1 2011-11-11 00:03:31 SomeoneWeirdzzzz has quit (Ping timeout: 276 seconds)
   2 2011-11-11 00:10:03 <erus`> xor with "fuckyou" repeated
   3 2011-11-11 00:10:09 <erus`> unbreakable
   4 2011-11-11 00:10:27 <cjdelisle> xor with /dev/urandom
   5 2011-11-11 00:10:33 <cjdelisle> really is unbreakable
   6 2011-11-11 00:10:51 glitch-mod has quit (Ping timeout: 245 seconds)
   7 2011-11-11 00:11:37 * cjdelisle waits for gmaxwell to show up and explain all about recreating random seeds and random generator vulns
   8 2011-11-11 00:14:27 iocor has quit (Quit: Computer has gone to sleep.)
   9 2011-11-11 00:16:53 Sedra has quit (Ping timeout: 252 seconds)
  10 2011-11-11 00:17:01 Sedra has joined
  11 2011-11-11 00:19:17 Guest99952 has quit (Ping timeout: 248 seconds)
  12 2011-11-11 00:21:27 traviscj has quit (Remote host closed the connection)
  13 2011-11-11 00:21:51 MBS has quit (Read error: Connection reset by peer)
  14 2011-11-11 00:25:54 jarpiain has joined
  15 2011-11-11 00:26:21 jarpiain is now known as Guest44250
  16 2011-11-11 00:26:44 shadders has quit (Read error: Connection reset by peer)
  17 2011-11-11 00:28:06 djoot has quit (Read error: Operation timed out)
  18 2011-11-11 00:28:46 djoot has joined
  19 2011-11-11 00:30:59 shadders has joined
  20 2011-11-11 00:31:30 <luke-jr> gavinandresen: 0.5 or 0.4.1
  21 2011-11-11 00:32:15 <gavinandresen> luke-jr: haven't decided yet, still working on a robust solution....
  22 2011-11-11 00:32:19 <luke-jr> also, only 0.4.0 is affected, not 0.4.1
  23 2011-11-11 00:32:40 Guest30746 has joined
  24 2011-11-11 00:32:40 <gavinandresen> (I'll change 0.4 to 0.4.0)
  25 2011-11-11 00:32:45 btc_novice has left ()
  26 2011-11-11 00:33:43 <luke-jr> 1. this only affects users who distributed their encrypted wallet files (so don't do it!) 2. if so, assume your addresses are all compromised 3. do <this> immediately to get secure 4. upgrade to 0.4.1 or newer to trust encryption again
  27 2011-11-11 00:34:13 <luke-jr> yep, looks like you covered it all
  28 2011-11-11 00:38:25 gjs278 has quit (Remote host closed the connection)
  29 2011-11-11 00:39:45 xsteve has joined
  30 2011-11-11 00:43:40 traviscj has joined
  31 2011-11-11 00:44:48 chrisb__ has quit (Quit: Ex-Chat)
  32 2011-11-11 00:45:49 NickelBot has quit (Read error: Connection reset by peer)
  33 2011-11-11 00:45:49 knotwork has quit (Read error: Connection reset by peer)
  34 2011-11-11 00:47:06 marf_away has quit (Ping timeout: 258 seconds)
  35 2011-11-11 00:50:44 NickelBot has joined
  36 2011-11-11 00:59:09 localhost has quit (Remote host closed the connection)
  37 2011-11-11 01:02:31 localhost has joined
  38 2011-11-11 01:06:13 caedes has quit (Ping timeout: 248 seconds)
  39 2011-11-11 01:07:14 bodom has joined
  40 2011-11-11 01:08:15 Kolky has quit (Quit: Bye bye!)
  41 2011-11-11 01:08:53 xsteve has quit (Quit: leaving)
  42 2011-11-11 01:09:32 gjs278 has joined
  43 2011-11-11 01:15:22 <phungus> that sounds bad
  44 2011-11-11 01:16:35 <phungus> I'm still on 0.3.24 because I didn't want to be a .0 adopter. :-)
  45 2011-11-11 01:16:44 <phungus> now I'm thinking that might have been a wise choice
  46 2011-11-11 01:16:57 wasabi1 has joined
  47 2011-11-11 01:17:08 <sipa> i believe 0.3.24 was one of the least stable releases we had
  48 2011-11-11 01:17:46 TheZimm has joined
  49 2011-11-11 01:19:36 <luke-jr> phungus: why?
  50 2011-11-11 01:19:54 <luke-jr> Gentoo stable is on (patched) 0.3.20.3 IIRC
  51 2011-11-11 01:20:05 <luke-jr> phungus: 0.4.1 will be out as soon as we get a fix finished
  52 2011-11-11 01:20:16 * luke-jr pokes gavinandresen for being slow
  53 2011-11-11 01:20:36 <luke-jr> oh, and people need to volunteer to build binaries…
  54 2011-11-11 01:21:24 <gavinandresen> slow == careful....  there are issues....
  55 2011-11-11 01:21:28 TheZimm has quit (Client Quit)
  56 2011-11-11 01:21:34 <gavinandresen> Actually, who knows the most about berkeley db here?
  57 2011-11-11 01:21:57 <luke-jr> *crickets chirp*
  58 2011-11-11 01:22:10 <gavinandresen> sigh
  59 2011-11-11 01:22:57 <gavinandresen> In particular, I need to know if there is a way to tell bdb to clean up after itself-- remove any temporary files, log files, etc.
  60 2011-11-11 01:23:23 <gavinandresen> ... or do "we" need to write code to clean up after it ourselves.
  61 2011-11-11 01:24:50 <sipa> gavinandresen: http://pybsddb.sourceforge.net/api_c/env_remove.html
  62 2011-11-11 01:26:08 crazy_imp has quit (Ping timeout: 255 seconds)
  63 2011-11-11 01:27:47 crazy_imp has joined
  64 2011-11-11 01:28:25 rdponticelli has joined
  65 2011-11-11 01:33:03 <phungus> 0.3.24 has been pretty stable on Windows 7
  66 2011-11-11 01:33:17 <phungus> and I suppose on Linux too but I haven't used my dev instance much
  67 2011-11-11 01:33:30 <phungus> but yes, I will upgrade at some point. :-)
  68 2011-11-11 01:33:35 <phungus> I use truecrypt for now anyway
  69 2011-11-11 01:51:26 wolfspraul has joined
  70 2011-11-11 01:51:37 wolfspraul has quit (Client Quit)
  71 2011-11-11 01:51:51 wolfspraul has joined
  72 2011-11-11 01:51:52 wolfspraul is now known as Guest77365
  73 2011-11-11 01:52:34 Guest77365 has quit (Client Quit)
  74 2011-11-11 01:52:55 wolfspra1l has joined
  75 2011-11-11 01:57:35 roconnor_ has joined
  76 2011-11-11 01:59:14 rdponticelli_ has joined
  77 2011-11-11 01:59:55 rdponticelli has quit (Ping timeout: 240 seconds)
  78 2011-11-11 02:03:23 gjs278 has quit (Remote host closed the connection)
  79 2011-11-11 02:03:36 erus` has quit (Quit: ChatZilla 0.9.87 [Firefox 8.0/20111104165243])
  80 2011-11-11 02:04:50 copumpkin has quit (Ping timeout: 255 seconds)
  81 2011-11-11 02:05:28 Turingi has quit (Read error: Connection reset by peer)
  82 2011-11-11 02:08:30 Cusipzzz has joined
  83 2011-11-11 02:12:21 pickett has quit (Ping timeout: 248 seconds)
  84 2011-11-11 02:20:23 mizerydearia has quit (Ping timeout: 256 seconds)
  85 2011-11-11 02:20:43 <roconnor_> what happens in the bitcoin client when CastToBigNum throws a runtime exception?
  86 2011-11-11 02:20:57 <roconnor_> presumably it is caught somewhere?
  87 2011-11-11 02:21:25 <roconnor_> ah it is caught in EvalScript
  88 2011-11-11 02:21:28 <roconnor_> nevermind
  89 2011-11-11 02:21:58 pickett has joined
  90 2011-11-11 02:25:53 bodom has quit (Remote host closed the connection)
  91 2011-11-11 02:35:39 mizerydearia has joined
  92 2011-11-11 02:38:13 copumpkin has joined
  93 2011-11-11 02:38:37 <gavinandresen> luke-jr: I've finished the 'make sure private keys aren't left on disk when bitcoin exits' code done.  Just have to write the 'obsolete the keypool' code.
  94 2011-11-11 02:49:08 <roconnor_> anyone know if the integers processed in scripts are big-endian or little-endian?
  95 2011-11-11 02:52:06 knotwork has joined
  96 2011-11-11 02:52:47 theymos has joined
  97 2011-11-11 02:54:01 copumpkin has quit (Quit: Computer has gone to sleep.)
  98 2011-11-11 02:54:35 <luke-jr> roconnor_: if it matters, I'd be afraid it would cause a block chain split
  99 2011-11-11 02:56:41 <roconnor_> luke-jr: it is exactly one of those subtle things that could go unnoticed for a long time
 100 2011-11-11 02:57:35 <roconnor_> luke-jr: after looking carefully, it seems to use the same code that the serializer uses, which makes it little endian
 101 2011-11-11 03:03:23 osmosis has joined
 102 2011-11-11 03:07:58 twobitcoins has quit (Ping timeout: 260 seconds)
 103 2011-11-11 03:09:20 <roconnor_> I take that back, I'm not sure whether it is little endian or not
 104 2011-11-11 03:13:34 <luke-jr> gavinandresen: btw, how do you go from README.md -> README?
 105 2011-11-11 03:14:56 <luke-jr> oh
 106 2011-11-11 03:14:58 <luke-jr> a symlink
 107 2011-11-11 03:23:25 eastender has joined
 108 2011-11-11 03:23:44 twobitcoins has joined
 109 2011-11-11 03:36:09 copumpkin has joined
 110 2011-11-11 03:49:20 TheZimm has joined
 111 2011-11-11 03:51:25 wolfspra1l has quit (Quit: leaving)
 112 2011-11-11 03:53:44 wasabi2 has joined
 113 2011-11-11 03:54:52 wasabi1 has quit (Ping timeout: 258 seconds)
 114 2011-11-11 03:57:59 B0g4r7_ has joined
 115 2011-11-11 04:07:10 denisx has quit (Quit: denisx)
 116 2011-11-11 04:10:30 Cusipzzz has quit (Ping timeout: 260 seconds)
 117 2011-11-11 04:11:12 TheSeven has quit (Disconnected by services)
 118 2011-11-11 04:11:25 [7] has joined
 119 2011-11-11 04:12:14 osmosis has quit (Quit: Leaving)
 120 2011-11-11 04:12:46 eoss has quit (Quit: Leaving)
 121 2011-11-11 04:20:40 rdponticelli has joined
 122 2011-11-11 04:20:53 rdponticelli_ has quit (Ping timeout: 248 seconds)
 123 2011-11-11 04:20:56 SomeoneWeird has joined
 124 2011-11-11 04:23:00 Diablo-D3 has quit (Ping timeout: 255 seconds)
 125 2011-11-11 04:33:18 etotheipi__ has quit (Quit: Ex-Chat)
 126 2011-11-11 04:36:59 slush has quit (Ping timeout: 259 seconds)
 127 2011-11-11 04:39:54 imsaguy has quit (Ping timeout: 260 seconds)
 128 2011-11-11 04:42:43 MimeNarrator has quit (Remote host closed the connection)
 129 2011-11-11 04:43:01 MimeNarrator has joined
 130 2011-11-11 04:44:57 glitch-mod has joined
 131 2011-11-11 04:45:56 RobinPKR_ has joined
 132 2011-11-11 04:46:31 dan__ has joined
 133 2011-11-11 04:48:12 RobinPKR has quit (Ping timeout: 255 seconds)
 134 2011-11-11 04:48:12 RobinPKR_ is now known as RobinPKR
 135 2011-11-11 04:52:20 amtal has quit (Ping timeout: 240 seconds)
 136 2011-11-11 04:53:53 imsaguy has joined
 137 2011-11-11 04:53:53 imsaguy has quit (Changing host)
 138 2011-11-11 04:53:53 imsaguy has joined
 139 2011-11-11 04:57:15 roconnor_ has quit (Ping timeout: 240 seconds)
 140 2011-11-11 04:58:24 <gavinandresen> luke-jr: heading to sleep-- patch for the problem will have to wait until tomorrow, I'm too tired to test carefully
 141 2011-11-11 04:58:40 <luke-jr> oh well :<
 142 2011-11-11 05:00:19 <gavinandresen> luke-jr: progress so far:  https://github.com/gavinandresen/bitcoin-git/commits/encryptionbug
 143 2011-11-11 05:00:39 <luke-jr> gavinandresen: not ready for merging, I presume?
 144 2011-11-11 05:00:43 <luke-jr> or ready, but needs testing?
 145 2011-11-11 05:00:51 <gavinandresen> not ready-- at least one problem
 146 2011-11-11 05:01:06 <gavinandresen> (encrypting the wallet should top-up the keypool with new keys before locking it)
 147 2011-11-11 05:01:24 <gavinandresen> (and there is code to do that... but it doesn't work)
 148 2011-11-11 05:01:28 casascius has joined
 149 2011-11-11 05:01:55 <luke-jr> gavinandresen: fwiw, it would make things a little easier for me if the commits were built on top of the v0.4.0 tag instead of master, but I can backport if needed
 150 2011-11-11 05:02:08 <gmaxwell> gavinandresen: probably some lock race that prevent the keypool toping?
 151 2011-11-11 05:02:27 <casascius> Hello everyone.  I think I am about to appear on the news in a TV news story about Bitcoin that starts in about 15 minutes at CBS news station KOLD (www.kold.com)
 152 2011-11-11 05:02:29 <casascius> It is being advertised on the home page of the news station
 153 2011-11-11 05:02:47 <luke-jr> casascius: be sure to promote Tonal too!
