1 2011-10-23 04:20:19 <CIA-101> bitcoin: phantomcircuit * rddeedffaf2b6 bitcoin-alt/ (4 files in 3 dirs): significant progress finding heads is still very expensive
  2 2011-10-23 04:59:35 <wumpus> tcatm: the translations are built into the executable automatically
  3 2011-10-23 05:00:36 <wumpus> tcatm: and  I'm not sure that would be sane; it would get rid of one place where the translations are specified, but you still need to add them to the qrc file
  4 2011-10-23 05:01:55 <wumpus> tcatm: but maybe it's a good idea, if we add a warning :-)
  5 2011-10-23 05:02:27 <phantomcircuit> ;;bc,blocks
  6 2011-10-23 05:02:28 <gribble> 150314
  7 2011-10-23 05:02:33 <gribble> 150314
  8 2011-10-23 05:02:33 <phantomcircuit> ;;bc,blocks
  9 2011-10-23 05:06:16 <PK> wumpus / tcatm: btc translations? I love to keep them separate so I can delete them. I want bitcoin in english, not in my native language. If you compile it into the binary then please and an option to select it manually.
 10 2011-10-23 05:06:50 <wumpus> PK: then make sure your LANG_XX environment variables are set to english
 11 2011-10-23 05:06:59 <wumpus> I'm in the netherlands and always installa an english ubuntu too
 12 2011-10-23 05:07:14 <PK> wumpus: windows.
 13 2011-10-23 05:07:22 <wumpus> you don't have t resort to hacks like deleting files for that
 14 2011-10-23 05:07:29 <wumpus> just fix the damn thing <:
 15 2011-10-23 05:07:58 <PK> although my windows is english by default. I have the swiss german locales. But some applications (like bitcoin) still don't get it.
 16 2011-10-23 05:08:21 <lfm> the LANG_XX would effect lots more stuff than just bitcoin
 17 2011-10-23 05:08:51 <wumpus> lfm: yes, but I suppose he doesn't want any language in his native language
 18 2011-10-23 05:08:56 <wumpus> lfm: application
 19 2011-10-23 05:08:58 <PK> wumpus: you have the windows language, there is a separate language setting for office and then you have the locales and the keyboard layout. Four places where you can set the language. And that isn't including the browser settings even. :(
 20 2011-10-23 05:09:17 <wumpus> PK: I agree with your problem
 21 2011-10-23 05:09:19 <lfm> wumpus ya, thats not what he said, he said he just wanted bitcoin in english
 22 2011-10-23 05:09:23 <wumpus> PK: just not with your solution :-)
 23 2011-10-23 05:09:54 <wumpus> this means qt takes the wrong language setting on windows, I guess
 24 2011-10-23 05:10:00 <wumpus> we just use the qt system locale now to set the language
 25 2011-10-23 05:10:04 <PK> wumpus: you're wrong there. Some things, like financial stuff I want in my native language.Because I learned all the accounting stuff years ago in german.
 26 2011-10-23 05:10:19 <CIA-101> bitcoin: phantomcircuit * r01e60bf9c2a2 bitcoin-alt/ (bitcoin.py bitcoin/peer.py bitcoin/storage/postgres.py): store transactions as well
 27 2011-10-23 05:10:24 <wumpus> feel free to send a fix to take the right setting, personally I don't use windows and know nothing of the regedit forests
 28 2011-10-23 05:10:26 <PK> but that's fine. I can write a batch script that sets the lang_xx just for that applicatino.
 29 2011-10-23 05:11:36 <lfm> having the language files separate is not a big deal
 30 2011-10-23 05:11:46 <wumpus> PK: that's an even easier solution, indeed
 31 2011-10-23 05:12:06 <PK> wumpus: I already have one to start bitcoind with -server
 32 2011-10-23 05:12:08 <wumpus> yes it is. I would prefere a statically linked executable that includes everything
 33 2011-10-23 05:12:17 <wumpus> it means you can launch it from usb stick etc
 34 2011-10-23 05:12:42 <wumpus> otherwise you have to hardcode paths to files and such
 35 2011-10-23 05:12:56 <wumpus> or add additional configuration files to specify the paths
 36 2011-10-23 05:13:01 <lfm> you can launch it from usb with separate files too
 37 2011-10-23 05:13:33 <wumpus> and while developing you have to make sure the right files are used (ie, not the system files but those in your local tree)
 38 2011-10-23 05:13:56 <lfm> linking ALL the text is just bloat that most people wont use
 39 2011-10-23 05:13:56 <PK> actually, I launch it from a network share right now. It works quite fine. Even the datadir works with UNC paths.
 40 2011-10-23 05:14:01 <wumpus> it's just text man
 41 2011-10-23 05:14:13 <wumpus> it's not bloat it hardly figures in the executable size
 42 2011-10-23 05:14:39 <wumpus> and it isn't loaded into memory if you don't use it either
 43 2011-10-23 05:15:09 <lfm> 836K    .
 44 2011-10-23 05:15:22 <wumpus> anyway, feel free to submit a patch to do it differently
 45 2011-10-23 05:15:44 <wumpus> I'm a bit tired of this
 46 2011-10-23 05:15:46 <lfm> simple just yank the patch that links them in
 47 2011-10-23 05:16:00 <wumpus> yes and make an installer that installs them
 48 2011-10-23 05:16:09 <wumpus> otherwise no one has translations anymore
 49 2011-10-23 05:16:15 <lfm> wasnt that done before too?
 50 2011-10-23 05:16:20 <wumpus> I have no idea
 51 2011-10-23 05:16:38 <wumpus> I don't worry about it I have other things to do, it works fine to me to link everything into one
 52 2011-10-23 05:16:40 <lfm> so you fixed something that you dont know if it needed fixing?
 53 2011-10-23 05:18:07 <wumpus> like getting cross-compile to windows to work
 54 2011-10-23 05:18:11 <PK> wumpus: don't you love it if you have a great idea, implement it and people hold it again you? Happens to me all the time.
 55 2011-10-23 05:18:30 <lfm> yup, thats very common
 56 2011-10-23 05:18:53 <wumpus> PK: yeah, it sometimes makes me wonder why I bother with open source development at all
 57 2011-10-23 05:19:24 <lfm> on the other hand "it works this way better for me" is not always the best way to do things
 58 2011-10-23 05:19:26 <wumpus> PK: people complain a lot less about implementation-specific details for paid projects
 59 2011-10-23 05:19:36 <wumpus> PK: they just want things fixed/working
 60 2011-10-23 05:20:18 <wumpus> well a single exe works and is easy
 61 2011-10-23 05:20:29 <PK> wumpus: actually no. It happens in paid projects too. You fix some really annoying bug and everything is like "Why did you take that feature away? We use that every day in production!"
 62 2011-10-23 05:20:42 <wumpus> mac and android have solved the problem by always bundling files for one application into one
 63 2011-10-23 05:21:17 <lfm> resource forks arnt exactly all one file
 64 2011-10-23 05:21:18 <wumpus> PK: yeah that's true
 65 2011-10-23 05:21:22 <PK> wumpus: teached me to always ask the users first before. More than once it turned out that I was just thinking differently.
 66 2011-10-23 05:21:50 <wumpus> I mean deleting file sis *not* the proper way to change translations for a program
 67 2011-10-23 05:22:02 <wumpus> install the next version and whoo, they're back
 68 2011-10-23 05:22:15 <FellowTraveler> hi all.
 69 2011-10-23 05:22:47 <wumpus> the best solution would be to add a configuration option for the language
 70 2011-10-23 05:22:54 <wumpus> in options...
 71 2011-10-23 05:22:57 <PK> wumpus: you could also rename the directories to mix the translations. :) Great way to play pranks on your friends if they forgot to screenlock their computer again.
