1 2012-11-17 01:36:08 <dparrish> ;;rate frend 1 Bought my BTC for bank deposit, he paid first
  2 2012-11-17 01:36:09 <gribble> Rating entry successful. Your rating of 1 for user frend has been recorded.
  3 2012-11-17 02:05:39 <etotheipi_> sipa, it's Bitcoin-Qt 0.7.1
  4 2012-11-17 02:06:15 <etotheipi_> I have a raw dump of the tx's that came in, but they came into the buffer all at once, so I guess there's a chance my code didn't parse them right
  5 2012-11-17 02:07:24 <etotheipi_> I just gotta manually parse them and confirm
  6 2012-11-17 02:17:18 <etotheipi_> jgarzik: sipa:   sorry, false alarm -- my code is somehow not breaking apart messages correctly when they all appear in the same buffer
  7 2012-11-17 02:48:19 <etotheipi_> oh man... this is one of those python-sometimes-things-are-references-sometimes-copy issues
  8 2012-11-17 02:52:30 <Luke-Jr> etotheipi_: pretty sure Python doesn't have that problem really?
  9 2012-11-17 02:52:45 <Luke-Jr> isn't everything a reference unless you use copy() or deepcopy()?
 10 2012-11-17 02:53:05 <etotheipi_> Luke-Jr: non-mutable objects are always copies
 11 2012-11-17 02:53:05 <Luke-Jr> and just the confusion stems from the inability to manipulate strings/numbers
 12 2012-11-17 02:53:17 <Luke-Jr> etotheipi_: non-mutable objects just aren't mutable ;)
 13 2012-11-17 02:53:25 <etotheipi_> mutable objects are always passed by reference
 14 2012-11-17 02:53:32 <Luke-Jr> if it's immutable, there is no distinction between a reference and a copy
 15 2012-11-17 02:53:40 <cjd> have you guys seen an issue with gcc segfaulting? I'm suddenly having people complain about it.. first time it was gentoo so kind of semi-expected, now centos-5.5 with gcc-4.4
 16 2012-11-17 02:54:02 <cjd> oops centos 6.3
 17 2012-11-17 02:55:14 <etotheipi_> Luke-Jr: the issue is when you're mixing lots of mutable and non-mutable
 18 2012-11-17 02:56:15 <etotheipi_> mainly containers that hold nonmutables and/or other containers
 19 2012-11-17 03:00:41 <xenland> How does one benchmark their financial software? Is there a predefined list of math formulas that display floating point errors that I could use and see if my integer only software is doing what it is suppsoed to?
 20 2012-11-17 03:01:27 <cjd> testing for rounding errors?
 21 2012-11-17 03:02:58 <xenland> cjd: Yeah i think thats the terminology im looking for.
 22 2012-11-17 03:03:08 <etotheipi_> xenland: http://xkcd.com/217/
 23 2012-11-17 03:03:14 <cjd> perhaps compare it to something known like openssl?
 24 2012-11-17 03:04:13 <xenland> ethotheipi_: lol cuz pi precision goes on foreva
 25 2012-11-17 03:04:30 <xenland> cjd:How do you mean?
 26 2012-11-17 03:05:30 <cjd> run random numbers in a loop and compare
 27 2012-11-17 03:06:37 <cjd> for example: https://ezcrypt.it/VB5n#uJeVApbiVpWCllLHLJmDAnsM
 28 2012-11-17 03:08:33 <xenland> interesting
 29 2012-11-17 03:08:52 <jgarzik> etotheipi_: cool
 30 2012-11-17 03:09:02 <jgarzik> etotheipi_: bitcoin will definitely send multiple messages in the same network packet, sometimes
 31 2012-11-17 03:10:09 <phantomcircuit> jgarzik, limited to high load
 32 2012-11-17 03:10:40 <jgarzik> phantomcircuit: ?
 33 2012-11-17 04:39:42 <phantomcircuit> jgarzik, you still around?
 34 2012-11-17 04:39:45 <phantomcircuit> cjd, ditto to you
 35 2012-11-17 04:39:48 <jgarzik> phantomcircuit: yes
 36 2012-11-17 04:40:00 <phantomcircuit> can you try and connect a bitcoind to metaexch.com:8333
 37 2012-11-17 04:40:24 <phantomcircuit> im currently attacking the hell out of it and i want to see if my anti ddos code works
 38 2012-11-17 04:41:12 <jgarzik> phantomcircuit: sorry, no
 39 2012-11-17 04:41:16 <jgarzik> busy
 40 2012-11-17 04:41:19 <phantomcircuit> :(
 41 2012-11-17 04:41:32 <cjd> also busy
 42 2012-11-17 04:41:34 <cjd> maybe later
 43 2012-11-17 04:41:37 <cjd> debugging gcc
 44 2012-11-17 04:41:39 <phantomcircuit> sipa, ?