 154 2011-11-11 05:02:51 <gavinandresen> gmaxwell: not locking, I don't think... the keystore/wallet APIs are being... ummm... subtle
 155 2011-11-11 05:03:04 <casascius> luke-jr: the interview is pre-recorded
 156 2011-11-11 05:03:07 <luke-jr> casascius: :P
 157 2011-11-11 05:03:18 <casascius> i gave it over skype a week ago
 158 2011-11-11 05:03:33 <casascius> they asked questions about security
 159 2011-11-11 05:03:35 <gavinandresen> gmaxwell: anyway, that'll wait until tomorrow, unless somebody else wants to take a crack at it while I'm asleep
 160 2011-11-11 05:03:36 * luke-jr wonders if Skype demands special licensing for them to use it for that
 161 2011-11-11 05:04:05 <casascius> luke-jr: doubt it
 162 2011-11-11 05:07:40 gavinandresen has quit (Quit: gavinandresen)
 163 2011-11-11 05:15:12 justmoon has quit (Read error: Operation timed out)
 164 2011-11-11 05:19:39 RAWRwins254 has quit (Read error: Operation timed out)
 165 2011-11-11 05:19:43 RAWRwins254 has joined
 166 2011-11-11 05:20:27 wasabi has quit (Ping timeout: 245 seconds)
 167 2011-11-11 05:21:26 glitch-mod has quit (Ping timeout: 260 seconds)
 168 2011-11-11 05:23:25 wasabi has joined
 169 2011-11-11 05:24:39 amiller has quit (Quit: Leaving)
 170 2011-11-11 05:28:47 wasabi2 has quit (Read error: Connection reset by peer)
 171 2011-11-11 05:28:52 wolfspraul has joined
 172 2011-11-11 05:33:05 Beremat has joined
 173 2011-11-11 05:42:29 dan_____ has joined
 174 2011-11-11 05:43:46 amtal has joined
 175 2011-11-11 05:44:20 dan__ has quit (Ping timeout: 240 seconds)
 176 2011-11-11 05:44:20 dan_____ is now known as dan__
 177 2011-11-11 05:44:44 Nesetalis has quit (Read error: Connection reset by peer)
 178 2011-11-11 05:45:01 Nesetalis has joined
 179 2011-11-11 05:47:32 WakiMiko has quit (Ping timeout: 240 seconds)
 180 2011-11-11 05:49:40 WakiMiko has joined
 181 2011-11-11 05:49:48 WakiMiko has quit (Changing host)
 182 2011-11-11 05:49:48 WakiMiko has joined
 183 2011-11-11 05:52:09 justmoon has joined
 184 2011-11-11 05:53:37 wasabi has quit (Read error: Connection reset by peer)
 185 2011-11-11 05:54:04 amiller has joined
 186 2011-11-11 05:55:18 BurtyB2 has quit (Ping timeout: 248 seconds)
 187 2011-11-11 05:58:33 rdponticelli_ has joined
 188 2011-11-11 05:59:02 rdponticelli has quit (Ping timeout: 248 seconds)
 189 2011-11-11 06:01:24 BurtyB has joined
 190 2011-11-11 06:01:50 RazielZ has joined
 191 2011-11-11 06:02:38 casascius has quit (Quit: Page closed)
 192 2011-11-11 06:04:26 BurtyBB has joined
 193 2011-11-11 06:06:27 BurtyB has quit (Ping timeout: 240 seconds)
 194 2011-11-11 06:10:20 dan__ has quit (Ping timeout: 240 seconds)
 195 2011-11-11 06:13:57 rdponticelli has joined
 196 2011-11-11 06:14:30 rdponticelli_ has quit (Ping timeout: 248 seconds)
 197 2011-11-11 06:15:11 ThomasV has joined
 198 2011-11-11 06:20:21 cocktopus has quit (Disconnected by services)
 199 2011-11-11 06:20:34 cocktopus has joined
 200 2011-11-11 06:21:08 cocktopus has quit (Client Quit)
 201 2011-11-11 06:22:23 cocktopus has joined
 202 2011-11-11 06:23:45 SomeoneWeird has quit (Quit: Leaving)
 203 2011-11-11 06:24:05 dan__ has joined
 204 2011-11-11 06:24:49 TheZimm has quit (Quit: Computer has gone to sleep.)
 205 2011-11-11 06:26:02 BurtyB has joined
 206 2011-11-11 06:27:28 BurtyBB has quit (Ping timeout: 256 seconds)
 207 2011-11-11 06:29:23 num1 has joined
 208 2011-11-11 06:30:21 eastender has quit (Ping timeout: 255 seconds)
 209 2011-11-11 06:38:08 theymos has quit (Remote host closed the connection)
 210 2011-11-11 06:39:51 dan__ has quit (Quit: dan__)
 211 2011-11-11 06:40:39 dvide has quit ()
 212 2011-11-11 06:41:24 MimeNarrator has quit (Remote host closed the connection)
 213 2011-11-11 06:44:42 BurtyB has quit (Ping timeout: 260 seconds)
 214 2011-11-11 06:45:45 BurtyB has joined
 215 2011-11-11 06:54:30 B0g4r7_ has quit (Ping timeout: 248 seconds)
 216 2011-11-11 06:55:29 MimeNarrator has joined
 217 2011-11-11 07:01:13 shadders has quit (Remote host closed the connection)
 218 2011-11-11 07:02:41 SomeoneWeird has joined
 219 2011-11-11 07:03:47 shadders has joined
 220 2011-11-11 07:08:04 pickett has quit (Remote host closed the connection)
 221 2011-11-11 07:12:49 pickett has joined
 222 2011-11-11 07:19:35 frozensquirrel has quit (Remote host closed the connection)
 223 2011-11-11 07:27:20 molecular has quit (Read error: Operation timed out)
 224 2011-11-11 07:34:30 shadders has quit (Ping timeout: 248 seconds)
 225 2011-11-11 07:34:34 gjs278 has joined
 226 2011-11-11 07:42:34 shadders has joined
 227 2011-11-11 07:42:36 molecular has joined
 228 2011-11-11 07:49:09 pickett has quit (Remote host closed the connection)
 229 2011-11-11 07:53:03 pickett has joined
 230 2011-11-11 07:54:24 Guest44250 is now known as jarpiain
 231 2011-11-11 07:55:41 CaptainDDL has quit (Quit: I leave my first mate in charge!)
 232 2011-11-11 08:05:57 ThomasV has quit (Ping timeout: 240 seconds)
 233 2011-11-11 08:13:08 Burgundy has joined
 234 2011-11-11 08:20:27 enquirer has joined
 235 2011-11-11 08:50:05 <CIA-89> bitcoin: Con Kolivas * rd548670ed23e cgminer/README: Update README.
 236 2011-11-11 08:57:33 pickett has quit (Remote host closed the connection)
 237 2011-11-11 09:00:07 <CIA-89> bitcoin: Con Kolivas * r19cad2a61588 cgminer/NEWS: Update NEWS.
 238 2011-11-11 09:00:09 <CIA-89> bitcoin: Con Kolivas * rb0bfd816d621 cgminer/miner.h: Don't build VIA on apple since -a auto bombs instead of gracefully ignoring VIA failing.
 239 2011-11-11 09:00:11 <CIA-89> bitcoin: Con Kolivas * r0638155c3707 cgminer/main.c: Don't load the default config if a config file is specified on the command line.
 240 2011-11-11 09:01:44 pickett has joined
 241 2011-11-11 09:06:11 AStove has joined
 242 2011-11-11 09:10:01 <CIA-89> bitcoin: Con Kolivas * r79159c3bcc0e cgminer/NEWS: Update NEWS.
 243 2011-11-11 09:10:02 <CIA-89> bitcoin: Con Kolivas * r35a8b9f3c5ee cgminer/configure.ac: Bump version number to 2.0.8
 244 2011-11-11 09:10:45 larsivi has joined
 245 2011-11-11 09:27:18 erle- has joined
 246 2011-11-11 09:31:07 gjs278 has quit (Remote host closed the connection)
 247 2011-11-11 09:33:45 shadders has quit (Ping timeout: 258 seconds)
 248 2011-11-11 09:35:38 gjs278 has joined
 249 2011-11-11 09:39:01 gjs278 has quit (Remote host closed the connection)
 250 2011-11-11 09:53:25 shadders has joined
 251 2011-11-11 09:55:26 abragin has joined
 252 2011-11-11 09:55:27 abragin has quit (Changing host)
 253 2011-11-11 09:55:27 abragin has joined
 254 2011-11-11 09:56:28 Edward_Black has quit (Ping timeout: 240 seconds)
 255 2011-11-11 10:01:44 ThomasV has joined
 256 2011-11-11 10:06:56 Edward_Black has joined
 257 2011-11-11 10:09:11 Rav3nSw0rd has quit (Read error: Operation timed out)
 258 2011-11-11 10:13:21 AStove has quit (Ping timeout: 260 seconds)
 259 2011-11-11 10:18:47 larsivi has quit (Ping timeout: 260 seconds)
 260 2011-11-11 10:19:49 larsivi has joined
 261 2011-11-11 10:23:34 datagutt has joined
 262 2011-11-11 10:26:16 Burgundy has quit (Ping timeout: 258 seconds)
 263 2011-11-11 10:30:23 iocor has joined
 264 2011-11-11 10:31:05 Beremat has quit (Read error: Connection reset by peer)
 265 2011-11-11 10:35:35 gjs278 has joined
 266 2011-11-11 10:36:53 RazielZ has quit ()
 267 2011-11-11 10:46:48 chrisb__ has joined
 268 2011-11-11 10:46:53 enquirer has quit (Quit: back soon)
 269 2011-11-11 10:56:12 iocor has quit (Quit: Computer has gone to sleep.)
 270 2011-11-11 11:08:35 iocor has joined
 271 2011-11-11 11:10:31 amiller has quit (Ping timeout: 248 seconds)
 272 2011-11-11 11:19:47 RazielZ has joined
 273 2011-11-11 11:29:24 Cusipzzz has joined
 274 2011-11-11 11:29:25 Cusipzzz has quit (Client Quit)
 275 2011-11-11 11:35:33 wooy has left ()
 276 2011-11-11 11:40:05 <CIA-89> bitcoin: Con Kolivas * rd5231ea57f59 cgminer/ (Makefile.am configure.ac): Fix build error with -ldl being passed to LDFLAGS instead of LDADD and update configure output.
 277 2011-11-11 11:40:06 <CIA-89> bitcoin: Con Kolivas * r1419719ce579 cgminer/NEWS: Update NEWS.
 278 2011-11-11 11:46:54 marf_away has joined
 279 2011-11-11 11:49:29 fgdfgfg has quit (Read error: Connection reset by peer)
 280 2011-11-11 11:49:41 fgdfgfg has joined
 281 2011-11-11 12:13:41 iocor has quit (Quit: Computer has gone to sleep.)
 282 2011-11-11 12:21:07 vragnaroda is now known as not_vragnaroda
 283 2011-11-11 12:22:07 not_vragnaroda is now known as vragnaroda
 284 2011-11-11 12:22:26 b4epoche_ has quit (Quit: Computer has gone to sleep.)
 285 2011-11-11 12:22:37 Diablo-D3 has joined
 286 2011-11-11 12:24:20 b4epoch__ has joined
 287 2011-11-11 12:25:21 b4epoc___ has joined
 288 2011-11-11 12:25:21 b4epoch__ has quit (Read error: Connection reset by peer)
 289 2011-11-11 12:26:15 b4epoche_ has joined
 290 2011-11-11 12:26:16 iocor has joined
 291 2011-11-11 12:28:22 disq has quit (Ping timeout: 240 seconds)
 292 2011-11-11 12:29:43 b4epoc___ has quit (Ping timeout: 258 seconds)
 293 2011-11-11 12:30:50 storrgie has quit (Ping timeout: 255 seconds)
 294 2011-11-11 12:30:52 b4epoche_ has quit (Ping timeout: 258 seconds)
 295 2011-11-11 12:32:42 Litt has quit ()
 296 2011-11-11 12:40:04 disq has joined
 297 2011-11-11 12:40:05 disq has quit (Changing host)
 298 2011-11-11 12:40:05 disq has joined
 299 2011-11-11 12:42:36 ElectRo` has quit (Read error: Connection reset by peer)
 300 2011-11-11 12:42:44 ElectRo` has joined
 301 2011-11-11 12:45:00 OneFixt_ has joined
 302 2011-11-11 12:47:34 OneFixt has quit (Ping timeout: 240 seconds)
 303 2011-11-11 12:47:58 urstroyer has joined
 304 2011-11-11 12:49:16 shadders has quit (Ping timeout: 276 seconds)
 305 2011-11-11 12:55:04 larsivi has quit (Ping timeout: 260 seconds)
 306 2011-11-11 12:59:58 fgdfgfg has quit ()
 307 2011-11-11 13:06:13 larsivi has joined
 308 2011-11-11 13:13:08 RazielZ has quit (Quit: Leaving)
 309 2011-11-11 13:14:43 cryptoxchange has quit (Read error: Connection reset by peer)
 310 2011-11-11 13:16:08 cryptoxchange has joined
 311 2011-11-11 13:19:48 RazielZ has joined
 312 2011-11-11 13:25:22 iocor has quit (Quit: Computer has gone to sleep.)
 313 2011-11-11 13:27:18 <tcatm> anyone interested in 32GB historic orderbook data? :)
 314 2011-11-11 13:27:58 glitch-mod has joined
 315 2011-11-11 13:30:19 erus` has joined
 316 2011-11-11 13:32:34 glitch-mod has quit (Ping timeout: 245 seconds)
 317 2011-11-11 13:41:41 marf_away has quit (Read error: Connection reset by peer)
 318 2011-11-11 13:49:22 agricocb has quit (Quit: Leaving.)
 319 2011-11-11 13:50:46 marf_away has joined
 320 2011-11-11 13:52:59 <Eliel> tcatm: from different bitcoin exchanges?