 72 2011-10-23 05:23:31 <wumpus> PK: right, you'd need a hex editor for that now :P
 73 2011-10-23 05:23:43 <wumpus> PK: that's one hell of a usecase though :-)
 74 2011-10-23 05:23:48 <lfm> just an environment var
 75 2011-10-23 05:23:56 <wumpus> the environment var already exists
 76 2011-10-23 05:24:09 <lfm> and you can change it for pranks
 77 2011-10-23 05:24:10 <wumpus> export LANG=nl_NL.utf8
 78 2011-10-23 05:24:13 <wumpus> instant dutch
 79 2011-10-23 05:25:29 <lfm> I always run bitcoin from scripts anyway so I could put the LANG= in the scripts if I had such a problem.
 80 2011-10-23 05:25:38 <wumpus> most people really couldn't give a fuck that all translations are in there, it could even be useful if they want to learn a language
 81 2011-10-23 05:25:51 <lfm> lol
 82 2011-10-23 05:26:02 <lfm> worst way to learn a language ever
 83 2011-10-23 05:26:19 <wumpus> yeah probably
 84 2011-10-23 05:26:30 <wumpus> though it helps to be confronted with a language everywhere
 85 2011-10-23 05:26:33 <gmaxwell> "so, all you know how to say is 'transaction fee' and 'address'??"
 86 2011-10-23 05:26:35 <wumpus> forces you to look up works
 87 2011-10-23 05:26:38 <wumpus> words
 88 2011-10-23 05:26:46 <gmaxwell> "what is this confirmation you keep talking about?!"
 89 2011-10-23 05:27:15 <wumpus> hehe
 90 2011-10-23 05:27:17 <lfm> confirmation is a sacrement in the roman church
 91 2011-10-23 05:27:45 <lfm> dictionairies are usefull!
 92 2011-10-23 05:29:13 <wumpus> so using bitcoin is like performaing a catholic ritual? wow, it seems we found an application
 93 2011-10-23 05:29:56 <lfm> so when you make a transaction it will be accepted when you are ten years old
 94 2011-10-23 05:30:38 <wumpus> so that is why we have bible texts in the block chain
 95 2011-10-23 05:30:59 <lfm> yup, its all a papist plot
 96 2011-10-23 05:32:45 <wumpus> do we have a vatican translation already?
 97 2011-10-23 05:32:55 <lfm> latin?
 98 2011-10-23 05:33:03 <chmod755> wumpus lol
 99 2011-10-23 05:33:30 <chmod755> wumpus: there's a channel for vatican city #bitcoin-va
100 2011-10-23 05:37:34 <PK> chmod755: how many bitcoins does an absolution cost?
101 2011-10-23 05:37:39 <PK> I could use one...
102 2011-10-23 05:38:00 <lfm> pay what you can afford to give of course
103 2011-10-23 05:38:35 <PK> So if I gave up all my bitcoins I get an absolution! great, 0.2 BTC sounds really cheap.
104 2011-10-23 05:38:38 <lfm> and if you short change the preist it wont work right
105 2011-10-23 05:39:18 <lfm> you still go to hell for 0.2 btc I think
106 2011-10-23 05:40:23 <PK> oh da.. I mean: oh, he.. or I better don't respond to that.
107 2011-10-23 05:42:31 <wumpus> has anyone here worked with gitian? any idea what kind of AWS image to create?
108 2011-10-23 05:43:26 <chmod755> PK: I'm not sure
109 2011-10-23 05:43:28 <PK> I haven't, sorry.
110 2011-10-23 05:44:22 <chmod755> PK: (09:43:38 AM) pope: absolution costs 250 BTC
111 2011-10-23 05:45:56 <PK> chmod755: lol. 750$ sounds very affordable too :)
112 2011-10-23 06:02:51 <wumpus> tcatm: TRANSLATIONS = $$files(src/qt/locale/bitcoin_*.ts)  seems to work btw
113 2011-10-23 06:25:09 <wumpus> locale nb is nb_NO right?
114 2011-10-23 06:25:50 <wumpus> Bokm??l, Norwegian
115 2011-10-23 06:39:07 <CIA-101> bitcoin: Wladimir J. van der Laan  * rf86ecd4 / (12 files in 3 dirs):
116 2011-10-23 06:50:56 <tcatm> wumpus: I'm not sure whether renaming language files to lang_terrritory is a good idea. e.g. a german translation will probably work well in most territories where people speak german
117 2011-10-23 06:51:06 <wumpus> it is a good idea
118 2011-10-23 06:51:09 <wumpus> all software does it
119 2011-10-23 06:51:30 <wumpus> you can use the qrc file to link de_<otherterritory> to the same language file if you want
120 2011-10-23 06:52:37 <tcatm> and if you miss one it will fall back to english instead of falling back to bitcoin_de.ts?
121 2011-10-23 06:52:46 <wumpus> yes but that was already the case
122 2011-10-23 06:52:56 <wumpus> only the names changed, not the lookup logic
123 2011-10-23 06:53:40 <wumpus> if you want it to pick *some territory* for a language if the language is present but not the territoriy, that would have to be explicitly implemented, but it's probably language dependent whehter that is appropriate
124 2011-10-23 06:54:24 <wumpus> wouldn't be too difficult though
125 2011-10-23 06:55:06 <wumpus> just default to  xx_XX if xx_YY is not present
126 2011-10-23 06:55:26 <tcatm> why not use the built-in fallback logic?
127 2011-10-23 06:55:53 <wumpus> if there is built-in fallback logic that's already used
128 2011-10-23 06:56:07 <wumpus> I'm not sure Qt has that though...
129 2011-10-23 06:56:24 <wumpus> but if you know more than me, feel free to fix it
130 2011-10-23 06:56:28 <wumpus> :P
131 2011-10-23 06:57:18 <wumpus> tcatm: does Qt have built-in fallback logic?
132 2011-10-23 06:57:32 <tcatm> oh, sorry. I didn't know that there was yet another file for finding the translations (bitcoin.qrc). I thought it just used the file name
133 2011-10-23 06:57:35 <wumpus> I have read the documentation but not found anything
134 2011-10-23 06:57:42 <wumpus> but that's some time ago
135 2011-10-23 06:58:36 <wumpus> I'm not much of a geography expert, I didn't even know about territory-specific language differences when I started, that's why they were called bitcoin_nl_NL not bitcoin_nl_NL
136 2011-10-23 06:58:55 <wumpus> * were called bitcoin_nl not bitcoin_nl_NL
137 2011-10-23 06:59:34 <wumpus> but this brings it more in line with the 'usual'
138 2011-10-23 07:00:06 <tcatm> I wonder if I can force transifex to use xx_XX templates for files instead
139 2011-10-23 07:00:22 <tcatm> or can we generate bitcoin.qrc from a template?
140 2011-10-23 07:00:45 <wumpus> (though I had obviously seen en_EN and en_US before :p)
141 2011-10-23 07:01:23 <wumpus> yes that'd be better
142 2011-10-23 07:01:32 <wumpus> to generate the translations part of it with a script
143 2011-10-23 07:02:34 <wumpus> however it shouldn't generate the entire file from a template otherwise you cannot edit it anymore with qt creator
144 2011-10-23 07:04:04 <wumpus> but if you no longer want the translations in the resource file it's even simpler :p
145 2011-10-23 07:04:44 <tcatm> how?
146 2011-10-23 07:04:46 <wumpus> in that case you do need to change the lookup logic though, currently it looks in the resource file only
147 2011-10-23 07:05:04 <wumpus> well just cut it out there
148 2011-10-23 07:05:06 <tcatm> I'd like to have a simple system where I can easily merge new translations from transifex
149 2011-10-23 07:05:37 <wumpus> but that only moves the problem of maing sure the translations end up somewhere where bitcoin can find them
150 2011-10-23 07:06:23 <wumpus> well what about a python script that requests the contents of src/locale/qt/*.ts and updates the qrc file?
