1 2011-09-30 01:00:58 <shadders> TD: is wallet the only thing that uses java serialization now?  I can't find any other references to ObjectOutputStream
  2 2011-09-30 01:41:16 <luke-jr> louigi: are there coins in the wallet?
  3 2011-09-30 01:41:32 <louigi> hey guys!  I am on Ubuntu, used the stable version, the 0.3 one since July, several days ago blockcount stopped updating. Today tried running 0.4 - still not updating. Any advice?
  4 2011-09-30 01:41:37 <louigi> luke-jr, yep
  5 2011-09-30 01:41:43 <louigi> not a lot though - 1!
  6 2011-09-30 01:42:17 <luke-jr> louigi: do you have a MtGox account?
  7 2011-09-30 01:42:57 <luke-jr> also, what OS?
  8 2011-09-30 01:43:18 <louigi> luke-jr, I do have an mt gox account though I did not use it yet
  9 2011-09-30 01:43:23 <louigi> OS Ubuntu
 10 2011-09-30 01:43:36 <louigi> (it is in the first request there ;))
 11 2011-09-30 01:44:03 <luke-jr> louigi: could you 1) tar up your ~/.bitcoin directory, 2) send the 1 BTC to your MtGox, 3) post the (compressed) tar somewhere public
 12 2011-09-30 01:44:05 <luke-jr> please
 13 2011-09-30 01:44:26 <luke-jr> be sure bitcoin is not running before step 1
 14 2011-09-30 01:44:38 <luke-jr> (I assume you have already tried restarting it)
 15 2011-09-30 01:44:47 <louigi> luke-jr, you mean so that you can look at the wallet and it is empty?
 16 2011-09-30 01:45:02 <louigi> luke-jr, yeah, restarted it many times during several days
 17 2011-09-30 01:45:07 <luke-jr> louigi: I mean so that if someone you don't trust downloads it, they don't steal your 1 BTC ;)
 18 2011-09-30 01:45:29 <luke-jr> (including me, since you have no reason to trust me)
 19 2011-09-30 01:45:41 <louigi> luke-jr, k man, I'll send it! I hope this works too
 20 2011-09-30 01:45:42 <gmaxwell> If the blockcount isn't updating, I'd wonder if he's even connected at all?
 21 2011-09-30 01:45:52 <louigi> gmaxwell, it does show 8 connections
 22 2011-09-30 01:46:01 <luke-jr> gmaxwell: he's not the first time I've heard of this problem
 23 2011-09-30 01:46:13 <louigi> 8 connections is the usual number I have
 24 2011-09-30 01:46:26 <gmaxwell> luke-jr: Well, if it's connected and stuck my next guess would be that his blockchain has a data error and no future blocks will validate against it.
 25 2011-09-30 01:46:28 <luke-jr> gmaxwell: if I can reproduce it, I plan to bisect
 26 2011-09-30 01:46:55 <gmaxwell> But yea, if it's actually wallet based that would be interesting.
 27 2011-09-30 01:47:58 <gmaxwell> if it is just blockchain corruption then the send will work fine, though he won't see it confirm locally.
 28 2011-09-30 01:48:07 <luke-jr> louigi: btw, you didn't mention: which 0.3 version were you using before?
 29 2011-09-30 01:48:51 <louigi> luke-jr, sec, will tell you (got 0.4 running now)
 30 2011-09-30 01:49:37 <luke-jr> louigi: please shutdown 0.4, wait 2 minutes, and backup ~/.bitcoin, before doing anything more (like starting the old version again)
 31 2011-09-30 01:49:42 <louigi> luke-jr, bitcoin-0.3.24
 32 2011-09-30 01:49:45 <luke-jr> k
 33 2011-09-30 01:50:07 <luke-jr> the key to fixing this IMO is reproducing it right now
 34 2011-09-30 01:50:25 <louigi> k, I sent the bitcoin to my bitmarket.eu account, dunno if it actually arrives there
 35 2011-09-30 01:50:47 <gmaxwell> louigi: I think you have good odds of it arriving.
 36 2011-09-30 01:51:06 <louigi> k guys, shut down program, waiting a couple of minutes
 37 2011-09-30 01:51:08 <gmaxwell> (though you should leave bitcoin running for a few minutes after sending!)
 38 2011-09-30 01:51:11 <gmaxwell> doh.
 39 2011-09-30 01:51:49 <imsaguy> check bitmarket.eu address against bitcoincharts pending transactions
 40 2011-09-30 01:51:50 <gmaxwell> well, it's probably okay.
 41 2011-09-30 01:51:54 <luke-jr> ^
 42 2011-09-30 01:52:05 <luke-jr> fwiw, I didn't see it in #bitcoin-watch
 43 2011-09-30 01:52:58 <luke-jr> louigi: if you want to ensure you only send me the backup, I'll personally guarantee I don't steal it at least :P
 44 2011-09-30 01:53:16 <louigi> okay
 45 2011-09-30 01:53:27 <louigi> not much to steal there anyway
 46 2011-09-30 01:53:30 <luke-jr> heh
 47 2011-09-30 01:53:37 <luke-jr> how big is it, all said and done?
 48 2011-09-30 01:54:49 <louigi> still adding to archive
 49 2011-09-30 01:56:36 <louigi> luke-jr, 518 Mb
 50 2011-09-30 01:57:38 <luke-jr> o.o
 51 2011-09-30 01:57:47 <luke-jr> use this then: http://luke.dashjr.org/tmp/code/upload-to-me.php
 52 2011-09-30 01:57:50 <luke-jr> actually
 53 2011-09-30 01:57:54 <luke-jr> 518 MB is too big for that too
 54 2011-09-30 01:58:03 <imsaguy> split it
 55 2011-09-30 01:58:07 <luke-jr> louigi:  have an apache installed?
 56 2011-09-30 01:58:16 <louigi> luke-jr, I have a site where to put it, but later on today
 57 2011-09-30 01:58:20 <imsaguy> torrent it
 58 2011-09-30 01:58:21 <luke-jr> hmm
 59 2011-09-30 01:58:24 <louigi> luke-jr, actually yeah
 60 2011-09-30 01:58:28 <louigi> apache is running
 61 2011-09-30 01:58:39 <luke-jr> but you're behind a firewall/NAT
 62 2011-09-30 01:59:33 <louigi> luke-jr, I don't think so
 63 2011-09-30 01:59:37 <louigi> or am i
 64 2011-09-30 02:00:22 <luke-jr> louigi: cat yourfile > /dev/tcp/184.4.160.40/8337
 65 2011-09-30 02:00:41 <Diablo-D3> or just use netcat?
 66 2011-09-30 02:00:52 <louigi> guys, atm the internet uplink here is weak
 67 2011-09-30 02:01:07 <louigi> I can later come to my work and upload to my site
 68 2011-09-30 02:02:05 <louigi> EXCEPTION: 22DbRunRecoveryException
 69 2011-09-30 02:02:14 <louigi> this is what I get when running older client
 70 2011-09-30 02:02:42 <luke-jr> expected
 71 2011-09-30 02:03:26 <luke-jr> louigi: what if you just tar blk0001.dat and blkindex.dat ?
 72 2011-09-30 02:03:44 <louigi> luke-jr, sure
 73 2011-09-30 02:04:17 <luke-jr> also, what compressor are you using? xz is the best these days
 74 2011-09-30 02:04:25 <louigi> luke-jr, hehe, they both together are 600Mb
 75 2011-09-30 02:04:32 <luke-jr> louigi: compressed?
 76 2011-09-30 02:04:33 <louigi> dunno if it will change things much
 77 2011-09-30 02:04:34 <Diablo-D3> luke-jr: not even close
 78 2011-09-30 02:04:35 <Diablo-D3> lrzip is the best
 79 2011-09-30 02:05:32 <luke-jr> Diablo-D3: promoting your competition?
 80 2011-09-30 02:05:52 <Diablo-D3> you do know I've known con for years, right?
 81 2011-09-30 02:06:05 <Diablo-D3> the only reason he knews bitcoin exists is because I told him about it
 82 2011-09-30 02:06:30 <Diablo-D3> Ive used bfs for as long as hes been doing it as well
 83 2011-09-30 02:06:38 <Diablo-D3> luke-jr: you are so noob it hurts.