 321 2011-11-11 13:53:15 <cjdelisle> tactm yes I am
 322 2011-11-11 13:53:27 <cjdelisle> even if I don't use it for anything, it's too cool not to have :)
 323 2011-11-11 13:57:47 <tcatm> Eliel: yes
 324 2011-11-11 14:04:07 iocor has joined
 325 2011-11-11 14:04:15 imsagodII is now known as imsaguy2
 326 2011-11-11 14:04:58 Litt has joined
 327 2011-11-11 14:12:38 amtal has quit (Ping timeout: 260 seconds)
 328 2011-11-11 14:13:06 agricocb has joined
 329 2011-11-11 14:19:54 iocor has quit (Quit: Computer has gone to sleep.)
 330 2011-11-11 14:24:03 marf_away has quit (Read error: Connection reset by peer)
 331 2011-11-11 14:24:17 iocor has joined
 332 2011-11-11 14:25:47 Zarutian has joined
 333 2011-11-11 14:26:43 Rav3nSw0rd has joined
 334 2011-11-11 14:33:32 roconnor_ has joined
 335 2011-11-11 14:37:44 b4epoche_ has joined
 336 2011-11-11 14:39:41 larsivi has quit (Ping timeout: 240 seconds)
 337 2011-11-11 14:43:58 cryptoxchange has quit (Quit: Leaving)
 338 2011-11-11 14:47:15 b4epoche_ has quit (Quit: Textual IRC Client: http://www.textualapp.com/)
 339 2011-11-11 14:52:14 batouzo has joined
 340 2011-11-11 14:56:51 marf_away has joined
 341 2011-11-11 15:04:14 <Eliel> tcatm: I'd like that too :)
 342 2011-11-11 15:05:56 <tcatm> so we need a reliable way to transfer 32GB of data quickly
 343 2011-11-11 15:06:13 <justmoon> tcatm: BitTorrent maybe?
 344 2011-11-11 15:06:41 <helo> tcatm: from how many machines to how many machines?
 345 2011-11-11 15:07:01 <tcatm> not an option. I don't want to install bittorrent on the server
 346 2011-11-11 15:07:06 btc_novice has joined
 347 2011-11-11 15:07:24 <tcatm> helo: looks like 1 -> 2
 348 2011-11-11 15:07:45 <batouzo> tcatm: rsync?
 349 2011-11-11 15:07:46 <justmoon> tcatm, so the file only exists on your server right now?
 350 2011-11-11 15:08:01 <tcatm> justmoon: yep
 351 2011-11-11 15:08:36 <justmoon> tcatm: single file? maybe http is easiest/safest then?
 352 2011-11-11 15:08:49 <tcatm> single file, just a sql dump
 353 2011-11-11 15:09:34 <helo> i'd daisy chain A -> B -> C, using netcat on B into tee, one stream going to disk, the other going out a 2nd nic to C's netcat ;)
 354 2011-11-11 15:09:50 <justmoon> lol
 355 2011-11-11 15:10:02 <justmoon> elegant though ^^
 356 2011-11-11 15:10:20 <helo> that would avoid two copies clogging the same network
 357 2011-11-11 15:11:33 <tcatm> I could probably compress the data a lot
 358 2011-11-11 15:12:02 <justmoon> I could offer downloading from the server to my seedbox via http, then seeding it (100 MBit) via BitTorrent
 359 2011-11-11 15:12:08 eueueue has joined
 360 2011-11-11 15:12:22 <justmoon> it would be worth it if more people show up that want it
 361 2011-11-11 15:12:30 <justmoon> i.e. you don't have to send it to everybody manually
 362 2011-11-11 15:12:48 iocor has quit (Quit: Computer has gone to sleep.)
 363 2011-11-11 15:12:49 erle- has quit (Quit: erle-)
 364 2011-11-11 15:13:01 acdb has joined
 365 2011-11-11 15:15:00 spitefulpretzel has joined
 366 2011-11-11 15:15:28 <justmoon> you can also ftp up it to my seedbox as well if you prefer
 367 2011-11-11 15:15:36 <justmoon> offer stands, let me know what you decide to do
 368 2011-11-11 15:15:58 <tcatm> k. let's wait if there are enough people interested until tomorrow
 369 2011-11-11 15:16:06 <justmoon> tcatm, k
 370 2011-11-11 15:16:11 ThomasV has quit (Quit: Quitte)
 371 2011-11-11 15:16:32 ThomasV has joined
 372 2011-11-11 15:17:56 justmoon has left ("Leaving")
 373 2011-11-11 15:18:00 ByronJohnson has quit (Ping timeout: 248 seconds)
 374 2011-11-11 15:18:01 justmoon has joined
 375 2011-11-11 15:18:41 ByronJohnson has joined
 376 2011-11-11 15:19:33 <CIA-89> bitcoin: Kamil Domanski * rfde2b0b0280c gentoo/net-p2p/bitcoin-qt/ (bitcoin-qt-9999.1.ebuild bitcoin-qt-9999.ebuild): net-p2p/bitcoin-qt: propagated the last linguas fix
 377 2011-11-11 15:19:39 <CIA-89> bitcoin: Kamil Domanski * r57e586fa1e41 gentoo/net-p2p/bitcoin-qt/ (3 files): net-p2p/bitcoin-qt: removed useless ewarn
 378 2011-11-11 15:26:32 iocor has joined
 379 2011-11-11 15:32:23 gavinandresen has joined
 380 2011-11-11 15:32:52 copumpkin has quit (Quit: Computer has gone to sleep.)
 381 2011-11-11 15:35:07 <gavinandresen> good morning y'all
 382 2011-11-11 15:35:53 <CIA-89> bitcoinjs/node-bitcoin-p2p: Stefan Thomas master * r6380d90 / lib/schema/transaction.js : Added helper to set outpoint index. - http://git.io/qDV8ZQ
 383 2011-11-11 15:35:53 <CIA-89> bitcoinjs/node-bitcoin-p2p: Stefan Thomas master * r86861dd / native.cc : Improved memory handling for BitcoinKey. - http://git.io/1nzp9w
 384 2011-11-11 15:35:54 <CIA-89> bitcoinjs/node-bitcoin-p2p: Stefan Thomas master * rc86d5d6 / native.cc : Fixed segfault - cannot call Buffer::* methods from EIO thread. - http://git.io/ZKqeuA
 385 2011-11-11 15:36:07 <justmoon> that's my way of saying good morning back :)
 386 2011-11-11 15:37:51 <Eliel> looks like the number of transactions per day for bitcoin is growing exponentially. 9 months means 10 times more transactions.
 387 2011-11-11 15:38:58 <ThomasV> Eliel: it is in a plateau now
 388 2011-11-11 15:39:25 <ThomasV> http://blockchain.info/charts/n-transactions
 389 2011-11-11 15:39:57 <Eliel> it's increasing again, you can see it more clearly if you turn on smoothing
 390 2011-11-11 15:41:09 * roconnor_ discovers litecoin
 391 2011-11-11 15:42:25 <Eliel> scrypt works quite well at keeping GPUs from overpowering CPUs in hashing :)
 392 2011-11-11 15:43:15 <roconnor_> After reading about scrypt it seemed like a better use for bitcoin hashing for more or less the same reasons it is better for password protection
 393 2011-11-11 15:43:21 <roconnor_> I'm glad someone implemented it
 394 2011-11-11 15:43:37 <justmoon> I still don't get that to be honest
 395 2011-11-11 15:44:12 <justmoon> how do the big supercomputers fare with something like scrypt? better or worse than sha?
 396 2011-11-11 15:44:30 btc_novice1 has joined
 397 2011-11-11 15:44:31 <Eliel> the only practical issue with it scrypt for hashing algo is that botnets can potentially make up a bigger portion of the network than expected.
 398 2011-11-11 15:44:44 <roconnor_> I wonder if the litecoin people also fixed that moderate bug involving timestamp differencing during difficulty updating
 399 2011-11-11 15:45:10 <iddo> roconnor_: yes it's fixed
 400 2011-11-11 15:45:17 <roconnor_> iddo: in litecoin?
 401 2011-11-11 15:45:17 <Eliel> justmoon: I'd expect the current supercomputers to work better at litecoin hashing than bitcoin hashing.
 402 2011-11-11 15:45:32 <Eliel> because GPUs don't have any edge with litecoin.
 403 2011-11-11 15:45:55 btc_novice has quit (Ping timeout: 245 seconds)
 404 2011-11-11 15:46:01 eastender has joined
 405 2011-11-11 15:46:12 <Eliel> of course, a GPU supercomputer would not have that problem with even bitcoin.
 406 2011-11-11 15:46:15 <iddo> roconnor_: yes you cab see link to github source from https://github.com/coblee/litecoin/wiki/Comparison-between-Bitcoin-and-Litecoin
 407 2011-11-11 15:47:12 <iddo> s/cab/can
 408 2011-11-11 15:47:14 BlueMatt has joined
 409 2011-11-11 15:47:32 <Eliel> anyway, if the number of transactions keeps going up this fast, we'll be approaching the current limit on 18 months.
 410 2011-11-11 15:47:46 minimoose has joined
 411 2011-11-11 15:47:56 <roconnor_> iddo: thanks!
 412 2011-11-11 15:47:57 <Eliel> how much time is needed to prepare for increasing the limit?
 413 2011-11-11 15:49:06 <iddo> increasing the limit risks chain fork? or not?
 414 2011-11-11 15:49:09 <gavinandresen> Eliel: we need exactly 111 days to prepare.
 415 2011-11-11 15:49:31 <gavinandresen> ... and we should start now, because it is 11/11/11
 416 2011-11-11 15:49:38 <Eliel> :D
 417 2011-11-11 15:49:43 <BlueMatt> gavinandresen: note on your creation of a encryptionbug branch, sipa also just created a resilvering branch
 418 2011-11-11 15:49:46 <BlueMatt> (why duplicate effort?)
 419 2011-11-11 15:50:00 <gavinandresen> sipa had to go to sleep
 420 2011-11-11 15:50:14 agricocb has quit (Quit: Leaving.)
 421 2011-11-11 15:50:20 <Eliel> how come exactly 111 days are needed?
 422 2011-11-11 15:50:27 <gavinandresen> because 11 is my favorite number
 423 2011-11-11 15:50:35 <Eliel> :D
 424 2011-11-11 15:50:40 <BlueMatt> gavinandresen: fair enough, just wanted to make sure we avoid duplicating effort
 425 2011-11-11 15:51:13 <gavinandresen> BlueMatt: I'm about to sanity test what I HOPE will be a final fix, by the way-- just launching the GUI now to make sure it acts sanely when encrypting a wallet.
 426 2011-11-11 15:52:06 <BlueMatt> mmm, nice
 427 2011-11-11 15:52:11 <iddo> if i understood correctly, wallet encryption bdb bug will also happen if you create new wallet in 0.4.0 because the client first populates 100 keypairs and saves wallet.dat, and only then allows you to encrypt it?
 428 2011-11-11 15:52:37 <gavinandresen> Hard to believe so many smart people (me included) didn't catch it before the 0.4 release....sigh
 429 2011-11-11 15:53:53 <BlueMatt> iddo: yep
 430 2011-11-11 15:54:19 <BlueMatt> gavinandresen: yea, and the worst part is Im pretty sure at one point or another it was on the todo list to check to make sure this exact bug didnt exist
 431 2011-11-11 15:55:04 <Eliel> gavinandresen: I think the reason no core developers catched it is that all of you are reasonably familiar with bsddb and how to use the APIs for it. The person who found it, found it because he wasn't and had hacked his own solution that does the bare minimum :)
 432 2011-11-11 15:56:02 iocor has quit (Quit: Computer has gone to sleep.)
 433 2011-11-11 15:57:45 BlueMatt has quit (Quit: Ex-Chat)
 434 2011-11-11 15:58:01 <gmaxwell> I expected keys to be left on disk, but not in the wallet file. But I think because of this I didn't bother hunting for leaked keys because I _expected_ leaked keys.
 435 2011-11-11 15:58:08 BlueMatt_ has joined
 436 2011-11-11 15:58:28 <Eliel> it's even possible it was checked but for some reason it didn't occur in the check.
 437 2011-11-11 15:58:41 <BlueMatt_> gavinandresen: yea, and the worst part is Im pretty sure at one point or another it was on the todo list to check to make sure this exact bug didnt exist
 438 2011-11-11 15:58:50 <BlueMatt_> oh, that did go out...
 439 2011-11-11 15:58:52 BlueMatt_ is now known as BlueMatt
 440 2011-11-11 15:58:57 <gavinandresen> worth saying twice.
 441 2011-11-11 15:59:31 <gmaxwell> Eliel: I'm pretty sure I didn't check for it.
 442 2011-11-11 15:59:37 eueueue has quit (Quit: Page closed)
 443 2011-11-11 15:59:42 <BlueMatt> heh
 444 2011-11-11 16:00:00 <gmaxwell> I think I might have checked that the wallet did grow a bunch.. but there are all sorts of reasons that might fail to find it.
 445 2011-11-11 16:00:26 p0s has joined
 446 2011-11-11 16:00:31 copumpkin has joined
 447 2011-11-11 16:00:36 <gmaxwell> I was checking the size just to make sure that _changing_ the password didn't result in crazy duplication.
 448 2011-11-11 16:02:03 <BlueMatt> oh, you know, we should probably resilver the file after pass change
 449 2011-11-11 16:02:14 <BlueMatt> (because otherwise the old pass might be half-valid)
 450 2011-11-11 16:02:37 btc_novice1 has left ()
 451 2011-11-11 16:02:59 <Eliel> gavinandresen: do you think it might make sense to have the block size limit grow dynamically a little each block?
 452 2011-11-11 16:03:16 <sipa> BlueMatt: not really duplicate effort, gavin worked further on what i had :)
 453 2011-11-11 16:03:25 <Eliel> like based on the transaction volume in the past 2000 or so blocks.
 454 2011-11-11 16:03:53 <sipa> Eliel: i think we first need a better view on how we'll deal with transaction fees
 455 2011-11-11 16:03:57 <sipa> before we touch the block size
 456 2011-11-11 16:04:24 <gmaxwell> Eliel: personally I'd like to lower the block size. We've had quite a few blocks filled with hundreds of KB of garbage transactions. :)
 457 2011-11-11 16:04:36 zapnap has joined
 458 2011-11-11 16:04:44 <Eliel> gmaxwell: for the time being, if this got implemented, it'd likely reduce the maximum a lot.