 45 2012-11-17 04:41:53 <cjd> redhat gcc segfaults trying to compile cjdns with -O3 -g
 46 2012-11-17 04:42:17 <amiller> phantomcircuit, i'll connect a bitcoind to it
 47 2012-11-17 04:42:40 <phantomcircuit> horray
 48 2012-11-17 04:46:29 <gmaxwell> cjd: open a bug.
 49 2012-11-17 04:47:05 <gmaxwell> cjd: my median time to fix on reproducable ICE I've reported on GCC is <24 hours.
 50 2012-11-17 04:47:10 <amiller> plenty of "socket recv error 104" phantomcircuit
 51 2012-11-17 04:47:35 <phantomcircuit> hmm
 52 2012-11-17 04:47:37 <Luke-Jr> cjd: Is -O3 even supposed to be reliable?
 53 2012-11-17 04:47:51 <Luke-Jr> I know RHEL 6's GCC miscompiles with -O2
 54 2012-11-17 04:48:13 <gmaxwell> Luke-Jr: sure it is.
 55 2012-11-17 04:48:15 <phantomcircuit> ************************
 56 2012-11-17 04:48:16 <amiller> O2 is supposed to be reliable
 57 2012-11-17 04:48:21 <phantomcircuit> hmm
 58 2012-11-17 04:48:31 <amiller> O3 is reliable on every thing i know of
 59 2012-11-17 04:48:34 <phantomcircuit> Luke-Jr, it's reliable as long as you're following the standards
 60 2012-11-17 04:48:38 <phantomcircuit> which most people dont :)
 61 2012-11-17 04:48:54 <Luke-Jr> phantomcircuit: shrug, not my code. I just want it to work :p
 62 2012-11-17 04:49:07 <gmaxwell> Luke-Jr: there are a lot of negative opinions of O3 that come from aliasing violations (and now O2, since O2 picked up some of those optimizations)
 63 2012-11-17 04:49:45 <Luke-Jr> gmaxwell: I see. Gentoo has a bad reputation for enabling overoptimization, but ironically they don't advise any more than -O2 on x86
 64 2012-11-17 04:50:23 <Luke-Jr> I personally have a completely hand-picked set of optimizations, because I don't like that -O1 makes debugging harder, and -O0 omits obvious crap like inlining
 65 2012-11-17 04:50:52 <gmaxwell> Luke-Jr: there is a lot of broken code out there _and_ there certantly are compiler bugs. But there aren't supposted to be an and GCC devs give fixing them very high priority.
 66 2012-11-17 04:51:39 <Luke-Jr> gmaxwell: well, I would expect some optimizations are inherently broken in most cases, but still useful when you have complete control of the code and need the performance
 67 2012-11-17 04:51:53 <Luke-Jr> or maybe not most cases, but some standards-acceptable ones
 68 2012-11-17 04:52:09 <gmaxwell> Things have gotten a bunch better in recent years because some academics created a C language fuzzer that produces random but valid code... and then they compile with N compilers and detect differences in behavior.
 69 2012-11-17 04:52:34 <Luke-Jr> hmm
 70 2012-11-17 04:53:25 <Luke-Jr> gmaxwell: what I'd really like to see, but haven't yet (and I know how difficulty it is because I've tried to do it myself), is an emulator that follows all possible code paths and reports any way to reach (memory corruption | segfault | etc)
 71 2012-11-17 04:53:40 <Luke-Jr> difficult*  - been working on bitcoin too much! XD
 72 2012-11-17 04:54:01 <gmaxwell> Luke-Jr: "static analysis"  I mean thats what clang analize, coverity, and a zillion other tools do.
 73 2012-11-17 04:54:07 <gmaxwell> I recommend them.
 74 2012-11-17 04:54:12 <Luke-Jr> do they? :o
 75 2012-11-17 04:54:56 <gmaxwell> They're not sound they don't hit all possible. But they do catch a lot.