 84 2011-09-30 02:06:58 <louigi> luke-jr, hehe, 511 Mb
 85 2011-09-30 02:07:24 <louigi> luke-jr, I tell you what - I will send the file later on today
 86 2011-09-30 02:07:30 <louigi> will come here and leave you a link
 87 2011-09-30 02:07:36 <louigi> anyone of the devs can watch
 88 2011-09-30 02:07:54 <louigi> I did read about complaints before on forums but simply restarting did not help
 89 2011-09-30 02:08:29 <louigi> I can reinstall, but my linux mentality tells me - help the devs to fix it :)))
 90 2011-09-30 02:08:48 <imsaguy> louigi, wait a minute
 91 2011-09-30 02:08:50 <imsaguy> major netsplit
 92 2011-09-30 02:08:59 <louigi> weird
 93 2011-09-30 02:09:04 <imsaguy> hold on
 94 2011-09-30 02:13:34 <luke-jr> louigi: poke
 95 2011-09-30 02:13:40 <louigi> luke-jr, yo
 96 2011-09-30 02:13:51 <louigi> so I was saying
 97 2011-09-30 02:13:51 <luke-jr> how's it?
 98 2011-09-30 02:13:53 <louigi> I can reinstall, but my linux mentality tells me - help the devs to fix it :)))
 99 2011-09-30 02:14:01 <louigi> 511 Mb
100 2011-09-30 02:14:07 <louigi> luke-jr, I tell you what - I will send the file later on today
101 2011-09-30 02:14:12 <louigi> will come here and leave you a link
102 2011-09-30 02:14:16 <louigi> anyone of the devs can watch
103 2011-09-30 02:14:21 <louigi> I did read about complaints before on forums but simply restarting did not help
104 2011-09-30 02:14:25 <louigi> I can reinstall, but my linux mentality tells me - help the devs to fix it :)))
105 2011-09-30 02:14:40 <luke-jr> I'm so impatient. XD
106 2011-09-30 02:20:25 <louigi> me too, tbh. i want my client to work )))
107 2011-09-30 02:20:27 <louigi> but!
108 2011-09-30 02:20:31 <louigi> it is early morning here
109 2011-09-30 04:23:35 <AlexWaters> so when a pull becomes un-automatically mergable - it is because of conflicts in the commit?
110 2011-09-30 04:23:46 <AlexWaters> and, this can be remedied by a 'rebase'?
111 2011-09-30 04:23:58 <luke-jr> &
112 2011-09-30 04:24:11 <luke-jr> in short, yes
113 2011-09-30 04:24:31 <phantomcircuit> for the long version send milk and cookies
114 2011-09-30 04:25:05 <luke-jr> ^
115 2011-09-30 04:25:06 <AlexWaters> i'm trying to write a general purpose closure warning message. to kill off abandoned pulls, any recommendations?
116 2011-09-30 04:26:03 <luke-jr> "We're a bunch of slow guys who took too long to look at your code. In the meantime, we made the codebase conflict with it. You're our only hope! Rebase or we'll be forced to go on in ignorance forever!"
117 2011-09-30 04:26:07 <luke-jr> <.<
118 2011-09-30 04:26:42 <AlexWaters> luke-jr: i'm trying to target the pulls that are unpopular, that people don't want merged
119 2011-09-30 04:26:42 <phantomcircuit> luke-jr, lol'd
120 2011-09-30 04:26:47 <AlexWaters> and they just sit there collecting dust
121 2011-09-30 04:27:29 <luke-jr> AlexWaters: then "We really don't want your help. Go away. If you want to persist, rewrite it with our new codebase."
122 2011-09-30 04:27:35 <AlexWaters> lol
123 2011-09-30 04:27:38 <phantomcircuit> i think you'll find that most of those are simple patches which should have been merged but now require a rebase
124 2011-09-30 04:28:05 <phantomcircuit> there isn't overwhelming support for them probably because they scratch an itch
125 2011-09-30 04:32:31 <AlexWaters> https://gist.github.com/1252874 - if anyone has advice for better wording on my pull closure wording message
126 2011-09-30 04:32:42 <AlexWaters> warning*
127 2011-09-30 04:34:51 <AlexWaters> does anyone know if notifications are sent out whenever someone edits a comment on a pull? I hate to be flooding people every time I edit my comments with my OCD
128 2011-09-30 04:37:05 <luke-jr> I'd s/automatically/without conflicts
129 2011-09-30 04:37:21 <wumpus> no, notifications are not sent on edits
130 2011-09-30 04:37:25 <luke-jr> your semicolon should be a comma
131 2011-09-30 04:37:26 <wumpus> only when you post the message
132 2011-09-30 04:37:40 <AlexWaters> ok, thanks guys
133 2011-09-30 04:37:40 <luke-jr> s/to be mergable/to merge cleanly
134 2011-09-30 04:38:02 <wumpus> and remember that the original message is in the mail, so sometimes people don't see your edits at all :p
135 2011-09-30 04:38:41 <AlexWaters> yeah, I kind of like that - so that i can tell the participants something without it being very public after editing
136 2011-09-30 04:38:44 <AlexWaters> like, good job etc
137 2011-09-30 04:38:44 <wumpus> (especiallyl if they reply to the mail)
138 2011-09-30 04:38:52 <wumpus> right
139 2011-09-30 05:25:47 <CIA-101> bitcoin: Con Kolivas * r3e527c6af5b0 cgminer/main.c: Retry pools after a delay of 15 seconds if none can be contacted on startup unless a key is pressed.
140 2011-09-30 05:58:03 <vegard> gmaxwell: so global, decentralised, micro-payment system seems impossible
141 2011-09-30 09:04:47 <grondilu> Is encryption support enabled by default on the bitcoind command line tool?
142 2011-09-30 09:05:03 <grondilu> (I hopre not)
143 2011-09-30 09:07:12 <louigi> luke-jr, yo
144 2011-09-30 09:07:37 <sipa> grondilu: no
145 2011-09-30 09:07:51 <sipa> well, support is enabled, but it isn't used for new wallets by default
146 2011-09-30 09:08:14 <grondilu> ok
147 2011-09-30 09:09:36 <louigi> luke-jr, i am now uploading the file
148 2011-09-30 09:09:44 <louigi> btw, transaction did suceed
149 2011-09-30 09:09:49 <louigi> succeed
150 2011-09-30 09:10:32 <grondilu> Is the 'import key from PEM file' in the main branch yet?
151 2011-09-30 09:11:31 <sipa> no
152 2011-09-30 09:11:37 <grondilu> also:  are we going to have a manual page someday?
153 2011-09-30 09:12:00 <sipa> if someone writes one :)
154 2011-09-30 09:12:55 <grondilu> ok, thanks
155 2011-09-30 09:45:27 <louigi> luke-jr, www.disc-shelf.com/bitcoin/bitcoin.zip
156 2011-09-30 09:45:32 <louigi> and anyone interested
157 2011-09-30 09:45:43 <louigi> this is a zip file of my bitcoin client data
158 2011-09-30 09:45:52 <louigi> my client stopped updating block count
159 2011-09-30 09:46:00 <louigi> so if any of the devs are interested to take a look
160 2011-09-30 09:46:04 <louigi> you are welcome
161 2011-09-30 09:46:08 <louigi> the zip is 518 Mb
162 2011-09-30 09:46:19 <louigi> I am on Ubuntu 10.04
163 2011-09-30 09:46:25 <louigi> version of client 0.40
164 2011-09-30 09:46:49 <tcatm> louigi: try to delete addr.dat
165 2011-09-30 09:46:52 <louigi> before that used the stable 0.3.24
166 2011-09-30 09:47:01 <louigi> tcatm, thanks, will try now
167 2011-09-30 09:57:05 <lfm> ;;bc,blocks
168 2011-09-30 09:57:06 <gribble> 147508
169 2011-09-30 09:57:43 <diki> do you guys know
170 2011-09-30 09:57:57 <diki> if google charges the SMS verifications they send upon password recovery?
171 2011-09-30 09:58:09 <diki> i am not from the US but in Europe
172 2011-09-30 09:58:22 <diki> and while i did get the SMS fast, i dunno if it costed ME something
173 2011-09-30 09:58:51 <lfm> diki prolly depends on your phone company
174 2011-09-30 10:02:26 <MacRohard> they're free i think
175 2011-09-30 10:02:52 <MacRohard> unless you have some ridiculous plan that charges for inbound texts
176 2011-09-30 10:03:07 <diki> lfm:well they do rip us off but..
177 2011-09-30 10:03:19 <lfm> or international texts or something rediculous
178 2011-09-30 10:03:58 <louigi> tcatm, did not help
179 2011-09-30 10:04:21 <lfm> louigi: how many connections do you have?
180 2011-09-30 10:04:29 <louigi> lfm, 8
181 2011-09-30 10:04:30 <CIA-101> bitcoin: Gavin Andresen master * rf4769e4 / (6 files in 2 dirs):
182 2011-09-30 10:04:41 <louigi> lfm, always had 8
183 2011-09-30 10:04:44 <louigi> never more
184 2011-09-30 10:04:45 <lfm> and what block number do you have?