 459 2011-11-11 16:05:07 <gmaxwell> Eliel: if you go through the log here you'll see some discussion about that about two months ago.
 460 2011-11-11 16:05:16 <Eliel> ... although, if someone starts spamming, they can increase the limit that way
 461 2011-11-11 16:05:24 <gmaxwell> But I think its a safe bet that there wouldn't be any fork creating change unless there was an urgent need.
 462 2011-11-11 16:05:27 SomeoneWeird has quit (Quit: Leaving)
 463 2011-11-11 16:05:51 <Eliel> OP_EVAL will be a forking change, no?
 464 2011-11-11 16:06:23 erle- has joined
 465 2011-11-11 16:06:29 <gmaxwell> No.
 466 2011-11-11 16:06:41 cebaik has joined
 467 2011-11-11 16:06:43 Sedra- has joined
 468 2011-11-11 16:06:45 <gmaxwell> It's very carefully designed to not be one.
 469 2011-11-11 16:07:17 <Eliel> I thought old clients would reject transactions spending the coins from an OP_EVAL input.
 470 2011-11-11 16:07:21 <cebaik> halooo
 471 2011-11-11 16:07:21 <gmaxwell> Eliel: well, a key point is that the people controlling the hash power are probably not the spammers. Rather, some are just not digilent in maintaining their transaction policy to prevent spam.
 472 2011-11-11 16:07:38 <gmaxwell> Eliel: no, thats not correct. (and the BIP is pretty clear about that, have you read it?)
 473 2011-11-11 16:07:50 <Eliel> I did, it's a while ago though.
 474 2011-11-11 16:07:56 <Eliel> I'll reread
 475 2011-11-11 16:08:00 <gmaxwell> Eliel: old nodes see OP_EVAL as "anyone can spend this with no requirements!"
 476 2011-11-11 16:08:05 traviscj has quit (Remote host closed the connection)
 477 2011-11-11 16:08:14 cebaik has left ()
 478 2011-11-11 16:08:18 BlueMatt has quit (Ping timeout: 258 seconds)
 479 2011-11-11 16:08:21 <gmaxwell> (er, actually "anyone who provides the right input script but no other requirements")
 480 2011-11-11 16:09:17 Sedra has quit (Ping timeout: 240 seconds)
 481 2011-11-11 16:09:29 BlueMatt has joined
 482 2011-11-11 16:10:02 <Eliel> ah, yes, it's said quite clearly.
 483 2011-11-11 16:10:08 <Eliel> I wonder if I read an old version before.
 484 2011-11-11 16:10:10 <BlueMatt> gavinandresen: also make sure you overwrite log files with /dev/urandom
 485 2011-11-11 16:10:12 <iddo> if there's no fork then why the BIP says it will be rejected if less than 50% agree to update?
 486 2011-11-11 16:11:49 <gmaxwell> iddo: Thats unrelated.
 487 2011-11-11 16:12:07 <gavinandresen> BlueMatt: no, I'm just making sure log files are removed.  Overwriting is out of scope (and may not work, anyway)
 488 2011-11-11 16:12:08 <gmaxwell> iddo: the op_eval transactions are construted so that all nodes will already accept them.
 489 2011-11-11 16:12:20 <gmaxwell> (well except that they're non-standard)
 490 2011-11-11 16:13:05 <BlueMatt> gavinandresen: for most filesystems, overwriting will work 99% which is good enough, doing nothing is not really...
 491 2011-11-11 16:13:36 <BlueMatt> gavinandresen: are you not already overwriting wallet_old.dat with /dev/urandom?
 492 2011-11-11 16:13:45 <sipa> no
 493 2011-11-11 16:13:47 <gavinandresen> BlueMatt: if somebody has block access to your disks then you're screwed.
 494 2011-11-11 16:14:06 <sipa> you can argue that optimistically overwriting is better than nothin
 495 2011-11-11 16:14:08 <BlueMatt> for most filesystems, if you overwrite with random data, you are fine
 496 2011-11-11 16:14:13 <sipa> but peole shouldn't rely on it
 497 2011-11-11 16:14:23 <BlueMatt> yea, but if its better than nothing, we should do it...
 498 2011-11-11 16:15:43 <Eliel> I'd be in favor of the GUI giving the user information on how to clean the empty space on the hard disk.
 499 2011-11-11 16:15:53 <Eliel> perhaps even offer to do it, it's not a complex operation.
 500 2011-11-11 16:15:55 <gavinandresen> BlueMatt: I disagree.  I think it makes more sense to spend time on more effective security, like figuring out how to make OP_EVAL multidevice wallets work...
 501 2011-11-11 16:16:14 <gavinandresen> ... or implementing a good deterministic key solution
 502 2011-11-11 16:17:26 <gmaxwell> 07:59 < gavinandresen> BlueMatt: if somebody has block access to your disks then you're screwed.
 503 2011-11-11 16:17:39 btc_novice has joined
 504 2011-11-11 16:17:50 <BlueMatt> gavinandresen: or just born-encrypted wallets
 505 2011-11-11 16:17:53 <gmaxwell> It's not really that simple.... I sell one of my old computers to someone, for example.
 506 2011-11-11 16:18:12 <BlueMatt> gavinandresen: but since it takes like 20 seconds to write code to overwrite wallet.dat and the log files, it is totally worth it
 507 2011-11-11 16:18:14 <sipa> BlueMatt: flushing the keypool at encryption is equivalent to born-encrypted
 508 2011-11-11 16:18:27 <BlueMatt> sipa: is that what is being done?
 509 2011-11-11 16:18:34 <sipa> yes, that is the intent
 510 2011-11-11 16:18:35 <pierre`> gmaxwell: remove the disk and destroy them
 511 2011-11-11 16:18:41 <BlueMatt> sipa: fair enough
 512 2011-11-11 16:18:58 <gmaxwell> pierre`: yea, and what if I don't becuase I have _no idea_ that my long time encrypted wallet could be exposed?
 513 2011-11-11 16:19:07 <gavinandresen> yes, I'm working on code right now to correctly flush the keypool.
 514 2011-11-11 16:19:08 <BlueMatt> sipa: so new scene is to keep all existing privkeys, but flush pool?
 515 2011-11-11 16:19:15 <BlueMatt> scheme*
 516 2011-11-11 16:19:23 <sipa> gavinandresen: i wonder why your code doesn't work
 517 2011-11-11 16:19:24 <gmaxwell> Obviously the user can be secure, but they shouldn't have to know everything you and I know in order to have a fighting chance. :)
 518 2011-11-11 16:19:38 <pierre`> gmaxwell: dou you realize the swap partition (or the swap file in windows) contains ssh keys, passwords et decrypted stuff ?
 519 2011-11-11 16:19:40 eastender has quit (Ping timeout: 245 seconds)
 520 2011-11-11 16:19:41 <BlueMatt> hence why I say overwrite log files
 521 2011-11-11 16:19:48 <gmaxwell> BlueMatt: mark all the pool keys as used. .. keep them... but don't reuse them. I think thats pretty nice.
 522 2011-11-11 16:19:51 <BlueMatt> pierre`: thats why we mlock properly
 523 2011-11-11 16:19:57 <BlueMatt> gmaxwell: agreed
 524 2011-11-11 16:20:14 <BlueMatt> gmaxwell: but you should also overwrite on disk (if possible)
 525 2011-11-11 16:20:19 <gmaxwell> pierre`: What BlueMatt said. And thats a more obvious vulnerablity.
 526 2011-11-11 16:20:44 <gmaxwell> Absolutely, I use fully encrypted volumes in any case. But joe user would be really surprised if his encrypted wallet got exposed.
 527 2011-11-11 16:21:06 <gmaxwell> (also, highly recommended:
 528 2011-11-11 16:21:07 <gmaxwell> $ cat /etc/crypttab
 529 2011-11-11 16:21:07 <gmaxwell> swap /dev/sda3 /dev/urandom swap,cipher=aes-lrw-plain,size=256
 530 2011-11-11 16:21:09 <gmaxwell> )
 531 2011-11-11 16:21:32 <BlueMatt> or just no swap :)
 532 2011-11-11 16:22:09 <gmaxwell> BlueMatt: tmp on tmpfs, and use the space you would have used as /tmp as swap.
 533 2011-11-11 16:22:16 eastender has joined
 534 2011-11-11 16:22:39 <BlueMatt> gmaxwell: I prefer no swap, but thats just because I have enough mem to not have to care
 535 2011-11-11 16:22:57 <CIA-89> bitcoinjs/node-bitcoin-p2p: Stefan Thomas master * r3b18cdc / test/storage.js : Added test case for querying for a non-existent block. - http://git.io/3HrJ-A
 536 2011-11-11 16:22:57 <CIA-89> bitcoinjs/node-bitcoin-p2p: Stefan Thomas master * r1f40b5a / (lib/storage.js lib/db/leveldb/storage.js): Added LevelDB storage backend. - http://git.io/gaqpBQ
 537 2011-11-11 16:23:19 <sipa> 8GB ought to be enough for everyone
 538 2011-11-11 16:23:49 <AliciaC> "no one will ever need more than 1MB" :P
 539 2011-11-11 16:23:55 <BlueMatt> (well I hit oom issues when running too many vms, but other than that, 8G is plenty)
 540 2011-11-11 16:23:57 wasabi has joined
 541 2011-11-11 16:24:08 <BlueMatt> (miss my 12g desktop though)
 542 2011-11-11 16:24:11 <gmaxwell> BlueMatt: tmpfs visibly improves system performance (lots of silly shell scripts use it) but I don't think that I'd feel safe with only a handful of gigs available for ram/tmp.
 543 2011-11-11 16:24:30 <cocktopus> 48GB ftw
 544 2011-11-11 16:24:58 <gmaxwell> $ free -g total       used       free     shared    buffers     cached
 545 2011-11-11 16:24:58 <gmaxwell> Mem:            63          5         57          0          0          3
 546 2011-11-11 16:25:47 <BlueMatt> anyway, back to wallet talk
 547 2011-11-11 16:26:07 <BlueMatt> 11:11 11/11/11
 548 2011-11-11 16:26:13 <cocktopus> it is that
 549 2011-11-11 16:26:43 <iddo> gmaxwell: 50% reject in BIP 0012 is unrelated to OP_EVAL being risk-free from forks? what is it related to then?
 550 2011-11-11 16:26:44 <gmaxwell> In any case, the purging/refilling of the pool is basically as good as born encrypted. That closes off a lot of problems going forward.
 551 2011-11-11 16:26:46 batouzo has quit (Ping timeout: 259 seconds)
 552 2011-11-11 16:27:15 <gmaxwell> iddo: Yes. It's related to OP_EVAL being useful as OP_EVAL.
 553 2011-11-11 16:27:32 <BlueMatt> yea, but you still need to overwrite wallet_old.dat and logs with urandom
 554 2011-11-11 16:27:43 <BlueMatt> or 0, or 1, or, or, or
 555 2011-11-11 16:28:04 <iddo> gmaxwell: ahh so just if 50% don't care about it then it's better to reject?
 556 2011-11-11 16:28:08 <gmaxwell> BlueMatt: no, not for a new wallet at least. (Though I agree that it should be done to cover the other cases)
 557 2011-11-11 16:28:21 <gmaxwell> iddo: ...
 558 2011-11-11 16:29:05 <iddo> ok i guess i should read the wiki again
 559 2011-11-11 16:29:11 <gmaxwell> iddo: To nodes not supporting OP_EVAL the OP_EVAL txn look like insecure txn. If the network rejects OP_EVAL then you'd be a fool to use it because those transactions would simply be insecure.
 560 2011-11-11 16:29:15 <BlueMatt> gmaxwell: well yea
 561 2011-11-11 16:29:31 <BlueMatt> gmaxwell: but you never know if the user already posted their old address to mtgox and hasnt gotten a payment yet
 562 2011-11-11 16:29:48 <BlueMatt> gmaxwell: so as long as you keep old privkeys around for future use, you should make sure to overwrite wallet+logs
 563 2011-11-11 16:29:54 <gmaxwell> Well, and not just that— they may have widely circulated donation addresses and such that can't be changed.
 564 2011-11-11 16:29:57 <iddo> ah
 565 2011-11-11 16:30:01 Sedra has joined
 566 2011-11-11 16:30:58 <BlueMatt> yea
 567 2011-11-11 16:30:58 <BlueMatt> same scenario...
 568 2011-11-11 16:31:13 <BlueMatt> gavinandresen: I highly disagree that this would be outside the scope of this bug
 569 2011-11-11 16:31:18 <BlueMatt> in fact, overwriting is the bug
 570 2011-11-11 16:31:35 Sedra- has quit (Ping timeout: 276 seconds)
 571 2011-11-11 16:31:46 <gavinandresen> no, the fact that an 'encrypted' wallet.dat that you might think you can just backup as-is has private keys in it unencrypted is the bug
 572 2011-11-11 16:32:10 <BlueMatt> but why not overwrite now?
 573 2011-11-11 16:32:15 <gmaxwell> I agree with gavin. (though would still prefer the wiping— just like we're fixing the keypool too)
 574 2011-11-11 16:32:16 Davincij15 has quit ()
 575 2011-11-11 16:32:37 <gavinandresen> (I agree with gmaxwell that discarding an old computer is also a threat that is worth thinking about)
 576 2011-11-11 16:32:43 <BlueMatt> or here, let me file a new security bug report: basic expectations that bitcoin sanely tries to remove keys from drive are not taken
 577 2011-11-11 16:33:07 <gavinandresen> sipa:  did you do any work yesterday trying to overwrite?  bdb seems to really like to hang onto the files it is using....
 578 2011-11-11 16:33:20 agricocb has joined
 579 2011-11-11 16:33:26 <gavinandresen> (so the why not is "because it isn't only 20 seconds")
 580 2011-11-11 16:33:41 <sipa> gavinandresen: the first attempt at resilver tried to manually move files around, but bdb really doesn't like that
 581 2011-11-11 16:33:51 <luke-jr> BlueMatt: that's not a security bug, since it is a known impossibility for Bitcoin to ensure they are removed.