151 2011-10-23 07:08:05 <wumpus> or maybe even downloads the ts files from transifex and places them in src/locale/qt
152 2011-10-23 07:08:34 <tcatm> there's already a script for that: tx pull -a :)
153 2011-10-23 07:08:56 <tcatm> not all of them have the territory suffix, though
154 2011-10-23 07:09:37 <wumpus> afaik no territiory suffix is the same as xx_XX
155 2011-10-23 07:10:01 <wumpus> ah not really
156 2011-10-23 07:10:19 <wumpus> it was not for nb_NO... :/
157 2011-10-23 07:10:32 <wumpus> you really need a territory suffix
158 2011-10-23 07:12:41 <wumpus> we need to chang the lookup logic
159 2011-10-23 07:13:15 <tcatm> how do other big projects do it?
160 2011-10-23 07:13:23 <wumpus> https://www.virtualbox.org/wiki/Translating
161 2011-10-23 07:13:33 <wumpus> _TERRITORY is optional
162 2011-10-23 07:13:37 <wumpus> fuck I didn't know that
163 2011-10-23 07:14:50 <wumpus> so let's change the logic to: first look for _lang_TERRITORY, then for _lang, then fall back to default
164 2011-10-23 07:15:19 <tcatm> sounds good
165 2011-10-23 07:15:46 <wumpus> then again that's a code change so I'll do it some time after the 0.5.0 freeze
166 2011-10-23 07:16:29 <tcatm> so can you revert the renaming of language files for now so it stays in sync with transifex?
167 2011-10-23 07:17:54 <wumpus> i guess so... :/
168 2011-10-23 07:18:25 <wumpus> does github provide a way to remove an entire commit?
169 2011-10-23 07:19:00 <tcatm> nope, you can only add new commits that restore old behaviour
170 2011-10-23 07:19:09 <wumpus> fuck
171 2011-10-23 07:19:56 <tcatm> (well, I could force HEAD to point to the old commit but that will screw up for everyone who already pulled the changes)
172 2011-10-23 07:19:56 <wumpus> an automatic way to generate a reverse-commit would be great, like the wikipedia revert
173 2011-10-23 07:20:06 <tcatm> git revert
174 2011-10-23 07:20:07 <wumpus> I don't feel like messing it up even more
175 2011-10-23 07:20:20 <tcatm> or git revert $HASH_OF_COMMIT
176 2011-10-23 07:20:58 <CIA-101> bitcoin: Wladimir J. van der Laan  * r3a30f34 / (12 files in 3 dirs):
177 2011-10-23 07:20:59 <CIA-101> bitcoin: Revert "translation fixes"
178 2011-10-23 07:21:00 <wumpus> done...
179 2011-10-23 07:21:11 <wumpus> someone else can do the translation stuff, I won't touch it anymore with a ten foot pole :P
180 2011-10-23 07:22:20 <tcatm> yes, it's tricky. I already regret setting up the transifex account even though it works really well so far :)
181 2011-10-23 07:22:44 <wumpus> myself I've always used computers in english
182 2011-10-23 07:22:51 <wumpus> so it's all new to me
183 2011-10-23 07:23:07 <tcatm> same here
184 2011-10-23 07:23:29 <wumpus> most people here in the netherlands do use the dutch translations but I've always found translated programs.. annoying in  a way
185 2011-10-23 07:23:33 <tcatm> and I'm a really annoyed that I can't just do LC_MESSAGES=nb_NO ./bitcoin-qt to test translations :/
186 2011-10-23 07:23:48 <wumpus> why not?
187 2011-10-23 07:23:59 <wumpus> qt ignores lc_messages?
188 2011-10-23 07:24:08 <tcatm> no, it exists with an errormessage
189 2011-10-23 07:24:16 <tcatm> exits*
190 2011-10-23 07:24:19 <wumpus> shouldn't you add .utf8?
191 2011-10-23 07:24:26 <wumpus> or is that only for LANG
192 2011-10-23 07:24:55 <tcatm> doesn't work either
193 2011-10-23 07:24:56 <wumpus> some time  ago I had to dive in the Qt source code to find out how it really found out what the system locale was, but it's heavily platform specific
194 2011-10-23 07:25:15 <genjix> hey tcatm, can we put http://conference.bitgroups.org on bitcoin.org?
195 2011-10-23 07:25:16 <tcatm> I think it is missing some system translations or something
196 2011-10-23 07:25:30 <wumpus> export LANG=nl_NL.utf8  works for me at least
197 2011-10-23 07:25:31 <genjix> can use this logo: https://bitcointalk.org/index.php?action=dlattach;attach=5148;type=avatar
198 2011-10-23 07:25:58 <tcatm> terminate called after throwing an instance of 'std::runtime_error' what():  locale::facet::_S_create_c_locale name not valid
199 2011-10-23 07:26:05 <wumpus> it complains about a locale not being supported by the C library, but that's probably some package I'm missing
200 2011-10-23 07:26:07 <tcatm> with nl_NL.utf8
201 2011-10-23 07:26:25 <wumpus> huh that's a boost error...
202 2011-10-23 07:26:28 <wumpus> not qt
203 2011-10-23 07:28:09 <tcatm> can you prepare a pull request to change the lookup logic? I don't know much about qt...
204 2011-10-23 07:28:13 <wumpus> why does it work here?
205 2011-10-23 07:28:24 <wumpus> this doesn't have anything to do with the lookup logic
206 2011-10-23 07:28:42 <tcatm> probably because your system has nl_NL locales installed
207 2011-10-23 07:28:56 <tcatm> I was referring to: 11:14 < wumpus> so let's change the logic to: first look for _lang_TERRITORY, then for _lang, then fall back to default
208 2011-10-23 07:28:57 <wumpus> yes, but changing the look up logic for *Qt* doesn't change that
209 2011-10-23 07:29:20 <wumpus> well I can do it for any language
210 2011-10-23 07:29:54 <wumpus> I'm sure I don't have the Russian locale installed and can still run bitcoin in that language
211 2011-10-23 07:30:30 <wumpus> I do get this warning: (process:31676): Gtk-WARNING **: Locale not supported by C library.
212 2011-10-23 07:30:31 <wumpus> 
213 2011-10-23 07:30:33 <wumpus> but not an error
214 2011-10-23 07:32:03 <tcatm> seems to be a bug in boost: https://svn.boost.org/trac/boost/ticket/4688
215 2011-10-23 07:35:01 <wumpus> at least they fixed it in newer versions, I wonder if there is a way to tell Qt to use a locale that passes by boost
216 2011-10-23 07:43:00 <chmod755> ;;pgp auth
217 2011-10-23 07:43:00 <gribble> (gpg auth <nick>) -- Initiate authentication for user <nick>. You must have registered a GPG key with the bot for this to work. You will be given a random passphrase to clearsign with your key, and submit to the bot with the 'verify' command. Your passphrase will expire within 10 minutes.
218 2011-10-23 07:43:05 <chmod755> ;;pgp auth chmod755
219 2011-10-23 07:43:06 <gribble> Request successful for user chmod755, hostmask chmod755!~affilitat@unaffiliated/chmod755. Your challenge string is: freenode:#bitcoin-otc:00988257a69b0c62487b7c612e53e2e7f065f075f52309929533a598
220 2011-10-23 07:51:48 <CIA-101> libbitcoin: genjix * rba936d41c8d9 /src/storage/ (postgresql_blockchain.cpp postgresql_storage.cpp): bits extraction/compaction util functions - code reuse!