185 2011-09-30 10:04:49 <louigi> it always workd fine
186 2011-09-30 10:04:56 <louigi> 140757
187 2011-09-30 10:05:07 <louigi> 7000 late )
188 2011-09-30 10:05:21 <lfm> thats like a month
189 2011-09-30 10:05:31 <lfm> almos6t 2 months late
190 2011-09-30 10:05:56 <lfm> ;;bc.blocks
191 2011-09-30 10:05:57 <gribble> Error: "bc.blocks" is not a valid command.
192 2011-09-30 10:06:03 <lfm> ;;bc,blocks
193 2011-09-30 10:06:04 <gribble> 147508
194 2011-09-30 10:06:54 <louigi> lfm, well, I think it did update things in August
195 2011-09-30 10:07:02 <louigi> I did have my computer freeze
196 2011-09-30 10:07:23 <louigi> and as far as I understand, if I just delete the dat file, it will redownload all blocks from the start - right?
197 2011-09-30 10:07:37 <louigi> luke-jr here guessed there could be some error in blocks
198 2011-09-30 10:07:43 <louigi> so they do not get downloaded anymore
199 2011-09-30 10:07:57 <lfm> you gotta delete blk0001.dat AND blkindex.dat
200 2011-09-30 10:07:57 <louigi> I know this - other ppl also have this problem and it would be nice if it gets figured out
201 2011-09-30 10:08:02 <louigi> yeah
202 2011-09-30 10:08:14 <louigi> that's a lot to redownload though...
203 2011-09-30 10:08:16 <louigi> I might try
204 2011-09-30 10:08:30 <lfm> you can get a restarter file quicker
205 2011-09-30 10:08:31 <louigi> trying
206 2011-09-30 10:08:39 <louigi> from the sourceforge?
207 2011-09-30 10:10:03 <lfm> not sure where the .zip is for the restart data off hand .... looking ....
208 2011-09-30 10:19:29 <louigi> lfm, I found some here: http://sourceforge.net/projects/bitcoin/files/Bitcoin/blockchain/
209 2011-09-30 10:21:24 <louigi> but it might remove my wallet which I do not want...
210 2011-09-30 10:23:08 <lfm> make a backup then
211 2011-09-30 10:23:32 <lfm> it should be just the two files tho
212 2011-09-30 10:24:40 <louigi> lfm, if I make a backup of the wallet
213 2011-09-30 10:24:45 <louigi> wallet.dat?
214 2011-09-30 10:24:48 <lfm> ya
215 2011-09-30 10:24:48 <louigi> something else?
216 2011-09-30 10:24:58 <louigi> and then after I download and get the block count
217 2011-09-30 10:25:00 <louigi> I can copy it in?
218 2011-09-30 10:25:02 <lfm> just need to save wallet.dat
219 2011-09-30 10:25:06 <louigi> k
220 2011-09-30 10:25:12 <louigi> will do that
221 2011-09-30 10:25:26 <louigi> but the link to the "broken" data files still makes sense
222 2011-09-30 10:25:32 <lfm> put the wallet back right aftyer you unpack the saved block chain
223 2011-09-30 10:25:43 <louigi> for you guys to see where the error occurs
224 2011-09-30 10:25:51 <louigi> ah, ok
225 2011-09-30 10:26:05 <lfm> louigi: naw we dont want your broken files, thanks anyway
226 2011-09-30 10:26:23 <louigi> luke-jr asked for them )
227 2011-09-30 10:26:33 <lfm> oh ok if he wants them
228 2011-09-30 10:35:38 <luke-jr> vegard: only with a public txn log
229 2011-09-30 10:36:36 <luke-jr> lfm: n00b
230 2011-09-30 10:36:49 <luke-jr> lfm: the goal is to FIX the problem, not get him back online :P
231 2011-09-30 10:37:19 <lfm> his machine crashed garbling the files. good luck figuring it out.
232 2011-09-30 10:38:08 <luke-jr> he said nothing about a crash
233 2011-09-30 10:38:27 <lfm> <louigi> lfm, well, I think it did update things in August
234 2011-09-30 10:39:44 <luke-jr> shrug
235 2011-09-30 10:39:52 <luke-jr> not hard to verify the data works in 0.3.24 still
236 2011-09-30 10:40:26 <lfm> he was like 7000 blocks behind current so I assumed the two were related
237 2011-09-30 11:28:20 <ThomasV> I just noticed that verifymessage works even for addresses that are not in my wallet. Does it lookup for the public key in the blockchain ?
238 2011-09-30 11:28:32 <ThomasV> or does it not need the pubkey?
239 2011-09-30 11:29:05 <lfm> it just tests the format of the address and the checksum which is part of the address
240 2011-09-30 11:29:40 <lfm> it doesn't need the pubkey
241 2011-09-30 11:29:41 <gmaxwell> lfm: he's asking about verify message.
242 2011-09-30 11:29:55 <gmaxwell> ThomasV: the signature includes the address, which is all it needs.
243 2011-09-30 11:30:09 <ThomasV> oh it's in the signature
244 2011-09-30 11:30:11 <ThomasV> ty
245 2011-09-30 11:30:17 <gmaxwell> Well, the public key isn't.
246 2011-09-30 11:30:24 <gmaxwell> But it's recovered from the signature.
247 2011-09-30 11:30:31 <gmaxwell> (using the address)
248 2011-09-30 11:31:14 <ThomasV> so, with signature and address, you can recover a pubkey ?
249 2011-09-30 11:31:28 <gmaxwell> (this is a property of ecc signatures given the data and the sig you can recover public keys from them, assuming you have a way of telling if the key is right)
250 2011-09-30 11:31:55 <gmaxwell> (and for us we can use the address to tell if the key is right, and the address is much smaller than the public key itself)
251 2011-09-30 11:31:59 <sipa> ThomasV: using the signature and the message, you can recover the pubkey
252 2011-09-30 11:32:09 <ThomasV> cool
253 2011-09-30 11:32:15 <sipa> but there may be up to 4 candidate pubkeys for a given signature/message combination
254 2011-09-30 11:32:16 <ThomasV> I'm learning everyday
255 2011-09-30 11:32:29 <sipa> so the signature contains two extra bits to tell which it is
256 2011-09-30 11:32:43 <ThomasV> I see
257 2011-09-30 11:32:59 <sipa> so only one recovery operation is necessary, and that is compared to the provided address
258 2011-09-30 11:33:29 <ThomasV> ecdsa is magic!
259 2011-09-30 11:34:19 <gavinandresen> Is there consensus on what format public keys should be in the RPC interface?  hex?  base58? base64 ?  I wrote an extension to validateaddress to dump the public key for an address...
260 2011-09-30 11:35:34 <ThomasV> gavinandresen, sipa : thanks for merging #524, great work!
261 2011-09-30 11:39:26 <sipa> gavinandresen: the sec standard uses hex for displaying public keys and other data structures
262 2011-09-30 11:40:07 <gavinandresen> sipa:  cool, thanks.
263 2011-09-30 11:41:31 <luke-jr> ;;bc,blocks
264 2011-09-30 11:41:32 <gribble> 147513
265 2011-09-30 11:58:00 <luke-jr> 0.3.24 and 0.3.23 can't resume this blockchain either
266 2011-09-30 13:16:50 <edcba> ;;bc,mtgox
267 2011-09-30 13:16:50 <gribble> {"ticker":{"high":4.999,"low":4.6742,"avg":4.820610517,"vwap":4.846178909,"vol":25619,"last":4.95,"buy":4.95,"sell":4.96}}
268 2011-09-30 13:22:30 <gavinandresen> Whoever is mining on the testnet:  thanks!
269 2011-09-30 13:22:56 <hippich> ;;bc,stats
270 2011-09-30 13:23:18 <hippich> ignores me :(
271 2011-09-30 13:23:29 <edcba> ;;bc,difficulty
272 2011-09-30 13:23:29 <gribble> Error: "bc,difficulty" is not a valid command.