 582 2011-11-11 16:34:12 <BlueMatt> luke-jr: its a security bug because its not doing the sane basic expectation that it should be
 583 2011-11-11 16:34:14 <sipa> i believe the only way is do a rename to wallet_old.dat, and af dbenv shutdown overwrite that file
 584 2011-11-11 16:34:29 <luke-jr> BlueMatt: there is no such expectation.
 585 2011-11-11 16:34:45 <sipa> i believe there are such expectations, but they are unreasonable
 586 2011-11-11 16:34:51 <CIA-89> bitcoin: Luke Dashjr free_icons * r0fe19a3a2c7c bitcoind-personal/ (contrib/debian/copyright src/qt/res/icons/overview.png): Replace kfm_home with go-home for "overview"
 587 2011-11-11 16:34:55 <BlueMatt> sipa: I thought that is the way it was being done (resilver at shutdown)
 588 2011-11-11 16:34:55 <BlueMatt> luke-jr: shut up
 589 2011-11-11 16:35:00 <gmaxwell> sipa: they're also not going away.
 590 2011-11-11 16:35:02 <sipa> we can do what we can, but at no point try to claim we prevent all leaking of keys
 591 2011-11-11 16:35:03 <luke-jr> BlueMatt: no u
 592 2011-11-11 16:35:20 <BlueMatt> sipa: we arent, but the exectation is that you try, and currently we arent even doing that
 593 2011-11-11 16:35:28 <gavinandresen> Also, overwriting the transaction log and the old wallet.dat file are not sufficient-- I found private keys unencrypted in other files that are completely controlled by bdb
 594 2011-11-11 16:35:33 <gmaxwell> It's really surprising that you could _encrypt_ your wallet and people could later get your keys off your disk, and we'll never really manage to educate on that.
 595 2011-11-11 16:35:51 <sipa> BlueMatt: well i agree that we can try, under the condition that at no point we give the impression that we prevent all key leakage
 596 2011-11-11 16:35:55 <BlueMatt> gavinandresen: in which files?
 597 2011-11-11 16:35:56 <luke-jr> what I *would* expect of Bitcoin clients, is that if the wallet is encrypted, to only handle the keys in secured/locked memory so they never touch swap
 598 2011-11-11 16:36:01 <BlueMatt> sipa: true
 599 2011-11-11 16:36:02 <gavinandresen> __db.00N
 600 2011-11-11 16:36:02 <luke-jr> is that the case?
 601 2011-11-11 16:36:23 <gmaxwell> what sipa said. We should _try_ but we shouldn't bother telling the user that we're doing this. We should encourage them to think otherwise, because we can't promise.
 602 2011-11-11 16:36:25 <BlueMatt> gavinandresen: mmm, well I suppose those need overwritten/deleted then too
 603 2011-11-11 16:36:27 <sipa> luke-jr: a serious attempt is made for that
 604 2011-11-11 16:36:34 <BlueMatt> gmaxwell: yea, thats what Im saying
 605 2011-11-11 16:36:35 <sipa> but it's impossible to guarantee
 606 2011-11-11 16:36:43 <sipa> unless there entire application is mlocked
 607 2011-11-11 16:36:50 <BlueMatt> whereas everyone here seems to be saying we can try but we'll do it some time in the future, not now...
 608 2011-11-11 16:36:57 <gmaxwell> mlocking a big gui application is basically a non-starter.
 609 2011-11-11 16:37:18 <luke-jr> BlueMatt: it's possibly a bug, but not an immediate-urgency critical security flaw
 610 2011-11-11 16:37:21 <gavinandresen> Shutting down bitcoin is ALREADY insanely slow.  And overwriting once is almost certainly insufficient-- if you've been running bitcoin for a while there are almost certainly pieces of old transaction logs marked erased on your hD
 611 2011-11-11 16:37:22 <luke-jr> IMO
 612 2011-11-11 16:37:34 <gmaxwell> if you do mlockall MCL_FUTURE then the program will crash if any mlocking fails.
 613 2011-11-11 16:37:50 <sipa> i agree with luke-jr
 614 2011-11-11 16:38:14 <BlueMatt> gavinandresen: currently, bdb wont delete any parts of the logs until shutdown...
 615 2011-11-11 16:38:26 <sipa> BlueMatt: see gavin's branch
 616 2011-11-11 16:38:35 <sipa> it does delete logs and region files
 617 2011-11-11 16:38:38 <luke-jr> can't we "shutdown" bdb without shutting down the client, and start it back up?
 618 2011-11-11 16:38:42 <gmaxwell> BlueMatt: which is a halarious bug during initial syncup since it increase peak usage by an extra gig.
 619 2011-11-11 16:39:20 <sipa> luke-jr: if no database handles are present, yes
 620 2011-11-11 16:39:24 <BlueMatt> gmaxwell: yep
 621 2011-11-11 16:39:39 <iddo> with "born encrypted" wallet, will there be unencrypted private keys in log files on disk?
 622 2011-11-11 16:40:00 <sipa> iddo: possibly, yes, but not of keys that you'll ever use
 623 2011-11-11 16:40:14 <sipa> you get a new set of 100 reserve keys that never touched disk in unencrypted form
 624 2011-11-11 16:40:30 <iddo> that's good
 625 2011-11-11 16:41:16 <luke-jr> sipa: that's only after this fix, right?
 626 2011-11-11 16:41:23 <sipa> yes
 627 2011-11-11 16:41:27 <gmaxwell> luke-jr: right.
 628 2011-11-11 16:41:33 CaptainDDL has joined
 629 2011-11-11 16:41:35 jrmithdobbs has quit (Quit: quit)
 630 2011-11-11 16:42:14 <gmaxwell> One metric I like to use for these tradeoffs in my own software is that, admitting it's /possible/ users could be harmed as the result of a decision— how stupid will I feel(/look) if it happens?  It's an outlook that encourages fixing easily fixed/foreseen problems even if they're unlikely.
 631 2011-11-11 16:43:06 jrmithdobbs has joined
 632 2011-11-11 16:44:35 merde has quit ()
 633 2011-11-11 16:45:48 d4de has quit (Ping timeout: 258 seconds)
 634 2011-11-11 16:45:53 eueueu has joined
 635 2011-11-11 16:46:36 <iddo> should the announcement recommend that you create new receiving address and send the entire balance to it?
 636 2011-11-11 16:47:10 <luke-jr> iddo: iff your wallet was exposed somewhere untrustworthy
 637 2011-11-11 16:47:23 wasabi1 has joined
 638 2011-11-11 16:47:34 <iddo> right
 639 2011-11-11 16:48:11 <iddo> though if balance was controlled by many keys then it might involve a small fee
 640 2011-11-11 16:49:09 <sipa> the number of keys is irrelevant
 641 2011-11-11 16:49:13 <BlueMatt> gavinandresen: so currently, we close the db, then dont bother rewriting...instead just reopening a new db in the old file?
 642 2011-11-11 16:49:14 <BlueMatt> gavinandresen: so why cant we rewrite?
 643 2011-11-11 16:49:33 <luke-jr> perhaps someone should write a "recycle all private keys" function :p
 644 2011-11-11 16:49:49 <sipa> BlueMatt: i don't see what you mean
 645 2011-11-11 16:49:50 <luke-jr> ie, for every private key with a balance, create a new one, and send each output to it
 646 2011-11-11 16:49:53 <luke-jr> and update the address book
 647 2011-11-11 16:50:03 <BlueMatt> sipa: maybe Im not reading the code right
 648 2011-11-11 16:50:04 <luke-jr> (marking the old ones as hidden/compromised)
 649 2011-11-11 16:50:19 <sipa> BlueMatt: currently, we delete the old db, and rename the new one; through bdb all
 650 2011-11-11 16:50:19 <BlueMatt> sipa: but from what I read, you are closing the db and then just reopening a new db in the file?
 651 2011-11-11 16:50:25 <sipa> no
 652 2011-11-11 16:51:26 <BlueMatt> excuse me, you also rm the files
 653 2011-11-11 16:51:39 <BlueMatt> why cant you just change the file deletion to file overwrite then deletion
 654 2011-11-11 16:51:41 <BlueMatt> ?
 655 2011-11-11 16:52:24 dvide has joined
 656 2011-11-11 16:52:33 <sipa> we could try to overwrite the file before the db.remove() call
 657 2011-11-11 16:52:46 <sipa> hopefully bdb won't die on that
 658 2011-11-11 16:53:18 <gavinandresen> I'm still working on the basics-- getting unencrypted private keys from wallet.dat (looks like topping up the keypool is doing the Wrong Thing)
 659 2011-11-11 16:53:19 <luke-jr> know what, I'd be willing to offer a one-time no-fee unlimited-size transaction for people who are at risk of loss from this, on Eligius.
 660 2011-11-11 16:53:25 <BlueMatt> as long as you lock it, and then dont close gracefully, I dont see why not?
 661 2011-11-11 16:53:37 <luke-jr> is there a sane way to make sure people don't abuse it?
 662 2011-11-11 16:53:48 <BlueMatt> luke-jr: no
 663 2011-11-11 16:54:02 <sipa> luke-jr: i think it would be hard to prevent people from doing that again and again
 664 2011-11-11 16:54:22 <luke-jr> sipa: well, it'd be limited time offer at least
 665 2011-11-11 16:54:45 <gavinandresen> BlueMatt: here's what I'm seeing, maybe you can help spot the bug:  Rewrite wallet:  all OK, no unencrypted private keys in the new wallet.dat.
 666 2011-11-11 16:54:45 <sipa> BlueMatt: bdb dies eg. if you manually delete the file while the dbenv exists (even though the db file isn't opened)
 667 2011-11-11 16:54:45 <luke-jr> perhaps "all inputs must be older than block X" where X is when the announcement goes out
 668 2011-11-11 16:55:11 <gavinandresen> Call TopUpKeyPool:  left with 1 unencrypted private key in wallet.dat  (why 1, I don't know yet....)
 669 2011-11-11 16:55:25 <sipa> what?
 670 2011-11-11 16:55:27 <gavinandresen> (that is:  Unlock then TopUp, of course)
 671 2011-11-11 16:55:47 <sipa> gavinandresen: is that a new key, unencrypted?
 672 2011-11-11 16:56:12 <gavinandresen> I don't know.  I have a tool that just searches for byte sequences in a file, and I'm looking for any of the old private keys in the file.
 673 2011-11-11 16:56:25 <sipa> i'll try it
 674 2011-11-11 16:56:41 <gavinandresen> Hmm.... could be bdb re-uses memory as it pages to/from disk
 675 2011-11-11 16:56:46 roconnor_ is now known as roconnor
 676 2011-11-11 16:57:21 <BlueMatt> why are we not just restarting bitcoin to do this?
 677 2011-11-11 16:57:30 <BlueMatt> (I thought that was the original intent)
 678 2011-11-11 16:57:46 <BlueMatt> (also I dont think people will think it too odd if bitcoin restarts after they encrypt their wallet)
 679 2011-11-11 16:57:47 <luke-jr> hmm
 680 2011-11-11 16:58:10 <luke-jr> could also move the "encrypt wallet and resilver" code to another EXE even
 681 2011-11-11 16:58:36 <sipa> ah, don't have the tool here
 682 2011-11-11 16:58:37 <BlueMatt> not necessary (and I always have hated the idea of distributing multiple exes)
 683 2011-11-11 16:58:41 <gavinandresen> I'm going to commit my changes to far to my working branch...
 684 2011-11-11 16:59:13 <BlueMatt> anyway, I have to go...I think it should be absolutely possible to just close bdb non-cleanly and then delete all the logs/wallet.dat securely
 685 2011-11-11 16:59:40 <sipa> non-clean shutdown? bah
 686 2011-11-11 16:59:55 traviscj has joined
 687 2011-11-11 16:59:57 d4de has joined
 688 2011-11-11 17:00:00 <BlueMatt> oh, yea we still dont remove logs on shutdown do we?
 689 2011-11-11 17:00:06 <sipa> we do
 690 2011-11-11 17:00:06 <BlueMatt> yea, nevermind then you can cleanly shut it down
 691 2011-11-11 17:00:17 <BlueMatt> I thought we only did on startup (for now)
 692 2011-11-11 17:00:24 larsivi has joined
 693 2011-11-11 17:00:24 <BlueMatt> you have to avoid bdb deleting its logs
 694 2011-11-11 17:01:01 <gavinandresen> BlueMatt: we can't avoid bdb deleting its logs, that would involve time travel....
 695 2011-11-11 17:01:18 <BlueMatt> gavinandresen: and why cant we do that?
 696 2011-11-11 17:01:27 <sipa> sure we can, not set any autodelete
 697 2011-11-11 17:01:34 <sipa> it will keep all its logs
 698 2011-11-11 17:01:42 <BlueMatt> sipa: we currently dont set autodelete (but should)
 699 2011-11-11 17:01:50 <sipa> BlueMatt: see gavin's code
 700 2011-11-11 17:01:51 <BlueMatt> only on startup from what someone said yesterday
 701 2011-11-11 17:01:58 OneFixt_ has quit (Changing host)
 702 2011-11-11 17:01:58 OneFixt_ has joined
 703 2011-11-11 17:02:04 OneFixt_ is now known as OneFixt
 704 2011-11-11 17:02:06 <gavinandresen> shutdown, actually
 705 2011-11-11 17:02:36 <BlueMatt> why are we manually deleting bdb logs???
 706 2011-11-11 17:02:50 <BlueMatt> we should NEVER do that
 707 2011-11-11 17:02:56 <BlueMatt> (unless we are trying to be secure)
 708 2011-11-11 17:03:06 <Eliel> it's that unless, no?
 709 2011-11-11 17:03:08 <BlueMatt> bdb has an option to make it delete logs on shutdown
 710 2011-11-11 17:03:16 <BlueMatt> which we should use
 711 2011-11-11 17:03:35 <BlueMatt> gotta go
 712 2011-11-11 17:03:36 <sipa> BlueMatt: no, bdb only removes its "unused" logs
 713 2011-11-11 17:03:46 <sipa> and there is always some log file left it considers "in use"
 714 2011-11-11 17:03:54 <sipa> i suppose just to know where the last checkpoint was
 715 2011-11-11 17:04:07 <sipa> even if fully synced
 716 2011-11-11 17:04:31 <BlueMatt> sipa: well we dont even tell it to do that
 717 2011-11-11 17:04:34 <BlueMatt> (on shutdown)
 718 2011-11-11 17:04:44 <sipa> BlueMatt: please, see gavin's code
 719 2011-11-11 17:04:50 <sipa> he does that
 720 2011-11-11 17:04:58 <gavinandresen> Lets see, what to do, what to do...