221 2011-10-23 08:01:04 <CIA-101> bitcoin: Con Kolivas * r8c63061eaa39 cgminer/ (adl.c main.c miner.h): Allow a fixed speed difference between memory and GPU clock speed that will change memory speed when GPU speed is changed in autotune mode.
222 2011-10-23 08:21:24 <CIA-101> libbitcoin: genjix * r8e5128ee2d0d / (7 files in 3 dirs): src/storage/postgresql/ sub-directory
223 2011-10-23 09:20:25 <SomeoneWeird> I wanna start a project todo with btc (ofc), but don't know what todo, suggestions/ideas anyone?
224 2011-10-23 09:23:08 <wumpus> I'd recommend choosing a project to make it easier for (web) merchants to accept btc
225 2011-10-23 09:25:01 <SomeoneWeird> mmm
226 2011-10-23 09:25:53 <genjix> SomeoneWeird: you want to help proofread the BIPs
227 2011-10-23 09:26:03 <SomeoneWeird> BIPs?
228 2011-10-23 09:26:53 <genjix> Bitcoin Improvement Proposals
229 2011-10-23 09:26:55 <tcatm> or translate bitcoin, improve the webdesign
230 2011-10-23 09:27:23 <SomeoneWeird> can't translate, what webdesigns?
231 2011-10-23 09:27:27 <genjix> tcatm: did you see my message above about the conference?
232 2011-10-23 09:27:31 <SomeoneWeird> and I suppose genjix, links?
233 2011-10-23 09:27:40 <genjix> https://en.bitcoin.it/wiki/Bitcoin_Improvement_Proposals
234 2011-10-23 09:27:40 <tcatm> genjix: probably not
235 2011-10-23 09:28:15 <genjix> tcatm: can we put the conference http://conference.bitgroups.org on bitcoin.org?
236 2011-10-23 09:28:24 <genjix> we can use this logo: https://bitcointalk.org/index.php?action=dlattach;attach=5148;type=avatar
237 2011-10-23 09:28:51 <genjix> a lot of work has gone into organising that. now we need to swell the attendees
238 2011-10-23 09:31:31 <tcatm> can you get ACKs from other people involved with bitcoin.org (gavin, jgarzik, ...) and provide a short text (about 50..100 words) that ends with something like "Visit <http://...> for more information"?
239 2011-10-23 09:31:55 <genjix> ok
240 2011-10-23 09:32:11 <genjix> the best place is the mailing list, right?
241 2011-10-23 09:32:38 <tcatm> I think just CC'ing everyone involved will be enough
242 2011-10-23 09:32:41 <genjix> ok
243 2011-10-23 09:32:45 <genjix> thanks
244 2011-10-23 09:33:03 <genjix> also, let me know if you need a letter or something to get a day off if you want to attend
245 2011-10-23 09:34:24 <tcatm> well, the problem is I'm already involved in organizing a local event on the same weekend
246 2011-10-23 09:35:14 <genjix> k
247 2011-10-23 09:36:42 <terrytibbs> zhoutong: Your connection seems pretty... spotty! Have you considered a bouncer?
248 2011-10-23 09:37:08 <genjix> zhoutong: or use irssi inside screen on a vps :p
249 2011-10-23 09:38:29 <terrytibbs> terrytibbs: or use irssi inside tmux on a vps!
250 2011-10-23 09:40:55 <wumpus> lol
251 2011-10-23 09:41:54 <tcatm> irssi inside tmux inside screen inside nx :)
252 2011-10-23 09:42:48 <wumpus> tcatm: https://github.com/laanwj/bitcoin/tree/translation_lookup_logic
253 2011-10-23 09:43:10 <wumpus> that's my final word on translations, if I still get it wrong then I give up :)
254 2011-10-23 09:44:22 <nathan7> tmux always gave me problems
255 2011-10-23 09:44:37 <SomeoneWeird> INCEPTIOOOON
256 2011-10-23 09:44:43 <tcatm> wumpus: thanks! :)
257 2011-10-23 09:44:44 <SomeoneWeird> SHELLCEPTION
258 2011-10-23 09:44:50 <AAA_awright> MONEY WITHIN MONEY
259 2011-10-23 09:44:55 <AAA_awright> within money
260 2011-10-23 09:45:03 <AAA_awright> within... money
261 2011-10-23 09:45:16 <AAA_awright> and it all vanishes in the blink of an eye
262 2011-10-23 09:45:28 <AAA_awright> ;_;
263 2011-10-23 09:45:31 <SomeoneWeird> lol
264 2011-10-23 09:45:34 <SomeoneWeird> awww
265 2011-10-23 09:45:35 <SomeoneWeird> right
266 2011-10-23 09:55:47 <CIA-101> libbitcoin: genjix * rdfe6f7317959 / (bitcoin.sql src/storage/postgresql/postgresql_blockchain.cpp): No more block_status.
267 2011-10-23 09:55:48 <CIA-101> libbitcoin: genjix * r6b799ef695bb / (bitcoin.sql src/storage/postgresql/postgresql_blockchain.cpp): Avoid reloading data in validate loop
268 2011-10-23 10:23:19 <talpan> hello, i'm having a hard time compiling bitcoin qt under ubuntu 11.10 after a fresh install. the error i get is: /usr/include/db_cxx.h:59:22: error iostream.h not found
269 2011-10-23 10:25:14 <sipa> talpan: you have all dependencies installed?
270 2011-10-23 10:25:57 <talpan> as far as i know, i already tried various things that i had googled.
271 2011-10-23 10:26:00 <sipa> in particular libdb4.8++-dev ?
272 2011-10-23 10:26:44 <talpan> do i need 4.8 or isn't 4.7 enough?
273 2011-10-23 10:26:56 <sipa> 4.7 is enough
274 2011-10-23 10:27:14 <talpan> additionally i can only find libdb4.7-dev not libdb4.7-dev, is there a difference?
275 2011-10-23 10:27:23 <talpan> *libdb4.7++
276 2011-10-23 10:27:35 <sipa> yes, without ++ is for C
277 2011-10-23 10:27:39 <sipa> with ++ is for C++
278 2011-10-23 10:28:53 <talpan> mhm okay, i download the source and try to compile it, since it isn't in the 11.10 repos
279 2011-10-23 10:32:54 <wumpus> iostream.h, that *should* be part of c++
280 2011-10-23 10:34:28 <wumpus> I somehow don't think this is a dbcxx version issue
281 2011-10-23 10:37:56 <talpan> i think this could be an issue with ubuntu 11.10 since i can't install libdb++
282 2011-10-23 10:38:01 <talpan> weird
283 2011-10-23 10:39:33 <SomeoneWeird> sure it isn't in the repos talpan?
284 2011-10-23 10:39:48 <talpan> http://packages.ubuntu.com/oneiric/allpackages yes
285 2011-10-23 10:39:59 <talpan> double checked it with the package list
286 2011-10-23 10:40:38 <talpan> but it is in the ubuntu natty (.04) repos
287 2011-10-23 10:41:27 <SomeoneWeird> apt-cache search libdb4.8 | grep dev ?
288 2011-10-23 10:41:32 <SomeoneWeird> :P
289 2011-10-23 10:41:41 <SomeoneWeird> https://launchpad.net/ubuntu/oneiric/amd64/libdb4.8
290 2011-10-23 10:41:48 <SomeoneWeird> just the dev package thats not?
291 2011-10-23 10:41:50 <wumpus> talpan: the error message refers to /usr/include/db_cxx.h being installed, so somehow it is installed
292 2011-10-23 10:41:53 <talpan> i need libdb4.x++
293 2011-10-23 10:42:07 <talpan> weird
294 2011-10-23 10:42:30 <talpan> currently i have libdb4.7-dev installed
295 2011-10-23 10:42:51 <talpan> ill test it with 4.8
296 2011-10-23 10:42:54 <genjix> tcatm: kthx. i sent the email.