273 2011-09-30 13:23:34 <edcba> ;;bc,help
274 2011-09-30 13:23:34 <gribble> Alias bc,24hprc, Alias bc,altprofit, Alias bc,avgprc, Alias bc,bcm, Alias bc,bitpenny, Alias bc,blocks, Alias bc,bounty, Alias bc,btceur, Alias bc,btcgbp, Alias bc,btcguild, Alias bc,btcrub, Alias bc,btcto, Alias bc,calc, Alias bc,calcd, Alias bc,channels, Alias bc,convert, Alias bc,deepbit, Alias bc,diff, Alias bc,diffchange, Alias bc,eligius, Alias bc,estimate, Alias bc,exchb, Alias bc,fx, (2 more messages)
275 2011-09-30 13:23:56 <edcba> ;;bc,help
276 2011-09-30 13:23:56 <gribble> Alias bc,24hprc, Alias bc,altprofit, Alias bc,avgprc, Alias bc,bcm, Alias bc,bitpenny, Alias bc,blocks, Alias bc,bounty, Alias bc,btceur, Alias bc,btcgbp, Alias bc,btcguild, Alias bc,btcrub, Alias bc,btcto, Alias bc,calc, Alias bc,calcd, Alias bc,channels, Alias bc,convert, Alias bc,deepbit, Alias bc,diff, Alias bc,diffchange, Alias bc,eligius, Alias bc,estimate, Alias bc,exchb, Alias bc,fx, (2 more messages)
277 2011-09-30 13:25:23 <hippich> ;;bc,diff
278 2011-09-30 13:40:44 <gavinandresen> sending/spending multisig transactions are working on the testnet: http://blockexplorer.com/testnet/block/0000000000cfe382f690777e3b52ea255717122465db6fcfd80efdf2d0080d8b
279 2011-09-30 13:42:32 <sipa> gavinandresen: nice!
280 2011-09-30 13:43:18 <gavinandresen> Only issues are with CBitcoinAddress/ExtractAddress... but that's something to tackle for a future release, I think
281 2011-09-30 13:43:44 <gavinandresen> (and what to show in listtransactions output )
282 2011-09-30 13:43:44 <luke-jr> gavinandresen: care to clarify the version issue?
283 2011-09-30 13:44:29 <gavinandresen> Hmm?  what version issue?  multisignature will be a two-stage pull:  first, get them relayed/accepted.  THen in a future release, add full support
284 2011-09-30 13:44:40 <luke-jr> gavinandresen: is mainline doing 0.4.1 or 0.5 next?
285 2011-09-30 13:44:49 <gavinandresen> oh, THAT version issue.
286 2011-09-30 13:44:56 <luke-jr> gavinandresen: your ML post said 0.5, but you committed 0.4.1
287 2011-09-30 13:45:21 <gavinandresen> It'll be 0.5.  But I don't want to bump to 0.5 until we have a release candidate.
288 2011-09-30 13:45:32 <luke-jr> O.o
289 2011-09-30 13:45:55 <luke-jr> maybe 0.4.99 in the meantime, like KDE does? :P
290 2011-09-30 13:46:21 <sipa> that's not unreasonable, i think
291 2011-09-30 13:46:30 <gavinandresen> Is that how KDE does it?  Release number versioning policy is something I should care more about, but just don't
292 2011-09-30 13:46:48 <luke-jr> well, KDE starts at like .98 for beta, then uses .99 for RCs
293 2011-09-30 13:47:05 <luke-jr> or something like that
294 2011-09-30 13:47:15 <luke-jr> maybe it's .8x for betas and .9x for RCs, I forget exactly
295 2011-09-30 13:48:01 <sipa> luke-jr: and then you'd use 0.4.{1-x} for the backport release?
296 2011-09-30 13:48:13 <luke-jr> sipa: right
297 2011-09-30 13:50:06 <genjix> major = big changes, minor = compatible changes, revisions = bug fixes
298 2011-09-30 13:50:30 <phantomcircuit> anybody know how to contact the guy who runs bitcoinity?
299 2011-09-30 13:50:33 <luke-jr> anyhow, why would a block get into mapBlockIndex, not be in mapOrphanBlocks, yet be higher than pindexBest?
300 2011-09-30 13:51:01 <genjix> luke-jr: define higher
301 2011-09-30 13:51:09 <luke-jr> genjix: pindexBest is the block's prevblock
302 2011-09-30 13:51:20 <CIA-101> bitcoin: various 0.4.x * r7944d8..4405b7 bitcoind-stable/ (288 files in 44 dirs): (964 commits)
303 2011-09-30 13:51:25 <genjix> that shouldn't ever happen
304 2011-09-30 13:51:32 <sipa> indeed, sounds like abug
305 2011-09-30 13:51:34 <luke-jr> genjix: it does with 0.3.24 apparently, at least
306 2011-09-30 13:51:43 <luke-jr> and no version from 0.3.20 to 0.4 can recover from it
307 2011-09-30 13:51:51 <luke-jr> the block download just locks up permanently
308 2011-09-30 13:51:59 <luke-jr> even across restarts
309 2011-09-30 13:52:13 <genjix> when ProcessBlock is called, it should check in ConnectBlock whether the new work is bigger than the last one + this current difficulty
310 2011-09-30 13:52:19 <genjix> then call Reorganize
311 2011-09-30 13:52:33 <luke-jr> genjix: ProcessBlock isn't called again, once it's in mapBlockIndex
312 2011-09-30 13:52:41 <luke-jr> though it doesn't explain how it got that way originally
313 2011-09-30 13:52:53 <genjix> it is
314 2011-09-30 13:52:57 <luke-jr> http://www.disc-shelf.com/bitcoin/bitcoin.zip is the .bitcoin dir
315 2011-09-30 13:52:59 <genjix> if it is in mapOrphanBlocks
316 2011-09-30 13:53:06 <genjix> then when the intermediary block is found
317 2011-09-30 13:53:06 <luke-jr> genjix: it isn't, though
318 2011-09-30 13:53:14 <luke-jr> there is no intermediary block either ;P
319 2011-09-30 13:53:17 <genjix> it recursively processes all the orphan blocks
320 2011-09-30 13:53:28 <genjix> yeah then it is a bug of some kind
321 2011-09-30 13:54:15 <luke-jr> anyhow, I'm stumped
322 2011-09-30 13:54:26 <luke-jr> hopefully someone else with knowledge of these internals can figure it out
323 2011-09-30 13:54:49 <imsaguy> hold the presses, luke-jr is stumped!
324 2011-09-30 13:55:42 <sipa> is it possible it is added to the block chain db, but not yet connected
325 2011-09-30 13:55:49 <sipa> and then the client crashed?
326 2011-09-30 13:56:56 <genjix> sipa: hmmm
327 2011-09-30 13:57:01 <genjix> let me check
328 2011-09-30 13:57:55 <genjix> sipa: yes it is a possibility
329 2011-09-30 13:57:57 <sipa> downloading again won't fix it, because the client considers the block already known (will not respoend to inv)
330 2011-09-30 13:57:59 <genjix> AddToBlockIndex
331 2011-09-30 13:58:11 <luke-jr> [08:07:02] <louigi> I did have my computer freeze
332 2011-09-30 13:58:18 <sipa> and subsequent block downloads will only try to connect to tip, making them all orphan
333 2011-09-30 13:58:38 <luke-jr> so the solution is to recover somehow I guess
334 2011-09-30 13:59:14 <sipa> removing that block from the blockdb should suffice then
335 2011-09-30 13:59:18 <genjix> its tough though since the further checks depend on it being added to the chain
336 2011-09-30 13:59:32 <luke-jr> sipa: this problem is fairly common-- recovery should be automatic ;p
337 2011-09-30 13:59:35 <genjix> so you cannot just add it to the chain *after* it's been validated
338 2011-09-30 13:59:51 <sipa> hmmm it seems not too hard
339 2011-09-30 14:00:08 <luke-jr> genjix: could you add it with a "not yet processed" flag?
340 2011-09-30 14:00:12 <sipa> when loading the chain, if the work is >> tip.work, reprocess
341 2011-09-30 14:00:18 <sipa> >
342 2011-09-30 14:01:46 <luke-jr> hmm
343 2011-09-30 14:01:50 <luke-jr> I think that makes sense
344 2011-09-30 14:02:05 <luke-jr> there should never be a work higher than the best, right?
345 2011-09-30 14:02:15 <luke-jr> since the existence of it implies the "best" isn't really the best
346 2011-09-30 14:02:28 <genjix> i don't think that's ideal
347 2011-09-30 14:02:33 <genjix> a flag is best
348 2011-09-30 14:02:46 <luke-jr> is height recorded on blocks in the db?
349 2011-09-30 14:02:53 <genjix> since there's still other things that happen after adding it to the block chain
350 2011-09-30 14:03:10 <luke-jr> genjix: after setting it as 'best'?
351 2011-09-30 14:03:23 <genjix> so a flag like 'fValidationComplete'
352 2011-09-30 14:03:41 <genjix> and once the entire ProcessBlock main body is done, that flag gets set on the block
353 2011-09-30 14:03:45 <phantomcircuit> so sad
354 2011-09-30 14:03:53 <phantomcircuit> i have to de normalize the intersango database
355 2011-09-30 14:03:57 <luke-jr> would adding it as an orphan first work?
356 2011-09-30 14:04:02 <genjix> when loading the chain, blocks without that set are re-processed
357 2011-09-30 14:04:03 <phantomcircuit> joining orders and limit orders is just too slow
358 2011-09-30 14:04:17 <phantomcircuit> and the schema was so pretty :(
359 2011-09-30 14:04:36 <sipa> the real solution is making the adding of the block to the db, and processing it atomic
360 2011-09-30 14:04:39 <sipa> no?