 721 2011-11-11 17:05:04 <BlueMatt> sipa: where?
 722 2011-11-11 17:05:17 <gavinandresen> you're late for class?
 723 2011-11-11 17:05:28 <BlueMatt> gavinandresen: no, Im sitting in class after Im done...
 724 2011-11-11 17:05:35 <sipa> BlueMatt: in DBFlush()
 725 2011-11-11 17:05:44 <sipa> search for DB_ARCH_REMOVE
 726 2011-11-11 17:06:03 <gavinandresen> That was there before, by the way.  I added EnvShutdown
 727 2011-11-11 17:06:12 <sipa> oh, right, indeed
 728 2011-11-11 17:06:20 <BlueMatt> sipa: hence why I didnt see it...
 729 2011-11-11 17:06:34 <sipa> forgot it was already there
 730 2011-11-11 17:06:58 <sipa> gavinandresen: overwriting wallet.dat right before the dbA.remove call sounds possible, imho
 731 2011-11-11 17:07:48 <gavinandresen> sipa:  yes.  I just don't want to bother, since there could be keys in swap, in old logfiles that grew bigger than 10MB and were deleted, in the _db.00N files, ....
 732 2011-11-11 17:08:03 <BlueMatt> sipa: mmm, so we do call remove before shutdown...odd that jonas still added DB_LOG_AUTOREMOVE to DB_CONFIG...
 733 2011-11-11 17:08:38 <sipa> gavinandresen: yes, it's not a 100% protection, but it's better than nothing, as long as we don't claim full prevention of leaking
 734 2011-11-11 17:08:45 <gavinandresen> Does DB_LOG_AUTOREMOVE actually work?  I'd love to get rid of the iterate-the-databse/ directory code....
 735 2011-11-11 17:08:54 <BlueMatt> gavinandresen: not any better than what is done now
 736 2011-11-11 17:08:57 <sipa> gavinandresen: didn't you try that yesterday?
 737 2011-11-11 17:09:06 <BlueMatt> gavinandresen: except its automatic, and we dont have to worry about manually calling remove
 738 2011-11-11 17:09:14 <BlueMatt> gavinandresen: but its not a difference
 739 2011-11-11 17:09:25 <BlueMatt> gavinandresen: you also should not be manually removing log files anyway
 740 2011-11-11 17:09:29 <gavinandresen> sipa:  mmmm....   maybe not.  I got sidetracked by the __db file issue
 741 2011-11-11 17:09:33 <BlueMatt> gavinandresen: bdb does keep those around for a reason...
 742 2011-11-11 17:09:47 <BlueMatt> gavinandresen: obviously when encrypting wallet you should, but otherwise not
 743 2011-11-11 17:09:48 <sipa> aren't the log files removed by dbenv.remove ?
 744 2011-11-11 17:09:57 <gavinandresen> nope
 745 2011-11-11 17:09:59 <BlueMatt> sipa: only *unused* ones
 746 2011-11-11 17:10:07 <BlueMatt> sipa: as you explained earlier afaict
 747 2011-11-11 17:10:15 <sipa> BlueMatt: removing the dbenv is different from autoremoval of log files
 748 2011-11-11 17:11:01 <BlueMatt> sipa: oh, well that I dont know...
 749 2011-11-11 17:11:04 <BlueMatt> if only I could read...
 750 2011-11-11 17:11:22 <BlueMatt> anyway, Im not being helpful here, I need to pay more attention...
 751 2011-11-11 17:11:31 <sipa> what class are you in?
 752 2011-11-11 17:11:35 <BlueMatt> gavinandresen: when you manually delete files, you might as well overwrite then first anyway
 753 2011-11-11 17:11:44 <BlueMatt> sipa: compsci recitation
 754 2011-11-11 17:11:53 <BlueMatt> sipa: iirc my average is something like 115%
 755 2011-11-11 17:11:56 <sipa> recitation?
 756 2011-11-11 17:12:11 <BlueMatt> ie go to class, do a problem when you are done you leave
 757 2011-11-11 17:12:13 <BlueMatt> Im done...
 758 2011-11-11 17:12:18 <gavinandresen> BlueMatt: ok, ok, if I manually remove them I'll let you spend 20 seconds overwriting them first.
 759 2011-11-11 17:12:49 <BlueMatt> gavinandresen: after dbenv.close(0) before dbenv.remove, you should overwrite stuff
 760 2011-11-11 17:12:55 <gavinandresen> Can we figure out the other show-stopper issue first?  That when I call TopUpKeyPool bdb decides to write old unencrypted keys to the file?
 761 2011-11-11 17:12:56 <BlueMatt> (Id think)
 762 2011-11-11 17:13:25 <sipa> gavinandresen: sure it's an old one?
 763 2011-11-11 17:13:31 BlueMatt has quit (Quit: Ex-Chat)
 764 2011-11-11 17:13:37 <gavinandresen> sipa: yes
 765 2011-11-11 17:13:44 <gavinandresen> "Found 5d903af9588014d4bd801e976f348a05f585837807861d099bd621dc5139fe38 at position 57646
 766 2011-11-11 17:13:44 <gavinandresen> "
 767 2011-11-11 17:13:44 <sipa> that's a relief
 768 2011-11-11 17:14:22 <gavinandresen> (good thing my test wallet has only 0.001 bitcoins in it)
 769 2011-11-11 17:14:55 BlueMatt-mobile has joined
 770 2011-11-11 17:15:14 <gavinandresen> sipa:  ... my tool doesn't look for new keys being unencrypted-- I should test for that, too.
 771 2011-11-11 17:19:03 <gavinandresen> So, assuming bdb is overwriting old pages or something... yeah, I think it might be 'restart bitcoin' time.  Hrmmm.
 772 2011-11-11 17:19:25 <gavinandresen> That's annoying.
 773 2011-11-11 17:19:53 ThomasV_ has joined
 774 2011-11-11 17:20:17 ThomasV has quit (Remote host closed the connection)
 775 2011-11-11 17:20:21 <luke-jr> gavinandresen: I imagine bdb has a strict mode for valgrind at least
 776 2011-11-11 17:20:25 ThomasV_ has quit (Client Quit)
 777 2011-11-11 17:20:43 ThomasV has joined
 778 2011-11-11 17:21:00 <gavinandresen> luke-jr: it isn't uninitialized memory, I think it just probably has a page cache that it is re-using.
 779 2011-11-11 17:21:13 <luke-jr> gavinandresen: it *would be* uninitialized if it hadn't been formerly used…
 780 2011-11-11 17:22:21 <gavinandresen> No, it might allocate and zero on disk, then mmap(), but when freed it just marks as 'all bytes free', etc.....
 781 2011-11-11 17:22:33 BlueMatt-mobile has quit (Ping timeout: 256 seconds)
 782 2011-11-11 17:23:25 <gavinandresen> ... and this is all a big assumption, maybe there's a bug in the keystore/wallet code that is causing the old unencrypted key to get written....
 783 2011-11-11 17:25:30 <luke-jr> gavinandresen: what if you set the cache size to 0 and back?:
 784 2011-11-11 17:25:56 theorb has joined
 785 2011-11-11 17:26:38 coblee has quit (Quit: coblee)
 786 2011-11-11 17:28:06 <gavinandresen> "Because databases opened within Berkeley DB environments use the cache specified to the environment, it is an error to attempt to set a cache in a database created within an environment."
 787 2011-11-11 17:28:22 <gavinandresen> Grumble grumble...
 788 2011-11-11 17:31:55 da2ce7 has joined
 789 2011-11-11 17:34:33 eueueu has quit (Quit: Page closed)
 790 2011-11-11 17:34:37 da2ce7_ has joined
 791 2011-11-11 17:34:37 da2ce7 has quit (Client Quit)
 792 2011-11-11 17:34:55 da2ce7_ has quit (Read error: Connection reset by peer)
 793 2011-11-11 17:35:00 <sipa> bah
 794 2011-11-11 17:35:03 da2ce7 has joined
 795 2011-11-11 17:35:03 da2ce7 has quit (Changing host)
 796 2011-11-11 17:35:04 da2ce7 has joined
 797 2011-11-11 17:35:12 <sipa> can't we just switch to json or xml files for the wallet?
 798 2011-11-11 17:36:24 <tcatm> I'd like a append-only format so we also have a journal (with a feature to compact the wallet)
 799 2011-11-11 17:37:15 <cjdelisle> json or something like, which people can print out, would give people a warm and fuzzy feeling...
 800 2011-11-11 17:37:56 <sipa> the advantage of bdb is that you don't need to load it in memory entirely
 801 2011-11-11 17:38:00 <sipa> ... but we do that anyway
 802 2011-11-11 17:40:22 BlueMatt has joined
 803 2011-11-11 17:40:34 BlueMatt has quit (Read error: Connection reset by peer)
 804 2011-11-11 17:41:54 <gavinandresen> I'm going to switch gears a little bit-- eat lunch, then write a tool to dump the keypool keys from an unlocked encrypted wallet.  Then I'll run some tests to see if unencrypted versions can end up in the wallet-- if 'born encrypted' doesn't even work, then we've got bigger problems.
 805 2011-11-11 17:42:37 <Eliel> gavinandresen: best scan the whole partition :)
 806 2011-11-11 17:44:50 <helo> will you find partial keys if they are there?
 807 2011-11-11 17:48:55 amiller has joined
 808 2011-11-11 17:49:59 Snapman is now known as Snapman[afkers]
 809 2011-11-11 17:59:58 iocor has joined
 810 2011-11-11 18:00:05 wasabi2 has joined
 811 2011-11-11 18:02:07 <sipa> gavinandresen: use etotheipi's tool?
 812 2011-11-11 18:02:37 <sipa> gavinandresen: i fear the only rock-solid way is to not wait for the wallet handle to be given up, but all handles, and then shutdown the dbenv entirely
 813 2011-11-11 18:02:47 wasabi1 has quit (Ping timeout: 256 seconds)
 814 2011-11-11 18:02:52 <sipa> to prevent old caches from ending up in the file
 815 2011-11-11 18:03:33 Turingi has joined
 816 2011-11-11 18:03:33 Turingi has quit (Changing host)
 817 2011-11-11 18:03:33 Turingi has joined
 818 2011-11-11 18:04:47 BlueMatt has joined
 819 2011-11-11 18:04:49 BlueMatt has quit (Changing host)
 820 2011-11-11 18:04:49 BlueMatt has joined
 821 2011-11-11 18:07:53 <helo> regarding recommending encrypted swap... shouldn't mlock() keep that from being an issue?
 822 2011-11-11 18:08:29 Beremat has joined
 823 2011-11-11 18:08:41 <BlueMatt> gavinandresen: got it to work?
 824 2011-11-11 18:08:47 <gavinandresen> no
 825 2011-11-11 18:08:53 <BlueMatt> helo: in theory, but in reality...
 826 2011-11-11 18:09:01 <helo> i guess suspend to disk could still write keys to disk
 827 2011-11-11 18:09:02 <BlueMatt> helo: and not on windows
 828 2011-11-11 18:09:04 <gmaxwell> helo: 08:22 < sipa> luke-jr: a serious attempt is made for that 08:22 < sipa> but it's impossible to guarantee 08:22 < gmaxwell> mlocking a big gui application is basically a non-starter. 08:23 < gmaxwell> if you do mlockall MCL_FUTURE then the program will crash if any mlocking fails.
 829 2011-11-11 18:09:27 <helo> ahh, missed that
 830 2011-11-11 18:09:42 <gmaxwell> helo: it attempts to mlock the segments known to contain private data but it's not completely unlikely that there are leaks.
 831 2011-11-11 18:09:50 d4de has quit (Ping timeout: 258 seconds)
 832 2011-11-11 18:10:08 <sipa> it's completely unlikely that there are no leaks
 833 2011-11-11 18:10:09 <BlueMatt> gavinandresen: did you push what you have?
 834 2011-11-11 18:10:17 <gavinandresen> yes
 835 2011-11-11 18:10:24 <sipa> gavinandresen: GUI inputs fields are not protected
 836 2011-11-11 18:10:44 <BlueMatt> gavinandresen: and what is the problem again? (sorry, I skipped over those comments earlier because I knew I didnt have time to fix them at that second)
 837 2011-11-11 18:11:04 <gmaxwell> and mlocking the whole thing basically wouldn't work. (esp since common systems limit users to 1mb of mlocked data or something like that) and still probably wouldn't be leak free regardless.
 838 2011-11-11 18:11:11 erle- has quit (Ping timeout: 240 seconds)
 839 2011-11-11 18:11:43 <gavinandresen> BlueMatt: stepping through EncryptWallet in the debugger... here's what happens:
 840 2011-11-11 18:12:08 <BlueMatt> win32 actually specifically states that VirtualLock does NOT guarantee that the memory locked will never be paged, only that it will be less likely to be paged
 841 2011-11-11 18:12:47 <gavinandresen> BlueMatt: CDB::Rewrite() rewrites a clean file-- no old, unencrypted private keys.
 842 2011-11-11 18:12:51 slush has joined
 843 2011-11-11 18:13:00 <helo> good to know...
 844 2011-11-11 18:13:16 * luke-jr blinks
 845 2011-11-11 18:13:32 <luke-jr> gavinandresen: did you just say you found a bdb method to do the whole process of resilvering?