297 2011-10-23 10:43:01 <genjix> you sould have one too
298 2011-10-23 10:43:16 <genjix> talpan: install libdb++-dev
299 2011-10-23 10:43:36 <talpan> thx :)
300 2011-10-23 10:43:40 <genjix> nw
301 2011-10-23 10:44:19 <genjix> talpan: also google "genjix ubuntu bitcoin" for the other needed packages
302 2011-10-23 10:44:19 <talpan> it is compiling :)
303 2011-10-23 10:45:43 <talpan> that was the only problem, i've compiled bitcoin many times before, but the upgrade to .10 did screw it up :)
304 2011-10-23 10:45:50 <genjix> same here
305 2011-10-23 10:46:06 <genjix> actually i was using oneiric for a while
306 2011-10-23 10:46:13 <genjix> g++ 4.6 is super cool
307 2011-10-23 11:00:17 <wumpus> why does github insist on nuking < and > characters in pull request descriptions
308 2011-10-23 11:01:10 <sipa> i suppose because it uses those for its own markup language
309 2011-10-23 11:01:24 <wumpus> bleh
310 2011-10-23 12:25:17 <nanotube> genjix: re conference: thanks for invite, but eur is a long way away for me :P
311 2011-10-23 12:34:49 <genjix> nanotube: ok :/
312 2011-10-23 18:44:10 <Diablo-D3> deepbit, slush, btcguild, ars, abc, btc.lc, bitclockers, nmcbit, btcmine, emc
313 2011-10-23 19:50:59 <Diablo-D3> current top ten: current top ten: deepbit, slush, ars, btcguild, abc, btc.lc, bitclockers, nmcbit, btcmine, emc
314 2011-10-23 19:52:28 <imsaguy> is there a ddos still going on?
315 2011-10-23 19:52:33 <Diablo-D3> no
316 2011-10-23 19:53:10 <imsaguy> oh
317 2011-10-23 20:00:36 <Diablo-D3> warning: deepbit now has half the network hashrate.
318 2011-10-23 20:01:05 <gmaxwell> :-/
319 2011-10-23 20:01:14 <gmaxwell> fuck, we're back to this crap again.
320 2011-10-23 20:01:42 <gmaxwell> And no one bothered writing automated shenanigans detection yet?
321 2011-10-23 20:01:50 <freewil> i guess bitcoin never the solved the problem of masses being idiots
322 2011-10-23 20:02:01 <gmaxwell> In other news, eligius is back up.. though only as of a few minutes, so it doesn't have real hashrate yet.
323 2011-10-23 20:02:37 <Diablo-D3> gmaxwell: well I left a reply on the deepbit thread
324 2011-10-23 20:02:40 <Diablo-D3> so a few people should jump
325 2011-10-23 20:02:53 <makomk> I suspect a lot of BTCGuild users may be migrating over to Deepbit...
326 2011-10-23 20:03:02 <Diablo-D3> makomk: they should be migrating to ars
327 2011-10-23 20:03:09 <Diablo-D3> ars is the largest no fee pool
328 2011-10-23 20:03:13 <gmaxwell> It was bad timing for eligius to be down.
329 2011-10-23 20:03:30 <Diablo-D3> ars is now the third largest pool too
330 2011-10-23 20:05:01 <gmaxwell> If people are looking for zero fee pools, as mentioned eligius is up again. Also eclipsemc.com has a healthy enough rate, zero fees, score based and has merged mining (like eligius)
331 2011-10-23 20:05:20 <Diablo-D3> abcpool is also zero fee as is bitcoins.lc
332 2011-10-23 20:06:04 <gmaxwell> perhaps the stickies should exclude #1 or #1-#N ? :)
333 2011-10-23 20:06:14 <Diablo-D3> lol
334 2011-10-23 20:06:18 <dikidera> diablo tell me
335 2011-10-23 20:06:23 <Diablo-D3> dikidera: what
336 2011-10-23 20:06:40 <dikidera> why does the 4way algorithm of cgminer uhm..check only 32 nonces on each loop?
337 2011-10-23 20:06:52 <dikidera> and the same ones at that
338 2011-10-23 20:07:04 <Diablo-D3> dunno, ask con
339 2011-10-23 20:07:18 <dikidera> I mean as far as i can read the code...it looks broken to me
340 2011-10-23 20:10:51 <gmaxwell> wow btcguild has only lost ~100 GH since they went uber-fees.
341 2011-10-23 20:11:09 <dikidera> no
342 2011-10-23 20:11:14 <dikidera> btcguild lost 1th
343 2011-10-23 20:11:20 <dikidera> i still remember when they were over 2
344 2011-10-23 20:11:37 <Diablo-D3> they were never fist
345 2011-10-23 20:11:38 <Diablo-D3> *first
346 2011-10-23 20:11:48 <Diablo-D3> deepbit has had first since it got it
347 2011-10-23 20:12:02 <gmaxwell> They were at 1TH last week before the switch, the whole network has lost rate.
348 2011-10-23 20:12:04 <Diablo-D3> it was deepbit, slush, btcguild for awhile
349 2011-10-23 20:12:14 <Diablo-D3> then slush dropped to third
350 2011-10-23 20:12:21 <Diablo-D3> and then btcguild dropped to fourth
351 2011-10-23 20:12:33 <Diablo-D3> so its deepbit, slush, ars, btcguild
352 2011-10-23 20:12:45 <Diablo-D3> and if ars keeps climbing like this, slush'll lose #2 again
353 2011-10-23 20:16:45 <Eliel> well, now that eligius is back, ars is going to lose some of their hashrate.
354 2011-10-23 20:16:56 <Diablo-D3> heh, btcguild is now down to 890
355 2011-10-23 20:18:00 <Diablo-D3> ;;bc,mtgox
356 2011-10-23 20:18:01 <gribble> {"ticker":{"high":3.26,"low":2.95,"avg":3.093004774,"vwap":3.085294059,"vol":39713,"last_all":3.1225,"last_local":3.1225,"last":3.1225,"buy":3.13055,"sell":3.18511}}
357 2011-10-23 20:31:33 <casascius> i'm wondering, are there a lot of nonstandard transactions being held up by eligius?  magicaltux was saying last night that the reason somebody's withdrawal wasn't going thru was cause of eligius, and oddly I have had a huge spike of physical bitcoin orders today with many of them leaving what
358 2011-10-23 20:31:50 <casascius> appears to be real contact info, but none of them paying.  seemed very odd.  but would make sense if they all relied on a single payment processor which depends on eligius.  (most of the orders were russian)
359 2011-10-23 20:32:03 <gmaxwell> casascius: thats not reasonable eligius doesn't solve many blocks.
360 2011-10-23 20:32:40 <gmaxwell> oh I think magicaltux has some deal with eligius to get zero fee processing.
361 2011-10-23 20:32:45 <gmaxwell> and eligius has been down.
362 2011-10-23 20:32:51 <gmaxwell> (it's back up now)
363 2011-10-23 20:33:08 <da2ce7> ahh. hence the free hosting...
364 2011-10-23 20:33:10 <da2ce7> lol
365 2011-10-23 20:33:27 <gmaxwell> Sounds pretty reasonable to me.
366 2011-10-23 20:33:33 <Eliel> yep :)
367 2011-10-23 20:35:31 <casascius> It seems as though with a little careful way to construct transaction outputs, one can avoid fees virtually all of the time... I lately never pay fees on anything, I wonder if my transaction load is unusual.
368 2011-10-23 20:35:43 <gmaxwell> casascius: Correct.
369 2011-10-23 20:35:46 <casascius> i load those 1 BTC coins by the 50 and 100 at a time, and never have to pay a fee
370 2011-10-23 20:36:22 <gmaxwell> Bitcoind is not all that careful though... and depending on your wallet composition you can end up constantly recycling change which results in fees.