361 2011-09-30 14:04:47 <genjix> yeah
362 2011-09-30 14:05:05 <sipa> since having a block in the database that has higher work, but isn't in the main chain, is considered an inconsistancy
363 2011-09-30 14:05:06 <genjix> but not much you can do i suppose if bitcoin freezes halfway
364 2011-09-30 14:05:17 <sipa> eh, bdb has locking
365 2011-09-30 14:05:34 <sipa> it should support transactionality
366 2011-09-30 14:05:56 <gavinandresen> it does
367 2011-09-30 14:06:29 <luke-jr> I like sipa's solution
368 2011-09-30 14:06:34 <luke-jr> reprocess blocks higher than best
369 2011-09-30 14:07:31 <genjix> and you commit at the end, right?
370 2011-09-30 14:07:35 <genjix> that should work too
371 2011-09-30 14:07:48 <genjix> probably the most elegant method
372 2011-09-30 14:08:25 <sipa> not committing before the block is processed is the only solution (note: i haven't checked the code now, all i say is based on assumptions)
373 2011-09-30 14:08:59 <genjix> well it should be ok afaik, since none of the current info is loaded. it only uses the data in memory
374 2011-09-30 14:09:14 <luke-jr> sipa: it doesn't solve it, though.
375 2011-09-30 14:09:20 <genjix> and the only things which are loaded are previous blocks/txs... which should be committed anyway
376 2011-09-30 14:09:27 <sipa> it won't solve the problem for old clients with corrupted databases
377 2011-09-30 14:09:47 <sipa> and recovery from almost-impossible situations in general is probably wanted, just for stability
378 2011-09-30 14:09:54 <luke-jr> we get the same transaction behaviour I think if we just delete blocks higher than <best> at load too
379 2011-09-30 14:10:09 <sipa> luke-jr: check, that's a lot easier
380 2011-09-30 14:10:14 <luke-jr> sipa: it's not almost impossible, lots of people have had this issue
381 2011-09-30 14:10:15 <genjix> might be an idea instead of trying to have catch alls for all situations
382 2011-09-30 14:10:23 <genjix> instead to have good tools to recover
383 2011-09-30 14:10:51 <sipa> luke-jr: that also fixes further orphans
384 2011-09-30 14:11:43 <luke-jr> downside to that is, it might potentially destroy real orphans which may be desired for forensics at some point
385 2011-09-30 14:12:03 <luke-jr> I suppose instead of deleting, the first octet could be changed to '1' (never happens in real blocks)
386 2011-09-30 14:25:47 <luke-jr> gavinandresen: still there?
387 2011-09-30 14:51:45 <gavinandresen> no, I was away
388 2011-09-30 14:52:55 <luke-jr> <luke-jr> gavinandresen: btw, kinda up in the air whether your "DoSprevention" code should be merged to stable or not (it's on the boundary of bugfix and feature); what do you think?
389 2011-09-30 14:53:04 <luke-jr> gavinandresen: never got a reply to that ^
390 2011-09-30 14:53:21 <gavinandresen> I dunno.  I'm not a huge fan of "stable" before a bitcoin 1.0 release
391 2011-09-30 14:53:52 <gavinandresen> "Here's a stable version of this beta experimental doo-hickey...." is kind of a mixed message
392 2011-09-30 14:54:36 <gmaxwell> luke-jr: if you don't merge it, and some transitive vulnerablity slips through, your 'stable' nodes will end up getting themselves partitioned from the rest of the network.
393 2011-09-30 14:54:59 <gmaxwell> e.g. if antiDOS nodes hang up on nodes that forward something only a non-antidos node would forward.
394 2011-09-30 14:55:04 <luke-jr> or vice-versa :p
395 2011-09-30 14:55:18 <gmaxwell> I guess I could argue it both ways based on that.
396 2011-09-30 14:55:27 <gmaxwell> Depending on what nodes I expect to be more popular.
397 2011-09-30 14:55:37 <luke-jr> even 0.4 is a minority still
398 2011-09-30 14:55:55 <gmaxwell> E.g. if non-stable nodes are more popular you very much want the antidos code.  If stable nodes are more popular then you could go either way.
399 2011-09-30 14:56:06 <gmaxwell> (assuming that kind of bug)
400 2011-09-30 14:56:09 <luke-jr> gmaxwell: but I guess your point is that the real question is "should stable track protocol changes?"
401 2011-09-30 14:56:27 <gavinandresen> gmaxwell: by the way, I've been thinking about a point you made RE: getting old clients to upgrade.  I'm starting to think that planned obsolescence might be a very good idea.
402 2011-09-30 14:56:53 <gmaxwell> obviously that would need to be done with great care.
403 2011-09-30 14:57:16 <luke-jr> gavinandresen: that defeats the entire point of keeping protocol compatibility
404 2011-09-30 14:57:27 <luke-jr> right now, at least
405 2011-09-30 14:57:45 <gmaxwell> It doesn't defeat it, because planned could still have a long horizon.
406 2011-09-30 14:57:55 <gmaxwell> (or even a way to override)
407 2011-09-30 14:58:34 <gavinandresen> If I recall correctly, very-very old version of bitcoin will stop working in February.
408 2011-09-30 14:58:52 <gavinandresen> Yeah:  net.h:        // Version 0.2 obsoletes 20 Feb 2012
409 2011-09-30 15:00:45 <gmaxwell> luke-jr: e.g. I don't know if we'd ever want want to maintain the code for it, but you could require that old nodes be placed behind new nodes which are explicitly ACLed to allow them.  Thus getting them off the public network even if people continue to run them in private.
410 2011-09-30 15:38:54 <gavinandresen> gmaxwell: re: 2-of-3 being more useful than (a or b)  :   okey doke.  Feel like writing some code and unit tests to support it, and packaging it up into an easy-to-review pull request?
411 2011-09-30 15:58:33 <gmaxwell> gavinandresen: Okay. I'll need to catch up on what the current multisig patches are, I haven't been following it for a while.
412 2011-09-30 16:00:41 <gavinandresen> gmaxwell: https://github.com/gavinandresen/bitcoin-git/tree/multisig and https://github.com/gavinandresen/bitcoin-git/tree/multisig_testing
413 2011-09-30 16:01:27 <gavinandresen> And groffer has code, too, not sure what the state of it is:  https://github.com/groffer/bitcoin/tree/escrow
414 2011-09-30 16:03:03 <genjix> when will multisig be added?
415 2011-09-30 16:03:15 <genjix> i.e when will we see the first block with it
416 2011-09-30 16:03:26 <tcatm> yay, git push -f fixed the messed up pull request :)
417 2011-09-30 16:03:32 <gavinandresen> On testnet?  earlier today....
418 2011-09-30 16:03:42 <genjix> no main
419 2011-09-30 16:04:14 <genjix> main chain
420 2011-09-30 16:04:22 <gavinandresen> On mainnet?  I suppose I could connect to Elgius and submit some test transactions...  (or somebody else could)
421 2011-09-30 16:04:43 <gavinandresen> tcatm: yay
422 2011-09-30 16:05:19 <gavinandresen> genjix: why do you ask?
423 2011-09-30 16:05:29 <gmaxwell> luke-jr: are nonstandard transactions actually working on eligius now?
424 2011-09-30 16:05:42 <genjix> just curious.
425 2011-09-30 16:05:45 <gmaxwell> When BlueMatt and I attempted a hash locked txn it wouldn't actually confirm on eligius.
426 2011-09-30 16:05:55 <tcatm> gavinandresen: so your plan is to remove cryptopp in 0.5?
427 2011-09-30 16:06:01 <luke-jr> gmaxwell: should be
428 2011-09-30 16:06:13 <genjix> i.e if i design a tx mem pool, then i also have scope for planning too
429 2011-09-30 16:06:19 <luke-jr> gmaxwell: you probably just had trouble keeping it in Eligius's queue :P
430 2011-09-30 16:06:23 <gavinandresen> tcatm: if possible.  No real hurry, but I don't see any reason to wait.
431 2011-09-30 16:12:18 <da2ce7> ok.
432 2011-09-30 16:13:38 <da2ce7> I was hoping for the groffer proposal: https://gist.github.com/dba89537d352d591eb36
433 2011-09-30 16:13:50 <da2ce7> where we could have (a / b) or (c / d)
434 2011-09-30 16:14:37 <tcatm> gavinandresen: kk. I'll try to a few more "interesting" unit test cases withint the next days so further work on the internal miner will be easier (I'm going to simplify quite a large chunk of code and finally make bitcoin hash the actual serialized block header).