 846 2011-11-11 18:13:33 <gavinandresen> BlueMatt: ... but after calling Unlock() and then TopUpKeyPool(), one old unencrypted private key is written into it (in unused space, I assume)
 847 2011-11-11 18:13:42 <gavinandresen> luke-jr: no
 848 2011-11-11 18:14:20 <gavinandresen> class CDB is a Satoshi creation
 849 2011-11-11 18:14:32 <BlueMatt> gavinandresen: hmmm
 850 2011-11-11 18:14:33 <gavinandresen> (CDB::Rewrite is mostly sipa)
 851 2011-11-11 18:14:42 <luke-jr> i c
 852 2011-11-11 18:15:29 <gavinandresen> I'm switching gears to test and see if we're completely screwed-- if 'born encrypted' keys can somehow end up unencrypted in wallet.dat.
 853 2011-11-11 18:19:26 amiller has quit (Remote host closed the connection)
 854 2011-11-11 18:19:34 <CIA-89> bitcoin: Kamil Domanski * r712c958289d4 gentoo/net-p2p/bitcoin-qt/Manifest: net-p2p/bitcoin-qt: forgot digest
 855 2011-11-11 18:20:41 amiller has joined
 856 2011-11-11 18:21:49 d4de has joined
 857 2011-11-11 18:23:08 <Diablo-D3> gavinandresen: completely screwed? oh noes!
 858 2011-11-11 18:23:11 <da2ce7> the wallet should be flat files, one line to private key; base64 encoded.
 859 2011-11-11 18:23:12 <Diablo-D3> time to sell all of our btc!
 860 2011-11-11 18:23:28 <Diablo-D3> I SAID, TIME TO SELL ALL OF OUR BTC!
 861 2011-11-11 18:23:33 <gavinandresen> da2ce7: get coding on that....
 862 2011-11-11 18:24:17 <gmaxwell> da2ce7: the wallet contains a more than just keys.
 863 2011-11-11 18:24:26 <gavinandresen> really, it isn't helpful to say "just rewrite the whole thing and it will all work perfectly."
 864 2011-11-11 18:24:47 erle- has joined
 865 2011-11-11 18:24:53 <da2ce7> gmaxwell, but the private parts of the keys can be stored in a super simple format.
 866 2011-11-11 18:25:38 <gmaxwell> da2ce7: they can, but as soon as you're mixing data of different types that complicates your 'just use a simple flatfile'... especially when you want it to be able to scale up to large amounts of randomly accessed data.
 867 2011-11-11 18:27:02 <gavinandresen> private keys in a simple binary file, like the blk00001.dat files, is a good idea.
 868 2011-11-11 18:27:28 <gavinandresen> ... and maybe that's the correct thing to do.  But even if we do that, there is still the problem of old keying material ending up in the wallet.dat file.
 869 2011-11-11 18:28:19 <da2ce7> gavinandresen, we should have a directory for the private keys;  so adding merging wallets, having seperate wallets is nothing more than creating annother file in the folder.
 870 2011-11-11 18:28:32 <gmaxwell> gavinandresen: going to >1 file to backup would be a pretty unfortunately step.
 871 2011-11-11 18:28:50 <gavinandresen> da2ce7: so get coding on that....
 872 2011-11-11 18:28:58 <luke-jr> gmaxwell: what needs to be backed up besides private keys?
 873 2011-11-11 18:29:00 <gavinandresen> gmaxwell: agreed.
 874 2011-11-11 18:29:05 <luke-jr> IIRC you can get the pubkey back from the privkey
 875 2011-11-11 18:29:06 * da2ce7 is just talking to my friend about it.
 876 2011-11-11 18:29:17 <gmaxwell> luke-jr: account data, knowing what keypool keys are used, txn comments.
 877 2011-11-11 18:29:18 <luke-jr> da2ce7: any such change doesn't fix the problem at hand
 878 2011-11-11 18:29:19 <da2ce7> maybe we can get tasked with this.
 879 2011-11-11 18:29:36 <da2ce7> luke-jr, we can just re-write the wallet with random data once it is coded.
 880 2011-11-11 18:29:38 <gavinandresen> "get tasked with" ???
 881 2011-11-11 18:29:53 <gmaxwell> da2ce7: I hereby assign you to get to it.
 882 2011-11-11 18:29:53 <da2ce7> gavinandresen, we have other project happening atm also.
 883 2011-11-11 18:29:56 <luke-jr> da2ce7: no, you're talking about a major design change. that wouldn't go into stable versions
 884 2011-11-11 18:30:23 <gavinandresen> yes, you are hereby tasked with coming up with the perfect secure wallet solution.
 885 2011-11-11 18:30:28 <gmaxwell> da2ce7: also please solve the case where you split up a wallet and you have txns in your history where you have some but not all of the keys. Make sure that case does nothing confusing.
 886 2011-11-11 18:31:07 <da2ce7> gmaxwell,  we will need to have some 'send to seperate wallets
 887 2011-11-11 18:31:15 <gavinandresen> (pardon me for being extra special grumpy today.  I was hoping to have a more relaxing eleven-day....)
 888 2011-11-11 18:31:54 <gmaxwell> da2ce7: hm? send to another wallet isn't the same as moving the keys, however.
 889 2011-11-11 18:33:33 <da2ce7> it depeneds if wallets are additive only or not.  I recommend they should be.
 890 2011-11-11 18:34:54 <gmaxwell> da2ce7: the filesystem centric thing you suggested above would expose a lot of corner cases (e.g. less than 100% merges, removals, etc).
 891 2011-11-11 18:35:59 <da2ce7> gmaxwell, well the goal (for me) is that the wallet isnt' stored in (appdata)/bitcoin/wallet.  But (documents)/wallet... so it is an easy to manage file for end-uses.
 892 2011-11-11 18:36:34 <da2ce7> if we have per-key encryption this sort of system makes logcal sence.
 893 2011-11-11 18:38:21 <gmaxwell> If your goal is to change where its stored, you should do that.
 894 2011-11-11 18:38:59 <da2ce7> but also logical and simple wallet management.
 895 2011-11-11 18:39:36 <gmaxwell> da2ce7: editing your wallet from outside the program is not going to get you "logical and simple wallet management"
 896 2011-11-11 18:40:08 <gmaxwell> it's just going to make available corner cases which are impossible to handle in a user-understandable way.
 897 2011-11-11 18:40:42 OneFixt has quit ()
 898 2011-11-11 18:42:13 Kolky has joined
 899 2011-11-11 18:43:09 merde has joined
 900 2011-11-11 18:48:49 OneFixt has joined
 901 2011-11-11 18:48:50 OneFixt has quit (Changing host)
 902 2011-11-11 18:48:50 OneFixt has joined
 903 2011-11-11 18:50:30 slush has quit (Ping timeout: 240 seconds)
 904 2011-11-11 18:59:06 storrgie has joined
 905 2011-11-11 19:06:57 <makomk> gavinandresen: you found private keys in
 906 2011-11-11 19:07:11 <makomk> __d0.00N? Interesting.
 907 2011-11-11 19:08:01 roconnor has quit (Ping timeout: 260 seconds)
 908 2011-11-11 19:15:19 <luke-jr> gavinandresen: http://cve.mitre.org/cve/obtain_id.html
 909 2011-11-11 19:18:42 jfg5878 has joined
 910 2011-11-11 19:21:28 <luke-jr> gavinandresen: see also the email I CC'd you
 911 2011-11-11 19:22:01 <gavinandresen> luke-jr: thanks
 912 2011-11-11 19:24:53 Sedra- has joined
 913 2011-11-11 19:27:35 Sedra has quit (Ping timeout: 245 seconds)
 914 2011-11-11 19:28:02 sgstair has quit (Ping timeout: 258 seconds)
 915 2011-11-11 19:28:11 urstroyer has left ()
 916 2011-11-11 19:30:51 Sedra- has quit (Remote host closed the connection)
 917 2011-11-11 19:31:00 Sedra has joined
 918 2011-11-11 19:35:24 gfinn has joined
 919 2011-11-11 19:42:19 <gavinandresen> good, it looks like a 'born encrypted' wallet does the right thing.  So shutting down bitcoin and restarting after wallet encryption should work.
 920 2011-11-11 19:42:35 storrgie has quit (Read error: Connection reset by peer)
 921 2011-11-11 19:42:43 <gavinandresen> ... I wonder if we could shut down all of the databases and dbenv and restart.....
 922 2011-11-11 19:46:18 zapnap has quit (Remote host closed the connection)
 923 2011-11-11 19:53:30 iocor has quit (Quit: Computer has gone to sleep.)
 924 2011-11-11 19:57:00 Snapman[afkers] is now known as Snapman
 925 2011-11-11 19:58:06 storrgie has joined
 926 2011-11-11 20:00:23 Sedra has quit (Remote host closed the connection)
 927 2011-11-11 20:00:42 Sedra has joined
 928 2011-11-11 20:06:23 sgstair has joined
 929 2011-11-11 20:08:07 Sedra has quit (Remote host closed the connection)
 930 2011-11-11 20:13:08 Sedra has joined
 931 2011-11-11 20:14:02 RazielZ has quit (Quit: Leaving)
 932 2011-11-11 20:14:16 datagutt is now known as superwar-imitato
 933 2011-11-11 20:14:45 superwar-imitato is now known as datagutt
 934 2011-11-11 20:17:42 Sedra has quit (Ping timeout: 240 seconds)
 935 2011-11-11 20:23:31 <gavinandresen> I just tried it, and DB_LOG_AUTO_REMOVE does not automatically remove all the log files (it keeps one)
 936 2011-11-11 20:23:40 jfg5878 has quit (Ping timeout: 265 seconds)
 937 2011-11-11 20:25:26 erle- has quit (Read error: Connection reset by peer)
 938 2011-11-11 20:25:41 erle- has joined
 939 2011-11-11 20:25:43 <gmaxwell> gavinandresen: that sounds generally like an improvement, though it doesn't solve the current issues.
 940 2011-11-11 20:29:39 <gavinandresen> Anybody here know a little bit of qt programming?
 941 2011-11-11 20:30:44 <gavinandresen> ... or more specific question:  anybody want to volunteer to write code to popup a dialog box after the wallet is encrypted that says "bitcoin needs to restart blah blah blah"
 942 2011-11-11 20:36:24 erle- has quit (Ping timeout: 240 seconds)
 943 2011-11-11 20:37:08 p0s has quit (Remote host closed the connection)
 944 2011-11-11 20:46:01 sshc has quit (Ping timeout: 248 seconds)
 945 2011-11-11 20:49:53 iocor has joined
 946 2011-11-11 20:49:53 iocor has quit (Changing host)
 947 2011-11-11 20:49:53 iocor has joined
 948 2011-11-11 21:00:57 <luke-jr> (or for wxBitcoin for that matter, if anyone wants a wx 0.4.1)
 949 2011-11-11 21:06:27 erle- has joined
 950 2011-11-11 21:06:32 Snapman is now known as Snapman[afkers]
 951 2011-11-11 21:06:43 eastender has quit (Ping timeout: 260 seconds)
 952 2011-11-11 21:06:53 erle- has quit (Client Quit)
 953 2011-11-11 21:07:27 erus` has quit (Read error: Connection reset by peer)
 954 2011-11-11 21:07:48 Snapman[afkers] is now known as Snapman
 955 2011-11-11 21:07:59 erus` has joined
 956 2011-11-11 21:13:23 storrgie has quit (Ping timeout: 258 seconds)
 957 2011-11-11 21:16:20 <wumpus> gavinandresen: yeah no problem...
 958 2011-11-11 21:16:25 amiller has quit (Ping timeout: 248 seconds)
 959 2011-11-11 21:16:42 <gavinandresen> wumpus: thanks
 960 2011-11-11 21:19:06 <wumpus> gavinandresen: there is already a popup after wallet encryption that says "Remember that encrypting your wallet cannot fully protect your bitcoins from being stolen by malware infecting your computer."  ... we should probably add it to that message
 961 2011-11-11 21:19:36 <gavinandresen> good idea
 962 2011-11-11 21:20:21 <Disposition> sounds good, we should have like a "things you should know" section from the menu that explains some misconceptions
 963 2011-11-11 21:21:11 <wumpus> Disposition: yes, we probably need more help messages here and there
 964 2011-11-11 21:22:08 storrgie has joined
 965 2011-11-11 21:22:48 <marf_away> a lot more!
 966 2011-11-11 21:23:07 <marf_away> dont forget people are in general stupit
 967 2011-11-11 21:23:09 <marf_away> like me
 968 2011-11-11 21:23:10 <marf_away> ;D
 969 2011-11-11 21:29:02 <wumpus> I've changed the message to "Restart bitcoin now to finish the encryption process. Remember that encrypting your wallet cannot fully protect your bitcoins from being stolen by malware infecting your computer."
 970 2011-11-11 21:30:03 <gmaxwell> marf_away: people aren't generally stupid— they just have lots of other things to worry about than whatever technical detail you personally consider super important and interesting.
 971 2011-11-11 21:32:33 <wumpus> doesn't change the fact that a lot of people are genuinely stupid, but yeah :)
 972 2011-11-11 21:32:48 <Disposition> just misinformed rather, or miss-informed
 973 2011-11-11 21:33:10 <Disposition> conceptually that's wrong since digital wallets are far different than physical ones
 974 2011-11-11 21:33:21 Sedra has joined
 975 2011-11-11 21:33:43 <wumpus> indeed the analogy goes only so far
 976 2011-11-11 21:34:13 <wumpus> making a copy of your physical wallet would be pretty harmless :-)
 977 2011-11-11 21:35:33 graingert has joined
 978 2011-11-11 21:36:18 <wumpus> gavinandresen: is there a branch I should push the message change to?
 979 2011-11-11 21:36:47 <gavinandresen> wumpus: no, should be completely independent of the internal changes
 980 2011-11-11 21:37:07 <wumpus> it is, but it sill makes sense to merge it at the same time...