371 2011-10-23 20:36:33 <gmaxwell> casascius: Are you using sendmay to load your coins?
372 2011-10-23 20:36:37 <gmaxwell> er sendmany
373 2011-10-23 20:36:39 <casascius> yep, i am doing sendmany
374 2011-10-23 20:36:49 <casascius> and that's just it I suppose...the methodology I use minimizes wallet recycling
375 2011-10-23 20:37:18 <casascius> it sort of happens by accident due to the way i use paper wallets for my btc
376 2011-10-23 20:37:47 <gmaxwell> Yep.  Though mtgox should _also_ use send many.. they should delay each transaction for 30 seconds or so and batch up all the parallel transactions.
377 2011-10-23 20:38:05 <gmaxwell> But IIRC we haven't even gotten [Tycho] to use sendmany. :-/
378 2011-10-23 20:38:36 <gmaxwell> It would be nice it bitcoin could do that batching for you, but the API returns the transaction ID.
379 2011-10-23 20:38:51 <casascius> i receive all my btc to an offline wallet.  once in a while, i will go to the offline wallet, and create a single sendmany that aggregates them all together, and outputs them in 7-10 useful-sized chunks that I load onto a doublesided Bitaddress.org paperwallet (14 addresses)
380 2011-10-23 20:39:00 <gmaxwell> For batching it would need to return a handle that you can query to get the txnid later.
381 2011-10-23 20:39:23 <casascius> and then when it's time to laod coins, i will only scan in what I need, and load coins from that.  in virtually all cases, i will never be reusing fresh change, and that is why I guess I never pay fees
382 2011-10-23 20:39:36 <gmaxwell> Yep.
383 2011-10-23 20:39:55 <gmaxwell> bitcoind makes almost no effort to avoid instantly recycling change once it has one confirm.
384 2011-10-23 20:40:07 <gmaxwell> but it your case, everything is nicely delayed.
385 2011-10-23 20:40:22 <casascius> that makes sense....and i spend my wallet down to zero each time (if i have "change", I will send it to a fresh paper bitcoin address)
386 2011-10-23 20:40:40 <gmaxwell> All the coin selection stuff was written prior to the anti-dos fees.. so you can't blame it.
387 2011-10-23 20:40:59 <casascius> i went and bought a handheld Wasp 2D barcode scanner for this purpose so that makes it really convenient.... a $400 expense that most people aren't likely to incur just to save a fee or two...but i suppose i'm the paper wallet fanatic
388 2011-10-23 20:41:20 <casascius> that said, when you ahve one of those, dealing in paper bitcoins is really easy
389 2011-10-23 20:42:03 <casascius> i assume there has got to be some sort of QR code reader for webcam out there that types on keyboard?  i mean i wonder how else do all qr-code lovers type their qr codes
390 2011-10-23 20:42:57 <gmaxwell> yes, but the wedges seem to all be expensive.
391 2011-10-23 20:43:17 <gmaxwell> normall people use software solutions w/ regular cameras.
392 2011-10-23 20:43:46 <casascius> SOmehow I suspect that it would be possible to still spam the system... I mean if I can do free transactions without limit as long as I let a few days elapse between each one, then it means one could spam the network if they simply took 50 BTC, split it between 5000 addresses, and ping ponged 0.01 BTC as long as each bitcoin only moved once every couple days
393 2011-10-23 20:44:53 <casascius> if true, that sort of spam is scalable to any amount
394 2011-10-23 20:45:17 <gmaxwell> casascius: your rate limited by the ping pongy...
395 2011-10-23 20:45:38 <gmaxwell> each 0.01 would have to sit still for 100 days.
396 2011-10-23 20:45:52 <casascius> why 100 days?  is that dependent on the value?  (i.e. 1/value)
397 2011-10-23 20:46:25 <gmaxwell> The priority calculation for a txn is   sum_over_all_inputs(value * confirms)/tx_data_size
398 2011-10-23 20:46:41 <gmaxwell> And to not have to pay a fee that number must be at least 51m.
399 2011-10-23 20:46:57 <casascius> so then if I had 5000 BTC and split them up into 1 BTC chunks then it would work?  (I suppose that raises the bar some)
400 2011-10-23 20:46:58 <gmaxwell> (value is in 1e-8 units)
401 2011-10-23 20:47:12 <gmaxwell> Yes, increasing the amount of coin increases the volume you can produce.
402 2011-10-23 20:47:40 <gmaxwell> But thats the point of the system: in order to do the attack you must have a lot of coins tied up in the attack, and it doesn't really matter how you split them up.
403 2011-10-23 20:47:41 <casascius> i suppose if people started to refuse to confirm those transactions, then that woudl put a stop to it
404 2011-10-23 20:48:03 <casascius> i suppose iw ould have to tie up coins, but if they are coins one already is sitting on anyway, they aren't losing any utility
405 2011-10-23 20:48:11 <gmaxwell> Right, so say you did ammass 5000 btc for attacking ... eventually we could just greylist those coins.
406 2011-10-23 20:48:13 <casascius> (not that I'm contemplating the attack by any means)
407 2011-10-23 20:48:30 <casascius> Miners would just refuse to include those transactions into blocks
408 2011-10-23 20:49:01 <gmaxwell> And presumably if you've pooled up 5000 btc you actually care something about the health of bitcoin. :)
409 2011-10-23 20:49:07 <gmaxwell> Even if you're not currently spending them. :)
410 2011-10-23 20:50:03 <gmaxwell> 5000 BTC would let you sustain 5000 TXN per day... we can technically handle that in a block or two, thought it would be kind ugly.
411 2011-10-23 20:50:34 <casascius> i mean if i wanted to be a bitcoinexpress and do the attack to create a motivation for the developers to do something about it, i suppose that would be a scenario where one could attack while "caring" about bitcoin
412 2011-10-23 20:50:57 <gmaxwell> You'd probably need around 200k btc to really start breaking things with the current rules.
413 2011-10-23 20:53:10 <gmaxwell> I don't think I've heard any anti-DOS suggestions that were actually any better.
414 2011-10-23 20:53:20 <gmaxwell> Mostly we see complaints that what we have is too agressive.
415 2011-10-23 20:53:44 <gmaxwell> (if it were possible to get a small amount of bitcoin through cpu mining which could be used to pay fees even those complaints would probably go away)
416 2011-10-23 20:55:48 <casascius> i suppose i agree - ultimately, the most powerful antidos would be some way that allows miners and node operators to react and reject threats
417 2011-10-23 20:57:06 <casascius> that's probably way down the road though
418 2011-10-23 20:57:26 <gmaxwell> Unfortunately the privacy properties of bitcoin make this hard.. e.g. every attack txn uses a new address.. the attack funds are frequently cycled through bank like sites.
419 2011-10-23 20:59:18 <casascius> if a future bitcoin client had a means to accept a signed interpreted validation script from some party trusted to manage it, that would be evolved enough to handle it (similar to how antivirus apps "phone home" to get the latest signatures)
420 2011-10-23 20:59:43 <casascius> presumably that's centralization, but by then perhaps there'd be as many bitcoin clients as we have antivirus vendors today
421 2011-10-23 21:00:25 <casascius> obviously hoping we don't need that any time soon
422 2011-10-23 21:00:56 <gmaxwell> That doesn't solve it in any case. How are you going to identify the txn involved in the attack pattern I described above?
423 2011-10-23 21:01:30 <gmaxwell> The attacker will make his traffic look maximally indistinguishable from normal traffic.
424 2011-10-23 21:02:01 <casascius> because it would be a transaction based on coins that are within set {...}... and if they washed them through a bank, that would give the operator time to react (since they already have to wait some time like a day before pingponging each coin)
425 2011-10-23 21:02:15 <gmaxwell> The best we can do is make it costly enough (in terms of fees or coin laying idle) that the attacker would go broke long before the system would be disrupted, so he doesn't even try.