435 2011-09-30 16:15:36 <da2ce7> btw is groffer on irc, if so what is his/her username?
436 2011-09-30 16:34:04 <eT-eB> Does anyone know where to get some testnet BTC?
437 2011-09-30 16:34:28 <gavinandresen> http://testnet.freebitcoins.appspot.com/
438 2011-09-30 16:40:55 <eT-eB> Thanks gavinandresen
439 2011-09-30 16:44:16 <genjix> sipa: haha this is classy, "Bitcoin: mining our own business since 2009"
440 2011-09-30 16:47:37 <Blitzboom> hahaha
441 2011-09-30 16:50:39 <ThomasV> indeed, good one
442 2011-09-30 16:51:28 <genjix> more here https://bitcointalk.org/index.php?topic=46247.20
443 2011-09-30 16:51:37 <genjix> some of them are really good :)
444 2011-09-30 16:52:09 <ThomasV> "Bitcoin: getting ourselves rid of bankers since 2009"
445 2011-09-30 16:54:47 <helo> sters
446 2011-09-30 16:55:01 <ThomasV> "an idea that's too big to fail" <-- imo the best one
447 2011-09-30 16:56:18 <Blitzboom> We need new suckers, and we need them NOW.
448 2011-09-30 16:56:22 <Blitzboom> thats good too
449 2011-09-30 16:56:23 <genjix> except that it's not
450 2011-09-30 16:56:44 <genjix> one economist i spoke to is worried of bitcoin suffering a deflationary spiral in the future
451 2011-09-30 16:56:59 <Blitzboom> genjix: lol, i wish!
452 2011-09-30 16:57:19 <luke-jr> "suffering"
453 2011-09-30 16:57:25 <Blitzboom> give me all the deflationary death spiral thats possible
454 2011-09-30 16:57:27 <luke-jr> people spending less is a GOOD thing
455 2011-09-30 16:57:39 <genjix> im not an economist
456 2011-09-30 16:57:55 <genjix> but this applies to everyone in the bitcoin community, http://rationalwiki.org/wiki/Engineers_and_woo
457 2011-09-30 16:58:17 <genjix> lets not talk hand wave and dismiss ideas far outside our knowledge domain
458 2011-09-30 16:58:44 <Blitzboom> im not dismissing it
459 2011-09-30 16:58:56 <genjix> "But engineers as a group have a noted, and indeed quantified, tendency to pontificate on things well outside their area of expertise, to the point of actual fallacy."
460 2011-09-30 16:58:58 <Blitzboom> just saying i would welcome a deflationary spiral as long as i profit from it
461 2011-09-30 16:59:09 <genjix> i would not
462 2011-09-30 16:59:25 <luke-jr> genjix: that wiki is obviously ridiculous
463 2011-09-30 16:59:29 <log0s> the idea of something becoming so valuable it becomes worthless contradicts itself
464 2011-09-30 16:59:37 <Blitzboom> luke-jr: because it is not religious?
465 2011-09-30 16:59:43 <luke-jr> it fails on even the most basic concepts
466 2011-09-30 16:59:48 <genjix> bitcoin has value because it is a mechanism of trade
467 2011-09-30 16:59:53 <luke-jr> for example, mathematics is not only science, but it is one of the few /formal/ sciences
468 2011-09-30 16:59:54 <genjix> not because it is rare
469 2011-09-30 16:59:58 <Blitzboom> no, bitcoin has value because it has value
470 2011-09-30 17:00:06 <Blitzboom> it has nothing to do with trade, its circular
471 2011-09-30 17:00:16 <luke-jr> yet that wiki basically outright says mathematics is distinct from science
472 2011-09-30 17:00:21 <genjix> my spit is rare. doesn't make it valuable.
473 2011-09-30 17:00:35 <Blitzboom> genjix: bitcoin is valuable because of future projections
474 2011-09-30 17:00:39 <genjix> luke-jr: it is different.
475 2011-09-30 17:00:42 <Blitzboom> and because it was valuable in the past
476 2011-09-30 17:00:49 <luke-jr> genjix: nope
477 2011-09-30 17:00:51 <ThomasV> my spit is $100/ml
478 2011-09-30 17:00:59 <genjix> mathematics uses axioms to prove theorems
479 2011-09-30 17:01:11 <Blitzboom> it doesnt matter what kind of trade is conducted with it as long as peoples "illusion" of value is maintained
480 2011-09-30 17:01:27 <genjix> science uses empirical evidence and theories to explain things as best as it can
481 2011-09-30 17:01:41 <genjix> but nothing is ever proven. difference in methodology
482 2011-09-30 17:01:46 <luke-jr> genjix: your own cluelessness doesn't make this wiki any more clueful ;)
483 2011-09-30 17:02:08 <genjix> ipse dixit
484 2011-09-30 17:08:44 <genjix> luke-jr: have you seen this guy? http://www.youtube.com/watch?v=nLbTE56HqLE - even if you don't agree with him (and i can understand both ways), it's really interesting for me to understand the other POV
485 2011-09-30 17:08:54 <genjix> not having had any religious background :)
486 2011-09-30 17:09:22 <genjix> idk thought you might enjoy that...
487 2011-09-30 17:09:29 <luke-jr> why would I?
488 2011-09-30 17:09:50 <genjix> his story is thought provoking and interesting
489 2011-09-30 17:10:44 <genjix> well he has a strong maths/computing background and describes using mathematical analogies how/why he went from changing his entire belief system.
490 2011-09-30 17:10:59 <genjix> it works both ways, and is an interesting expose into psychology
491 2011-09-30 17:12:08 <genjix> starts getting good at part 2.0
492 2011-09-30 17:13:50 <luke-jr> he didn't change his entire belief system really
493 2011-09-30 17:14:18 <genjix> how? seems he did.
494 2011-09-30 17:14:31 <luke-jr> if he did, he would abandon morality
495 2011-09-30 17:14:33 <luke-jr> entirely
496 2011-09-30 17:14:52 <genjix> where do you believe morality comes from? yourself?
497 2011-09-30 17:14:55 <luke-jr> God
498 2011-09-30 17:15:06 <genjix> so can god decide what is moral and what is not?
499 2011-09-30 17:15:14 <luke-jr> He defines morality, yes.
500 2011-09-30 17:15:20 <genjix> can he say murder and rape is moral just because he says so?
501 2011-09-30 17:15:25 <luke-jr> He could.
502 2011-09-30 17:15:26 <genjix> or is morality outside god?
503 2011-09-30 17:15:40 <luke-jr> but then it wouldn't be murder, since murder is defined as unlawful
504 2011-09-30 17:16:20 <gavinandresen> so speaking of morality... I haven't had any feedback on my DoSorphans patch
505 2011-09-30 17:16:20 <genjix> so what is moral, is simply what god decides is moral?
506 2011-09-30 17:16:43 <luke-jr> genjix: moral/good = conforming to God's will
507 2011-09-30 17:17:26 <genjix> why do people pray then? isn't that trying to change god's will?
508 2011-09-30 17:17:34 <luke-jr> no
509 2011-09-30 17:17:45 <luke-jr> prayer changes the person praying
510 2011-09-30 17:17:46 <genjix> but god has a plan for you.
511 2011-09-30 17:17:53 <luke-jr> their disposition, etc
512 2011-09-30 17:18:10 <genjix> ok so would you pray for a pair of wings?
513 2011-09-30 17:18:17 <luke-jr> no
514 2011-09-30 17:18:19 <genjix> or do you moderate your prayer to realistic goals
515 2011-09-30 17:19:15 <nathan7> this isn't a place to discuss religion, mkay
516 2011-09-30 17:19:20 <luke-jr> http://www.newadvent.org/cathen/12345b.htm
517 2011-09-30 17:20:23 <genjix> nathan7: yeah sorry, it's not religion but i find belief systems interesting. how like no belief system is inherently wrong but just different. belief systems can be equally consistent internally.
518 2011-09-30 17:20:33 <nathan7> You can spend forever discussing circular and internally inconsistent logic.
519 2011-09-30 17:20:48 <neofutur> eh interesting offtopic, move it to ##mtgox-chat ;)
520 2011-09-30 17:23:51 <nathan7> (=
521 2011-09-30 18:47:35 <genjix> what is the reason/argument for using int64 with bool MoneyRange(value) { 0 <= value && value <= MAX_MONEY; }  versus  using uint64_t and bool MoneyRange(value) { value <= MAX_MONEY; } ?
522 2011-09-30 18:47:50 <genjix> surely the latter protects against overflow errors much better?
523 2011-09-30 18:48:09 <genjix> well to the same degree
524 2011-09-30 18:49:37 <gmaxwell> genjix: mixing unsigned int in C can lead to surprising bugs due to the implicit type promotion rules.
