25 22:55:14 <luke-jr> ie, the last version for 10.5
612 2012-05-25 22:55:37 <luke-jr> GTRsdk: my stuff is at https://gitorious.org/cross-osx
613 2012-05-25 22:56:09 <GTRsdk> Actually, 3.1.4 was the last version for 10.5 (installed on my PowerPC 10.5 machine), but I don't think there was anythign new in it...
614 2012-05-25 22:56:31 <luke-jr> hmm
615 2012-05-25 22:58:09 <sipa> later xcode doesn't run on non-osx?
616 2012-05-25 22:59:51 <luke-jr> sipa: no Xcode runs on non-OSX; Xcode is needed for OSX libs to link to
617 2012-05-25 23:00:06 <luke-jr> since it's dynamic linking, using 10.5 should be best compat
618 2012-05-25 23:00:18 <luke-jr> I did try newer GCCs, but they didn't build for me
619 2012-05-25 23:00:25 <sipa> ok
620 2012-05-25 23:01:34 <luke-jr> sipa: curiously, vragnaroda discovered pthreads isn't really supported by OSX until 10.7 tho
621 2012-05-25 23:01:52 <luke-jr> sipa: I had to change bdb to build with some "gcc assembly" mutex instead
622 2012-05-25 23:02:16 <luke-jr> (via configure option)
623 2012-05-25 23:02:21 <sipa> ok...
624 2012-05-25 23:02:43 <luke-jr> :p
625 2012-05-25 23:03:32 <sipa> i have no intent to venture into apple's internal dungeons, but if you can succeed in buildimg bitcoin-qt deterministically in gitian... great
626 2012-05-25 23:18:27 <sipa> ah, finally... some ipv6 nodes in seeds.txt!
627 2012-05-25 23:19:01 <luke-jr> curiously, some bithasher.com pool that's been unmaintained for a year is running a very old 0.3 branch of mine
628 2012-05-25 23:19:38 <Diablo-D3> lol
629 2012-05-25 23:20:01 <Ukto> heh
630 2012-05-25 23:20:16 <Ukto> guess I can throw up an ipv4/ipv6 client
631 2012-05-25 23:20:42 <luke-jr> sipa: is it on tor?
632 2012-05-25 23:20:48 <sipa> yes!
633 2012-05-25 23:20:54 <sipa> untested, however
634 2012-05-25 23:21:55 <luke-jr> sipa: there is at least one hidden service running by RedEmerald
635 2012-05-25 23:22:09 <sipa> sure, i have one too
636 2012-05-25 23:22:17 <luke-jr> so why aren't they listed yet? :p
637 2012-05-25 23:22:30 <sipa> oh, but it is not advertizing itself
638 2012-05-25 23:22:46 <sipa> i should update
639 2012-05-25 23:23:48 <sipa> as ipv6 nodes prefer relaying ipv6 addresses, and onion nodes prefer relaying onion addresses, you need some critical mass in each network before you'll actually encounter new nodes
640 2012-05-25 23:24:50 <luke-jr> how are hidden services listed in DNS O.o?
641 2012-05-25 23:24:59 <luke-jr> SRV records or something?
642 2012-05-25 23:25:29 <sipa> not exposed via DNS at all, for now
643 2012-05-25 23:25:48 <sipa> maybe i can encode them onioncat-style in a AAAA record
644 2012-05-25 23:26:10 <sipa> that would actually work in bitcoin now
645 2012-05-25 23:26:26 <luke-jr> but it'd break normal IPv6 resolvers
646 2012-05-25 23:26:31 <sipa> why?
647 2012-05-25 23:26:42 <luke-jr> because they can't reach them?
648 2012-05-25 23:27:07 <sipa> resolvers don't need to reach any hosts, only nameservers
649 2012-05-25 23:27:20 <luke-jr> I mean nodes resolving the DNS seed :p
650 2012-05-25 23:28:01 <luke-jr> I guess it'd be OK so long as you ensure there's normal AAAA mixed in
651 2012-05-25 23:28:17 <luke-jr> otoh&
652 2012-05-25 23:28:19 <sipa> every bitcoin node capable of reaching ipv6 has a special casing for treating the onioncat subnet as a separate network
653 2012-05-25 23:28:24 <luke-jr> does it make sense for Tor nodes to use DNS seeds?
654 2012-05-25 23:28:32 <sipa> good point!
655 2012-05-25 23:28:43 <sipa> it doesn't, i'd say
656 2012-05-25 23:29:10 <luke-jr> just a few hard-coded hidden services should work just the same for tor imo
657 2012-05-25 23:29:36 <sipa> maybe i'll add p2p listening code, which only answers getaddr requests and disconnects
658 2012-05-25 23:29:54 <sipa> though that would increase dos attack surface
659 2012-05-25 23:35:13 <gmaxwell> 18:28 < sipa> it doesn't, i'd say
660 2012-05-25 23:35:28 <gmaxwell> well, what happens when we hard code a couple onions and they're all down?
661 2012-05-25 23:35:50 <gmaxwell> DNS has some robusness properties that hiddenservices lack
662 2012-05-25 23:36:35 <luke-jr> gmaxwell: DNS servers can be down too
663 2012-05-25 23:36:45 <luke-jr> I'd think Tor should support anycast also