426 2011-10-23 21:03:10 <casascius> right now if i wanted to put $15k USD, I could have 5000 coins (well, I already have 5000 coins of my own, that's besides the point)... and if they only cycled once a day, i wouldn't lose any fees nor use of my coins since i'm already sitting on them
427 2011-10-23 21:03:37 <gmaxwell> You also wouldn't be noticed by bitcoin users.
428 2011-10-23 21:03:58 <gmaxwell> (though you might see a few people here notice)
429 2011-10-23 21:04:07 <casascius> i would just be filling up their hard disk is all
430 2011-10-23 21:04:22 <gmaxwell> fairly slowly, however.. about additional 1mb/day.
431 2011-10-23 21:04:54 <gmaxwell> And those transactions would be prunable, so it would probably just motivate someone implementing that in the official client.
432 2011-10-23 21:05:21 <casascius> Which ultimately woudl improve bitcoin, not hinder it.
433 2011-10-23 21:06:01 <gmaxwell> Sure, though we don't need an attack to get that. I'll be needed eventually anyways.
434 2011-10-23 21:06:47 <gmaxwell> If anyone gave a #@$@ about the alt chains it might be fun to attack them to get them to develop the software.   (evil laugh)
435 2011-10-23 21:07:07 <gmaxwell> The trouble NMC had with hash rate swinging got them to build merged mining.
436 2011-10-23 21:07:09 <Diablo-D3> gmaxwell: whats that one cpu only chain
437 2011-10-23 21:07:38 <gmaxwell> trixyloops or something
438 2011-10-23 21:08:01 <Diablo-D3> I should gpu ot.
439 2011-10-23 21:08:02 <gmaxwell> Although.. it only needs 128k of memory per instance, so it's hardly asic/gpu incompatible.
440 2011-10-23 21:08:33 <casascius> sum_over_all_inputs(value * confirms) / data size, units are satoshis, confirms, and bytes/kilobytes?
441 2011-10-23 21:08:54 <Diablo-D3> gmaxwell: I mean hell, Im writing a codec to troll xiph with.
442 2011-10-23 21:08:55 <casascius> must >= 5.1e6
443 2011-10-23 21:08:55 <gmaxwell> casascius satoshis, confirms, and bytes.
444 2011-10-23 21:09:06 <Diablo-D3> gmaxwell: it doesnt actually work, but at least Im trying.
445 2011-10-23 21:09:28 <Diablo-D3> casascius... why does that nick sound familiar
446 2011-10-23 21:10:05 <gmaxwell> Diablo-D3:
447 2011-10-23 21:10:11 <gmaxwell> Diablo-D3: https://www.casascius.com/
448 2011-10-23 21:10:18 <Diablo-D3> oh right! the coin guy!
449 2011-10-23 21:10:23 <JFK911> Diablo-D3: becauze hes the one selling gambling tokens as bitcoins
450 2011-10-23 21:10:44 <Diablo-D3> casascius: you should give coins to influencial bitcoin people.
451 2011-10-23 21:11:05 <casascius> gambling tokens would be cool...i considered bitcoin poker chips...these are more like carwash tokens
452 2011-10-23 21:11:19 <JFK911> vegas casinos dont have rfid metal tokens yet?
453 2011-10-23 21:11:19 <neofutur> interesting question on stackexchange : http://bitcoin.stackexchange.com/questions/1708/capabilities-of-bitcoins-and-their-place-in-the-future
454 2011-10-23 21:11:20 <Diablo-D3> casascius: so we can show them off to people to generate business
455 2011-10-23 21:11:30 <neofutur> any expert here to answer this ?
456 2011-10-23 21:11:31 <Diablo-D3> JFK911: oh they have for years
457 2011-10-23 21:11:34 <JFK911> the coin idea is really cool
458 2011-10-23 21:11:37 <casascius> diablo-d3: actually, I am interested, I generally want to keep the BTC though, anybody considered "influential" i'd sell them to at face value
459 2011-10-23 21:11:42 <JFK911> Diablo-D3: yeah plastic ones have been for a decade
460 2011-10-23 21:11:56 <casascius> in reasonable quantities of course (they take time to make)
461 2011-10-23 21:11:58 <Diablo-D3> JFK911: the high value ones are all metal at some places
462 2011-10-23 21:12:04 <JFK911> i found an rfid on a packet of socks
463 2011-10-23 21:12:16 <Diablo-D3> like remember that time someone stole a shitload of high value poker chips from vegas?
464 2011-10-23 21:12:22 <Diablo-D3> they were metal coins
465 2011-10-23 21:12:34 <Diablo-D3> casascius: even just 1 of the 1btc ones?
466 2011-10-23 21:12:55 <JFK911> well for samples/hype they dont need to have actual bitcoin sticker on them
467 2011-10-23 21:13:01 <casascius> i would probably do 1x1btc for free, but would rather give a handful
468 2011-10-23 21:13:06 <Diablo-D3> JFK911: unless I want to photograph it
469 2011-10-23 21:13:18 <JFK911> hes got some beautiful photos already
470 2011-10-23 21:13:31 <Diablo-D3> yes, but I have a blog. with a huge readership.
471 2011-10-23 21:13:32 <JFK911> i think he said you can use them too
472 2011-10-23 21:13:47 <Diablo-D3> and Im the author of diablominer. which makes me awesome.
473 2011-10-23 21:14:00 <casascius> maybe what i should do is create "redemption codes" kind of how mtgox says "we want to give you a free yubikey, order and use code xxxxxx when you check out"
474 2011-10-23 21:14:11 <JFK911> i havent seen a photo of a broken one yet
475 2011-10-23 21:14:12 <Diablo-D3> casascius: maybe
476 2011-10-23 21:14:28 <casascius> cause i'd be more than happy to give a half dozen of the 25btc coins to quite a few people, but would prefer not to give away 150btc
477 2011-10-23 21:14:34 <gmaxwell> casascius: you need someone to build you a robot for assembly. Your current prices are a hinderance for using them as a btc promotion tool.
478 2011-10-23 21:14:42 <casascius> gmaxwell: 100% agreed
479 2011-10-23 21:15:00 <JFK911> how many production steps do you have
480 2011-10-23 21:15:04 <casascius> and it sucks, cause i can blow a whole saturday just putting together a few hundred coins
481 2011-10-23 21:15:29 <gmaxwell> E.g. I'd like to go throw a hand full of btc in geocaches around me, but at 30% overhead on the 1btc coins  I'd rather write keys on paper and be able to put out more coins for the same spend.
482 2011-10-23 21:16:02 <casascius> if you're going to geocache, putting the coins on paper is almost better - you can take them back much later if no one cares to collect them
483 2011-10-23 21:16:16 <gmaxwell> Thats true too.
484 2011-10-23 21:16:24 <casascius> a coin on paper, with a 1 btc "token" glued to it, wouldn't be bad (in other words, the 1 btc coin I have, with no hologram or key)
485 2011-10-23 21:16:41 <casascius> i would part with just the physical coins for pretty cheap for promotional purposes...they are easy to box and ship
486 2011-10-23 21:16:52 <casascius> especially in bulk
487 2011-10-23 21:16:59 <gmaxwell> Though introducing people to the concept of your tokens would be good too.  ... hm. thats a reasonable half step.
488 2011-10-23 21:17:12 <casascius> i got like 8000+ of them left
489 2011-10-23 21:17:12 <JFK911> casascius: you wouldnt want to put an orange sticker or something on to make it obvious they werent bitcoins?