525 2011-09-30 18:50:44 <genjix> that works both ways
526 2011-09-30 18:53:43 <gmaxwell> Well, the promotion rule only works one way. Emperically on 32bit+ it's the unsigned type that tends to be the guilty one (I think my expirence on 16 bit machines is actually the opposite, and the promotion rule is pretty helpful there) ::shrugs:: ymmv.
527 2011-09-30 19:05:51 <lfm> I think the money type is signed to support debts and overdrawn accounts and stuff even tho its not used reall with bitcoin. and as you say, it works to the same degree The compiler might even optimize it internally the way you are asking us to manually.
528 2011-09-30 19:06:16 <gmaxwell> it almost certantly does.
529 2011-09-30 19:07:07 <gmaxwell> Usually signed types actually lead to better compiled code anyways signed overflow is formally undefined (not even implementation defined) and the compiler can optimize accordingly.
530 2011-09-30 19:10:38 <helo> it is dangerous to use unsigned for monetary values... when you accidentally make it < 0 it just asplode
531 2011-09-30 19:11:00 <Diablo-D3> gmaxwell: clearly we need unlimited integer support in psus
532 2011-09-30 19:11:00 <helo> if you write casino games ;)
533 2011-09-30 19:14:17 <lfm> in psus? power supply units?
534 2011-09-30 19:14:37 <lfm> so they don't asplode?
535 2011-09-30 19:14:47 <Diablo-D3> er
536 2011-09-30 19:14:49 <Diablo-D3> cpus
537 2011-09-30 19:16:01 <genjix> helo: how comes?
538 2011-09-30 19:16:18 <lfm> new superpower unit, with unlimited integer support. can power an infinite number of graphics cards!
539 2011-09-30 19:17:00 <genjix> helo: if the uint64_t overflows in bitcoin then that's ok since the max is 21 million
540 2011-09-30 19:17:52 <lfm> genjix ya, may be a tiny advantage in one place but offset other places.
541 2011-09-30 19:20:07 <lfm> genjix: also the unsigned is not as explicit therefore harder to understand that it is doing the underflow test implicitly
542 2011-09-30 19:20:55 <genjix> so it's not really a big deal either way?
543 2011-09-30 19:21:24 <lfm> I am saying the signed code is easier to follow
544 2011-09-30 19:22:44 <lfm> just seems like excesive optimization to use the unsigned underflow trick.
545 2011-09-30 19:25:25 <lfm> I admit I might use it myself at times but I hope its only with code I never intend to share
546 2011-09-30 20:14:40 <lfm> its kinda odd how it seems the difficulty has hit a wall!
547 2011-09-30 20:15:39 <JFK911> what wall is that
548 2011-09-30 20:15:42 <JFK911> ;;bc,stats
549 2011-09-30 20:15:46 <gribble> Current Blocks: 147556 | Current Difficulty: 1689334.4045971 | Next Difficulty At Block: 149183 | Next Difficulty In: 1627 blocks | Next Difficulty In About: 1 week, 5 days, 11 hours, 11 minutes, and 14 seconds | Next Difficulty Estimate: 1535184.94291375 | Estimated Percent Change: -9.12486368972
550 2011-09-30 20:15:51 <gribble> The average time to generate a block at 466391 Khps, given the supplied difficulty of 0.125, is 1 second
551 2011-09-30 20:15:51 <tcatm> ;;bc,calcd 466391  0.125
552 2011-09-30 20:16:02 <tcatm> hrm, something is broken
553 2011-09-30 20:16:28 <lfm> it peaked at 1888786 and has been dropping very slowly since
554 2011-09-30 20:16:45 <JFK911> oh yeah it costs more than its worth to mine for many
555 2011-09-30 20:16:51 <gribble> The average time to generate a block at 466391 Khps, given the supplied difficulty of 1, is 9 seconds
556 2011-09-30 20:16:51 <tcatm> ;;bc,calcd 466391 1
557 2011-09-30 20:18:15 <lfm> tcatm: you wanna show fractions of a sec if seconds is the only unit shown?
558 2011-09-30 20:18:53 <tcatm> lfm: no, I'm trying to test the internal miner and it looks like it does not find blocks
559 2011-09-30 20:19:14 <lfm> ah! ok
560 2011-09-30 20:20:09 <tcatm> what difficulty is first 32bits == 0?
561 2011-09-30 20:20:19 <lfm> 1.0
562 2011-09-30 20:21:02 <sipa> 65536/65535
563 2011-09-30 20:21:04 <lfm> but the internal miner reports hash speed in kh/s and it looks like you'r using the wrong unit in the ;;bc,calcd
564 2011-09-30 20:21:16 <lfm> or h/s
565 2011-09-30 20:21:30 <tcatm> oh, that might be the problem ;)
566 2011-09-30 20:21:38 <gribble> The average time to generate a block at 466 Khps, given the supplied difficulty of 0.125, is 19 minutes and 12 seconds
567 2011-09-30 20:21:38 <tcatm> ;;bc,calcd 466 0.125
568 2011-09-30 20:23:06 <lfm> assuming your cpu is an old p4 or something
569 2011-09-30 20:23:20 <tcatm> core i5
570 2011-09-30 20:24:09 <lfm> ok, its just one core and the internal miner is severly deoptimized from before then
571 2011-09-30 20:27:22 <lfm> thats odd, I get 1253420 h/s for one core on a core2
572 2011-09-30 20:27:59 <gribble> Error: "bc,cacld" is not a valid command.
573 2011-09-30 20:27:59 <lfm> ;;bc,cacld 1253.420 0.125
574 2011-09-30 20:28:11 <gribble> The average time to generate a block at 1253.420 Khps, given the supplied difficulty of 0.125, is 7 minutes and 8 seconds
575 2011-09-30 20:28:11 <lfm> ;;bc,calcd 1253.420 0.125
576 2011-09-30 20:28:57 <tcatm> it refuses to mine at all now :/
577 2011-09-30 20:29:36 <lfm> want me to try a testnet-in-a-box test too?
578 2011-09-30 20:31:55 <tcatm> is github down?
579 2011-09-30 20:32:12 <tcatm> mhm works again
580 2011-09-30 20:32:22 <tcatm> 64kbit/s internet can be annoying :)
581 2011-09-30 20:33:08 <tcatm> so if you could compile bitcoind for testnet-in-a-box can you try current master branch and compare results to https://github.com/tcatm/bitcoin/tree/no-cryptopp ?
582 2011-09-30 20:33:48 <lfm> tesnetinabox works with standard binaries
583 2011-09-30 20:34:37 <lfm> ill try with 0.4.0 first unless you really need me to get master branch
584 2011-09-30 20:35:23 <lfm> 0.4.0 is generating with my testnetinabox setup
585 2011-09-30 20:35:26 <tcatm> 0.4.0 should be fine, too
586 2011-09-30 20:35:36 <lfm> I just got a block
587 2011-09-30 20:36:58 <gribble> Error: "bc,cald" is not a valid command.
588 2011-09-30 20:36:58 <tcatm> ;;bc,cald 2100 0.125
589 2011-09-30 20:37:02 <tcatm> ;;bc,calcd 2100 0.125
590 2011-09-30 20:37:03 <gribble> The average time to generate a block at 2100 Khps, given the supplied difficulty of 0.125, is 4 minutes and 15 seconds
591 2011-09-30 20:38:18 <gribble> The average time to generate a block at 4321.054 Khps, given the supplied difficulty of 0.125, is 2 minutes and 4 seconds
592 2011-09-30 20:38:18 <lfm> ;;bc,calcd 4321.054 0.125
593 2011-09-30 20:41:00 <tcatm> I got a block :)
594 2011-09-30 20:41:30 <lfm> ok it seems ok then
595 2011-09-30 20:42:19 <tcatm> next test: does it work without alignup<16>? :)
596 2011-09-30 20:42:44 <lfm> ya I would think it would
597 2011-09-30 20:42:52 <tcatm> IIRC that was added to allow 4way to work
598 2011-09-30 20:43:34 <lfm> ya, the see alignments are special
599 2011-09-30 20:43:37 <lfm> sse
600 2011-09-30 20:44:51 <tcatm> IIRC that was added to allow 4way to work
601 2011-09-30 20:45:44 <tcatm> seems slower without alignup. only 943kh/s
602 2011-09-30 20:45:52 <lfm> hehe, ok there is still the internal generator, you just cant access it from the gui any more! you guys are so funny sometimes!
603 2011-09-30 20:46:37 <tcatm> yep and because it's there we still have to make sure it works :/
604 2011-09-30 20:46:42 <lfm> tcatm: that may be a core iX "feature"
605 2011-09-30 20:48:16 <lfm> whats your clock speed on that core i3? It seems my core2 is faster now, I wonder if it was with all the old sse code and stuff. (I know you dont care at this point ...)