 981 2011-11-11 21:37:56 <gavinandresen> I've been working in:  https://github.com/gavinandresen/bitcoin-git/tree/encryptionbug
 982 2011-11-11 21:38:22 <wumpus> well I'll push it to my own bitcoin clone for now
 983 2011-11-11 21:39:19 eastender has joined
 984 2011-11-11 21:39:57 <gavinandresen> wumpus: ok, I'll pull it into my branch and then submit a pull request
 985 2011-11-11 21:40:48 <wumpus> gavinandresen: it's there https://github.com/laanwj/bitcoin/commit/dfe5812b8ceea6949404c8e6ba031331d2f7db33
 986 2011-11-11 21:41:08 <gavinandresen> wumpus: thanks
 987 2011-11-11 21:41:34 <CIA-89> bitcoin: Gavin Andresen master * r6be6be2 / (doc/release-process.txt contrib/qt_translations.py): Tool/process for including correct qt translations in mac .App - http://git.io/QX3tPw
 988 2011-11-11 21:41:35 <CIA-89> bitcoin: Gavin Andresen master * rba1fce9 / (11 files): Merge branch 'nullmodels' of https://github.com/laanwj/bitcoin - http://git.io/dJ5qnQ
 989 2011-11-11 21:42:32 Titanium has joined
 990 2011-11-11 21:47:08 Sedra- has joined
 991 2011-11-11 21:47:56 JFK911 has quit (Ping timeout: 260 seconds)
 992 2011-11-11 21:48:03 JFK911 has joined
 993 2011-11-11 21:48:29 <gavinandresen> wumpus: it just tells the user to restart-- doesn't actually call shutdown?
 994 2011-11-11 21:48:41 Snapman is now known as Snapman[afkers]
 995 2011-11-11 21:49:03 <wumpus> gavinandresen: Yes. Is it supposed to?
 996 2011-11-11 21:49:27 <wumpus> gavinandresen: thought you were talking about adding a messagebox with a message, not killing the process :)
 997 2011-11-11 21:49:41 <gavinandresen> Yes, if it doesn't shutdown right after encryption there is a chance of old unencrypted private keys ending up in the wallet.dat file
 998 2011-11-11 21:50:09 <wumpus> hmm ok
 999 2011-11-11 21:50:13 Sedra has quit (Ping timeout: 255 seconds)
1000 2011-11-11 21:50:22 Sedra has joined
1001 2011-11-11 21:51:35 Sedra- has quit (Ping timeout: 260 seconds)
1002 2011-11-11 21:51:55 <wumpus> I'll add an application quit after the message box and change the message again then
1003 2011-11-11 21:53:46 copumpkin is now known as RadicalFeminist
1004 2011-11-11 21:55:33 Sedra- has joined
1005 2011-11-11 21:58:17 BlueMatt has quit (Read error: Operation timed out)
1006 2011-11-11 21:59:51 Sedra has quit (Ping timeout: 276 seconds)
1007 2011-11-11 22:00:14 RadicalFeminist is now known as copumpkin
1008 2011-11-11 22:01:56 nfriedly has joined
1009 2011-11-11 22:02:16 Snapman[afkers] is now known as Snapman
1010 2011-11-11 22:06:05 <wumpus> gavinandresen: updated
1011 2011-11-11 22:06:30 <gavinandresen> thanks
1012 2011-11-11 22:07:13 Sthebig has quit (Read error: Connection reset by peer)
1013 2011-11-11 22:12:10 <gavinandresen> Announcing the bug on the forum:  https://bitcointalk.org/index.php?topic=51604.0
1014 2011-11-11 22:12:15 <b4epoche> Hey all…  Bitcoin talk basically done.  Take a look at http://public.iwork.com/document/?d=Bitcoin.key&a=p451016
1015 2011-11-11 22:12:34 Turing_i has joined
1016 2011-11-11 22:12:51 <b4epoche> I'll be on the road the next couple hours.  If anyone sees any egregious error, please email me at:  emm10@psu.edu
1017 2011-11-11 22:13:03 <JFK911> joepa@psu.edu
1018 2011-11-11 22:13:05 Turing_i has quit (Read error: Connection reset by peer)
1019 2011-11-11 22:13:09 <JFK911> RIP
1020 2011-11-11 22:13:12 <b4epoche> dumb fuck
1021 2011-11-11 22:13:17 <b4epoche> him, not u
1022 2011-11-11 22:13:26 <JFK911> also the kids rioting because he got fired :(
1023 2011-11-11 22:13:35 <b4epoche> yea, them too...
1024 2011-11-11 22:13:43 <b4epoche> idiots
1025 2011-11-11 22:15:15 Turing_i has joined
1026 2011-11-11 22:15:15 Turing_i has quit (Changing host)
1027 2011-11-11 22:15:15 Turing_i has joined
1028 2011-11-11 22:15:32 BlueMatt has joined
1029 2011-11-11 22:16:09 Turingi has quit (Ping timeout: 260 seconds)
1030 2011-11-11 22:17:37 larsivi has quit (Read error: Connection reset by peer)
1031 2011-11-11 22:18:15 <b4epoche> I'm off.  If anyone wants the slides (made with Keynote) send me email.
1032 2011-11-11 22:18:29 spitefulpretzel has quit (Remote host closed the connection)
1033 2011-11-11 22:18:31 gavinandresen has quit (Quit: gavinandresen)
1034 2011-11-11 22:19:59 traviscj_ has joined
1035 2011-11-11 22:21:36 hostilewind has joined
1036 2011-11-11 22:22:39 abragin has quit ()
1037 2011-11-11 22:22:50 larsivi has joined
1038 2011-11-11 22:23:54 traviscj has quit (Ping timeout: 276 seconds)
1039 2011-11-11 22:25:11 datagutt has quit (Quit: Computer has gone to sleep.)
1040 2011-11-11 22:32:11 BlueMatt has quit (Ping timeout: 245 seconds)
1041 2011-11-11 22:36:43 Zarutian has quit (Quit: Zarutian)
1042 2011-11-11 22:42:23 agricocb has quit (Quit: Leaving.)
1043 2011-11-11 22:44:12 Turing_i is now known as Turingi
1044 2011-11-11 22:49:27 BlueMatt has joined
1045 2011-11-11 22:53:49 BlueMatt has quit (Client Quit)
1046 2011-11-11 22:57:39 d4de has quit (Ping timeout: 258 seconds)
1047 2011-11-11 22:58:22 <graingert> b4epoche: mee
1048 2011-11-11 22:58:37 eoss has joined
1049 2011-11-11 22:58:38 eoss has quit (Changing host)
1050 2011-11-11 22:58:38 eoss has joined
1051 2011-11-11 22:59:09 <graingert> b4epoche: you got a pdf of it
1052 2011-11-11 23:02:33 wasabi1 has joined
1053 2011-11-11 23:05:14 wasabi2 has quit (Ping timeout: 248 seconds)
1054 2011-11-11 23:06:56 osmosis has joined
1055 2011-11-11 23:07:30 sshc has joined
1056 2011-11-11 23:08:39 shadders has joined
1057 2011-11-11 23:16:23 agricocb has joined
1058 2011-11-11 23:19:15 marf_away has quit (Quit: Nettalk6 - www.ntalk.de)
1059 2011-11-11 23:20:30 traviscj has joined
1060 2011-11-11 23:20:42 traviscj_ has quit (Ping timeout: 256 seconds)
1061 2011-11-11 23:21:23 <graingert> does anyone have any documentation on what caused the encryption bug?
1062 2011-11-11 23:21:45 <gmaxwell> graingert: It's not clear what you're asking.
1063 2011-11-11 23:22:03 <graingert> where are the private keys leaked
1064 2011-11-11 23:22:20 <graingert> how can one repeat the bug
1065 2011-11-11 23:22:22 <gmaxwell> The issue is simply that although bitcoin deleted the old unencrpyted private keys in your wallet, the database library leaves some of the data laying around.
1066 2011-11-11 23:22:32 <graingert> ah of course
1067 2011-11-11 23:22:44 <graingert> a fix would be to copy everything into a new file perhaps
1068 2011-11-11 23:22:47 <gmaxwell> graingert: create a wallet. encrypt it.  Observe that there are still private keys in it when you scan through the data directly.
1069 2011-11-11 23:23:17 <gmaxwell> graingert: yes, thats part of the solution. There is a compact() call, but its not sufficient.
1070 2011-11-11 23:24:18 <gmaxwell> (key material ends up in the db transaction log files too, and compact also still sometimes leaves some in the wallet file.
1071 2011-11-11 23:24:22 <gmaxwell> )
1072 2011-11-11 23:24:55 <graingert> and if one deletes the file
1073 2011-11-11 23:25:02 <graingert> it will remain on the filesystem
1074 2011-11-11 23:25:10 <graingert> and if you shred it, it still won't work
1075 2011-11-11 23:25:17 <gmaxwell> yes, but that was always known. It was considered out of scope.
1076 2011-11-11 23:25:22 <graingert> the only solution is to send all bitcoin to new keys
1077 2011-11-11 23:25:23 <gmaxwell> (the filesystem part)
1078 2011-11-11 23:25:31 <gmaxwell> (and thats documented)
1079 2011-11-11 23:25:41 <graingert> that have never hit a disk
1080 2011-11-11 23:26:23 <graingert> but that will link the keys
1081 2011-11-11 23:26:42 <gmaxwell> What will link the keys?
1082 2011-11-11 23:26:53 <gmaxwell> oh, sending them to a single address. Sure.
1083 2011-11-11 23:27:02 <graingert> or sending them all in one go
1084 2011-11-11 23:27:08 <graingert> output by output
1085 2011-11-11 23:27:20 <gmaxwell> Though you'll randomly be mixing them up anyways.. it doesn't take long before any addresses you reuse are linked.
1086 2011-11-11 23:27:21 <graingert> each in a separate tx
1087 2011-11-11 23:27:31 <graingert> ?
1088 2011-11-11 23:27:50 <gmaxwell> graingert: its unusual for transactions to only spend a single input.
1089 2011-11-11 23:28:00 <graingert> yes
1090 2011-11-11 23:28:02 <graingert> as I said
1091 2011-11-11 23:28:06 <graingert> it would link the addresses
1092 2011-11-11 23:28:20 <gmaxwell> Your addresses are probably already linked if they're in the same wallet.
1093 2011-11-11 23:28:29 <gmaxwell> Or they will be if you actually use them.
1094 2011-11-11 23:29:12 sacarlson has quit (Ping timeout: 240 seconds)
1095 2011-11-11 23:29:23 <graingert> okay so it would be prudent to move all bitcoin spendable by potentially exposed keys into new unexposed keys
1096 2011-11-11 23:29:40 <graingert> when the wallet becomes encrypted
1097 2011-11-11 23:29:46 groffer has quit (Ping timeout: 248 seconds)
1098 2011-11-11 23:30:00 <graingert> unexposed keys/key
1099 2011-11-11 23:30:06 <gmaxwell> graingert: ::shrugs:: you can't do anything about persistent addresses people use.
1100 2011-11-11 23:30:13 <graingert> and from then
1101 2011-11-11 23:30:18 devrandom has quit (Ping timeout: 248 seconds)
1102 2011-11-11 23:30:30 <graingert> autosend all exposed key bitcoin into new addresses
1103 2011-11-11 23:31:19 <gmaxwell> yea, not going to happen.
1104 2011-11-11 23:31:23 <graingert> people worried about encryption would understand this
1105 2011-11-11 23:32:11 <gmaxwell> If someone has actually stolen those old keys, they'd just race you for funds sent to those old addresses... and they'd probably do better than the built in autosending.
1106 2011-11-11 23:32:29 <graingert> okay that's a bad idea
1107 2011-11-11 23:32:31 <gmaxwell> (or rather they'd just use the same functionality, but they'd run a half dozen nodes and crank up their connection counts so they'd usually beat you)
1108 2011-11-11 23:33:09 <graingert> I think giving the option for sending all bitcoin to guaranteed "green" key(s) after encryption is a good plan
1109 2011-11-11 23:33:22 <gmaxwell> graingert: it's trivially enough to do manually.
1110 2011-11-11 23:33:32 <gmaxwell> getnewaddres.. tada green address... send.
1111 2011-11-11 23:33:32 <graingert> okay in that case just a message suggesting it
1112 2011-11-11 23:33:34 eastender has quit (Ping timeout: 260 seconds)
1113 2011-11-11 23:33:38 <graingert> gmaxwell: no
1114 2011-11-11 23:33:42 <graingert> wrong!
1115 2011-11-11 23:33:49 <gmaxwell> No, not wrong.
1116 2011-11-11 23:33:52 <graingert> you still have the 100 in the queue
1117 2011-11-11 23:33:56 <gmaxwell> Nope.
1118 2011-11-11 23:34:02 <gmaxwell> Fixing that too.
1119 2011-11-11 23:34:03 <graingert> are they dropped on encryption?
1120 2011-11-11 23:34:10 <gmaxwell> marked as used.
1121 2011-11-11 23:34:17 <graingert> fair enough
1122 2011-11-11 23:34:23 <graingert> good solution
1123 2011-11-11 23:34:40 <gmaxwell> There is a readme that someone (gavin) mocked up that had that instruction. dunno where the link went.
1124 2011-11-11 23:36:59 erle- has joined
1125 2011-11-11 23:38:12 <graingert> gmaxwell: what instruction?
1126 2011-11-11 23:38:47 theymos has joined
1127 2011-11-11 23:40:05 <gmaxwell> If you're concerned your unencrypted keys may be compromised hit getnewaddress, send your balance there.
1128 2011-11-11 23:43:46 groffer has joined
1129 2011-11-11 23:45:19 Burgundy has joined
1130 2011-11-11 23:45:37 sacarlson has joined
1131 2011-11-11 23:48:21 kish has quit (Remote host closed the connection)
1132 2011-11-11 23:49:19 Edward_Black has quit (Ping timeout: 240 seconds)
1133 2011-11-11 23:52:11 da2ce7 has quit (Ping timeout: 245 seconds)
1134 2011-11-11 23:53:23 kish has joined
1135 2011-11-11 23:53:36 DaQatz has quit (Ping timeout: 240 seconds)
1136 2011-11-11 23:53:41 TheZimm has joined
1137 2011-11-11 23:54:21 Edward_Black has joined
1138 2011-11-11 23:54:38 DaQatz has joined
1139 2011-11-11 23:57:13 AStove has joined
1140 2011-11-11 23:59:16 iocor has quit (Quit: Computer has gone to sleep.)
1141 2011-11-11 23:59:22 nfriedly has left ()
1142 2011-11-11 23:59:38 devrandom has joined