490 2011-10-23 21:17:43 <casascius> possibly, but the point of shipping the coins in bulk was to get out of stickering them one by one.  if i'm going to sticker them, might as well sticker them with a hologram and make them legit! =)
491 2011-10-23 21:17:57 <casascius> however, at least someone else could do the stickering
492 2011-10-23 21:18:12 <JFK911> well im just thinking about coins and how people think their current coins are valid just because they can hold them in their hand
493 2011-10-23 21:18:36 <gmaxwell> does kind of add confusion "but I thought it wasn't valid without the hologram?"
494 2011-10-23 21:18:54 <casascius> another reasonable idea would be to just make a promotional edition coin
495 2011-10-23 21:18:58 <casascius> that doesn't purport to be "a bitcoin"
496 2011-10-23 21:19:04 <JFK911> but with your btc coins most people dont know why the sticker should be intact
497 2011-10-23 21:19:08 <gmaxwell> it could be square perhaps.
498 2011-10-23 21:19:17 <JFK911> or a stamp "sample"
499 2011-10-23 21:19:32 <gmaxwell> sized to make it easy to put a mailing label type sticker on it for a giftcard.
500 2011-10-23 21:20:00 <casascius> i have ordered some window holograms which I believe might enable me to delegate the production of the coins...the bitcoin address would show through the window....then there would be no concern that the private key was inside
501 2011-10-23 21:20:19 <casascius> then i could have a bunch of temp workers over for a day while they stuffed coins and I made sure no one kept any or copied any keys
502 2011-10-23 21:20:26 <JFK911> or take a grinder to a stack and flatten part of an edge
503 2011-10-23 21:20:30 <casascius> that's about the only viable solution i can think of
504 2011-10-23 21:20:57 <gmaxwell> JFK911: they'd need to be very distinct. If they look coinlike people will look for the hologram.
505 2011-10-23 21:21:08 <gmaxwell> (as they should)
506 2011-10-23 21:21:34 <JFK911> well, that's how they cancel drivers licenses, and record albums that cant be sold at full price
507 2011-10-23 21:21:39 <JFK911> just some cut
508 2011-10-23 21:21:53 <casascius> I would be willing to sell a bag of 1000 bitcoin blanks for $300 or BTC equivalent plus shipping
509 2011-10-23 21:22:30 <JFK911> casascius: where's the private key stored?  the sticker makes an envelope or there is storage inside the token?
510 2011-10-23 21:22:57 <casascius> these coins would have no private keys..they are the brass portion only....otherwise, the private key is under the sticker (there's a hollowed out spot for a little piece of paper so the sticker is flush)
511 2011-10-23 21:23:05 <gmaxwell> http://img199.imageshack.us/img199/6272/physbitcoinkey.jpg
512 2011-10-23 21:23:08 <JFK911> OH thats awesome
513 2011-10-23 21:23:31 <JFK911> neat, i hadn't seen this before
514 2011-10-23 21:23:39 <casascius> those pictures on my website where i show stacks of coins....that's one bag of 1000 coins worth
515 2011-10-23 21:23:45 <JFK911> And it's polymer or something else that should be laundry-proof?
516 2011-10-23 21:24:33 <casascius> the sticker is likely to hold up in the laundry but the inkjetted bitcoin address on the exterior will probably wear...It would probably survive a single wash, but won't survive a lot of solvents nor constant wear
517 2011-10-23 21:25:18 <JFK911> pretty good
518 2011-10-23 21:25:44 <JFK911> maybe the window will be easier production
519 2011-10-23 21:25:49 <JFK911> 2-sided laser printer
520 2011-10-23 21:25:57 <JFK911> then die cut the dots out
521 2011-10-23 21:26:16 <casascius> i would love to get the cost down, because i know that a 30-35% markup on a coin is bullcrap....but my time is just limited
522 2011-10-23 21:26:39 <casascius> hoping the 25 BTC coin makes more sense at this point...much better deal, because it's less time commitment per BTC
523 2011-10-23 21:26:52 <gmaxwell> yea, I don't fault you for that.
524 2011-10-23 21:27:13 <gmaxwell> The 25 BTC coin makes more sense as a token for actually obtaining bitcoins... and about no sense as a promotional tool.
525 2011-10-23 21:29:04 <casascius> Got to run, back later
526 2011-10-23 21:35:35 <Diablo-D3> back
527 2011-10-23 21:35:51 <Mad7Scientist> When a block is solved does that mean somebody find a hash with a certain number of leading zeros?
528 2011-10-23 21:36:12 <Mad7Scientist> I'll ask in -otc
529 2011-10-23 21:36:27 <Diablo-D3> wasnt someone doing btc cards of high values at one point?
530 2011-10-23 21:36:32 <Diablo-D3> Mad7Scientist: yes
531 2011-10-23 21:36:42 <Diablo-D3> although the zeros in this sense are binary
532 2011-10-23 21:36:47 <Diablo-D3> so you can just directly check the value
533 2011-10-23 21:37:32 <Mad7Scientist> Diablo-D3, how can you have difficulty levels that are not * of 2 with that method?
534 2011-10-23 21:38:01 <Diablo-D3> because diff is checked by just comparing the second 32 bits against the number of diff
535 2011-10-23 21:38:24 <Diablo-D3> ie, we're at 1.x million now, so the second 32 bits needs to be less than 1.x million
536 2011-10-23 21:38:33 <Mad7Scientist> I see
537 2011-10-23 21:38:40 <Mad7Scientist> I was thinking
538 2011-10-23 21:39:00 <Mad7Scientist> Could a miner solve future blocks at the same time?
539 2011-10-23 21:39:14 <Diablo-D3> well, less than 2**32 minus 1.x million, but you get the point
540 2011-10-23 21:39:19 <Diablo-D3> no
541 2011-10-23 21:39:19 <Mad7Scientist> Say it finds a hash that is short by 1 zero
542 2011-10-23 21:39:33 <Mad7Scientist> could it hang on to that and then when difficulty goes down in the future use that one
543 2011-10-23 21:39:38 <Diablo-D3> the block header has the previous block's hash in it
544 2011-10-23 21:40:09 <Mad7Scientist> But is the hash that the miner is trying to solve related to the previous block at all?
545 2011-10-23 21:40:14 <Diablo-D3> yes
546 2011-10-23 21:40:27 <Diablo-D3> because the miner is fed the header of the potential block to hash
547 2011-10-23 21:41:00 <ThomasV> what is the last byte of the signature in ScriptSig ?
548 2011-10-23 21:44:52 <lianj> ThomasV: hash type?
549 2011-10-23 22:20:51 <gmaxwell> Diablo-D3: looks like eligius is back up to 150GH/s.
550 2011-10-23 22:21:04 <Diablo-D3> not very much.
551 2011-10-23 22:21:06 <gmaxwell> Diablo-D3: it would be nice if you'd update the forum to not show 0.
552 2011-10-23 22:21:21 <Diablo-D3> erm, does luke know he can do it himself?
553 2011-10-23 22:21:50 <gmaxwell> Luke can't, his laptop is dead, he fixed the pool today over a cellphone.
554 2011-10-23 22:21:55 <Diablo-D3> lawl
555 2011-10-23 22:24:48 <Diablo-D3> hes now #10.
556 2011-10-23 22:35:47 <imsaguy> its funny you'll edit it to say 0, but you won't edit it to fix it
557 2011-10-23 22:36:01 <gmaxwell> imsaguy: he fixed it
558 2011-10-23 22:36:15 <imsaguy> excellent
559 2011-10-23 22:36:20 <Diablo-D3> [08:24:47] <Diablo-D3> hes now #10.
560 2011-10-23 22:36:24 <Diablo-D3> ^ proof I fixed it
561 2011-10-23 23:30:20 <CIA-101> bitcoin: phantomcircuit * ra419a0073feb bitcoin-alt/ (4 files in 3 dirs): add indexes to default template improve algorithm for getting blocks