606 2011-09-30 20:48:59 <tcatm> 2.53Ghz
607 2011-09-30 20:49:27 <luke-jr> meh
608 2011-09-30 20:49:36 <luke-jr> scrap the internal miner and embed cgminer
609 2011-09-30 20:49:37 <luke-jr> :P
610 2011-09-30 20:49:50 <luke-jr> ./configure --with-cgminer=/path/to/cgminer/code
611 2011-09-30 20:49:54 <tcatm> cgminer = GPU?
612 2011-09-30 20:49:58 <luke-jr> cgminer is CPU or GPU
613 2011-09-30 20:50:04 <lfm> mine is a Q9450 at 2.0ghz
614 2011-09-30 20:50:30 <lfm> tcatm: ya cgminer does either cpu or gpu
615 2011-09-30 20:50:53 <tcatm> 2167kh/s with genproclimit = 2
616 2011-09-30 20:51:29 <tcatm> luke-jr: the internal miner is still great to test all those functions getwork calls
617 2011-09-30 20:52:11 <lfm> "hashespersec" : 2439386,
618 2011-09-30 20:52:14 <lfm> "genproclimit" : 2,
619 2011-09-30 20:52:25 <tcatm> now back to 900kh/s. probably some cpu throttling
620 2011-09-30 20:52:45 <lfm> oh ok
621 2011-09-30 20:53:18 <lfm> put a bigger cooler on it -> itll go faster?
622 2011-09-30 20:53:43 <tcatm> probably, but not that easy on a laptop
623 2011-09-30 20:53:55 <lfm> oh ya ok
624 2011-09-30 20:54:22 <tcatm> hrm,  I could shutdown bitcoincharts for a few minutes and mine on that server
625 2011-09-30 20:55:15 <lfm> naw, those charts are more important than my curiosity
626 2011-09-30 20:55:45 <tcatm> well, I need to know whether the miner still works or not
627 2011-09-30 20:56:57 <lfm> seems it is working.
628 2011-09-30 20:57:50 <lfm> hehe if you do "bitcoind setgenerate true 0" it sets generate to false
629 2011-09-30 20:58:22 <tcatm> yep and if you setgenerate true twice it doesn't mine even though generate is true
630 2011-09-30 21:00:14 <lfm> ok then its not as smart as I thot
631 2011-09-30 21:18:18 <bitanarchy> I have a suggestion for conditional transactions in the block chain. Can this work? https://bitcointalk.org/index.php?topic=8821.msg552339#msg552339
632 2011-09-30 21:20:25 <tcatm> I don't think that will work with the current scripting implementation but I might not have fully understood it.
633 2011-09-30 21:21:02 <lfm> I dont think so either. It is not clear what are the "conditions" for awarding either A or B.
634 2011-09-30 21:21:07 <luke-jr> looks like nonsense imo
635 2011-09-30 21:22:09 <lfm> luke-jr I tend to agree but I was trying to be nice for once
636 2011-09-30 21:22:48 <bitanarchy> lfm: could be any readable condition like "superluminous neutrinos are confirmed at certain date"
637 2011-09-30 21:23:11 <bitanarchy> sure I probably make a mistake... but what is the mistake?
638 2011-09-30 21:23:44 <lfm> ok, thats sounding more certainly like intenable nonsense. the whole beauty of the bitcoin block chain is that it is self contained in a sense.
639 2011-09-30 21:24:56 <tcatm> won't work. whatever the condition is it would have to be verifyable by a computer.
640 2011-09-30 21:25:05 <bitanarchy> lfm: what do you mean with self-contained? By the way... I am not suggesting to implement it in bitcoin... but it could be a seperate experimental currency.
641 2011-09-30 21:25:15 <lfm> you seem to be adding all sorts of outside "deciders" like a wiki and now some neutrinos.
642 2011-09-30 21:26:05 <bitanarchy> lfm: yes, I want the community to decide whether a condition used for a transaction is false or true
643 2011-09-30 21:26:45 <lfm> with the block chain as it is you can figure out where every bitcoin ever created is now and every where it has ever been. (by location I mean attached to certain bitocin "addresses".
644 2011-09-30 21:26:54 <tcatm> bitanarchy: try to solve that problem (i.e. prevent cheating and all that) without adding a currency first :)
645 2011-09-30 21:28:54 <lfm> bitanarchy: it would be an incredible complication of the security and provability if we had to somehow take into account the status of some wiki page or some neutrino
646 2011-09-30 21:30:00 <gmaxwell> bitanarchy: one challenge is that if it has monolithic blocks like bitcoin you'll have problems where I accept 99% of the txn in a block but 1% I reject... and someone else rejects a different 1% and so on...
647 2011-09-30 21:30:56 <casascius> https://en.bitcoin.it/wiki/Casascius_Bitcoin_POS_system
648 2011-09-30 21:35:48 <bitanarchy> gmaxwell: why do miners reject false transactions right now?
649 2011-09-30 21:36:29 <bitanarchy> ah, ok because the whole block will be rejected...
650 2011-09-30 21:38:52 <gmaxwell> Right. Alternatively ask "why do miners not cheat and insert crap transactions that are invalid"   because it makes the whole block invalid.
651 2011-09-30 21:51:15 <bitanarchy> I suggest that invalid transactions must still be rejected, but I want to allow blocks that contain transactions based on wrong outcomes of conditions. The miners will have an incentive to ignore those because the corresponding fees are worthless.
652 2011-09-30 22:00:15 <bitanarchy> I like to see come into existence a decentralized system in which the users are rewarded for investigating the truth of certain statements. The best way to do that is to let these people bet on these statements like a prediction market.
653 2011-09-30 22:00:35 <elkingrey> Can anybody help me diagnose the problem? I've installed bitcoind on my server but get this error http://pastebin.com/GQfh5p4K
654 2011-09-30 22:01:04 <bd_> bitanarchy: what kind of conditions?
655 2011-09-30 22:01:31 <gmaxwell> elkingrey: tail -f ~/.bitcoin/debug.log
656 2011-09-30 22:01:41 <bitanarchy> bd_: why? does that matter?
657 2011-09-30 22:01:47 <gmaxwell> elkingrey: bitcoin can take a little time to become responsive after startup.
658 2011-09-30 22:02:03 <bd_> bitanarchy: Yes, it very much does matter. How will the bitcoin servers evaluate whether your condition is fulfilled, efficiently, and consistently?
659 2011-09-30 22:02:19 <bd_> As soon as bitcoin transactions rely on non-local information you open a HUGE can of worms, protocol-wise
660 2011-09-30 22:03:11 <elkingrey> gmaxwell: http://pastebin.com/rtS4pxNF
661 2011-09-30 22:03:57 <bd_> bitanarchy: The only way bitcoin can work is if any bitcoin server can look at the block chain up to point X, plus a transaction, and determine _consistently_, with _only that information_ whether the transaction is valid or not
662 2011-09-30 22:04:09 <bitanarchy> bd_: scientific statements can be evaluated... but sometimes there is some ambiguity in the interpretation... but that has to be decided by the intersubjective evaluation of the community of users
663 2011-09-30 22:04:19 <bd_> bitanarchy: Right. So this is beyond the scope of bitcoin.
664 2011-09-30 22:04:25 <gmaxwell> elkingrey: looks like its still loading addresses.
665 2011-09-30 22:04:36 <elkingrey> gmaxwell: What does that mean?
666 2011-09-30 22:04:38 <gmaxwell> elkingrey: is the tail output actively scrolling?
667 2011-09-30 22:04:44 <bd_> bitanarchy: You've basically said, "well, someone has to go and say, 'yes, this is valid'". So, why not have that someone just... pay normally when they say yes?
668 2011-09-30 22:04:52 <bd_> Why does the bitcoin protocol need to get involved?
669 2011-09-30 22:05:00 <elkingrey> gmaxwell: I don't know what you mean by scrolling, but it is changing, yes
670 2011-09-30 22:05:23 <gmaxwell> elkingrey: then bitcoin itself is working.
671 2011-09-30 22:05:34 <gmaxwell> elkingrey: did you set an rpc user and password in the configuration?
672 2011-09-30 22:05:50 <elkingrey> gmaxwell: Yes
673 2011-09-30 22:05:59 <gmaxwell> Could there be any local firewalling which is preventing it from connecting to itself?
674 2011-09-30 22:06:15 <elkingrey> Don't think so.
675 2011-09-30 22:06:24 <gmaxwell> I _think_ that by the time you see add-address scrolling it should be responding to getinfo.
676 2011-09-30 22:06:36 <bitanarchy> bd_: the decentralized prediction market needs a decentralized contract system/notary and it needs a decentralized currency or value system.