 76 2012-11-17 04:55:14 <gmaxwell> They also have false alarms (and minimizing them is why they don't trace all paths)
 77 2012-11-17 04:55:16 <Luke-Jr> hm
 78 2012-11-17 04:55:31 <cjd> gmaxwell: yeap, I'm golfing it down to the simplest code which will reproduce the bug, I'm down to 400 lines already.. Hopefully I can write a little cmake test to find bad compilers as well as giving RH something they can very easily act on (this is RH's issue because it's gcc-4.4)
 79 2012-11-17 04:55:44 <gmaxwell> Example output (a false alarm in this case, stride must be >=1 but it can't tell that): https://mf4.xiph.org/jenkins/view/opus/job/opus-scan-build/ws/scan-build/current/report-Dhq6sz.html#EndPath
 80 2012-11-17 04:55:50 <Luke-Jr> gmaxwell: I was thinking of working at the bytecode level, and storing memory contents as some kind of a quantum state
 81 2012-11-17 04:56:19 <gmaxwell> cjd: All you need to do is get the preprocessed output using gcc -E  and then thats a standalone reproduction. Don't bother reducing it further if you don't want.
 82 2012-11-17 04:56:52 <gmaxwell> cjd: you can attach that to a ticket (on the redhat system if you haven't reproduced with stock gcc; otherwise the gcc bugzilla) and thats all someone will need to reproduce.
 83 2012-11-17 04:57:14 <gmaxwell> Luke-Jr: Well tools like klee do that to drive fuzz testing.. but the state is exponential.
 84 2012-11-17 04:58:03 <Luke-Jr> of course. unless you can find a good way to identify loops ;)
 85 2012-11-17 04:58:14 <gmaxwell> Source code level analysis is also more likely to produce a result that makes any sense.
 86 2012-11-17 04:58:17 <Luke-Jr> identify when a path results in the same state you had before
 87 2012-11-17 04:58:48 <gmaxwell> E.g. if you had that clang-analysis output above where it only told you the [4] without showing you the trace that got it there, you'd waste a half hour trying to figure out if it was right or not and why. :P
 88 2012-11-17 05:06:25 <Luke-Jr> gmaxwell: well, something tracing actual possible code paths is never wrong :P
 89 2012-11-17 05:14:22 <phantomcircuit> uh
 90 2012-11-17 05:14:24 <phantomcircuit> is it just me
 91 2012-11-17 05:14:32 <phantomcircuit> or is continue not working in BOOST_FOREACH
 92 2012-11-17 05:15:16 <Diablo-D3> heh, boost =/
 93 2012-11-17 05:15:30 <Diablo-D3> [12:49:07] <gmaxwell> Luke-Jr: there are a lot of negative opinions of O3 that come from aliasing violations (and now O2, since O2 picked up some of those optimizations)
 94 2012-11-17 05:15:50 <Diablo-D3> doesnt -O3 enable the optimization that pretends all function arguments are c99 restrict?
 95 2012-11-17 05:22:28 <phantomcircuit> lol there's a pretty obvious logic error in my code
 96 2012-11-17 05:22:29 <phantomcircuit> sigh
 97 2012-11-17 05:22:41 <Diablo-D3> phantomcircuit: dont you love those?
 98 2012-11-17 05:22:55 <Diablo-D3> Ive been doing "sleep coding" lately
 99 2012-11-17 05:23:00 <Diablo-D3> and it produces horrid code
100 2012-11-17 05:23:16 <Diablo-D3> "hrm, I havent produced enough code today, and I want to go to bed, so lemme bang out a few lines"
101 2012-11-17 05:23:21 <phantomcircuit> that's how this code came to be
102 2012-11-17 05:23:25 <Diablo-D3> * two hours later *
103 2012-11-17 05:23:34 <Diablo-D3> "there, done, Ill see if it compiles tommorow"
104 2012-11-17 05:23:37 <Diablo-D3> * tommorow *
105 2012-11-17 05:23:40 <phantomcircuit> WAT
106 2012-11-17 05:23:42 <Diablo-D3> "WHAT THE FUCK IS THIS SHIT I WROTE"
107 2012-11-17 05:24:19 <Diablo-D3> atm Im trying to make re2c and lemon play nice, so thats much more fun
108 2012-11-17 05:24:51 <Diablo-D3> they're both candidates of being turned into seaking compiler plugins
109 2012-11-17 05:25:22 <jgarzik> bah, who needs lemon?  http://shop.oreilly.com/product/9780596155988.do
110 2012-11-17 05:25:38 <D34TH> Diablo-D3, http://static.fjcdn.com/pictures/fuck_545be4_531549.gif
111 2012-11-17 05:25:40 <Diablo-D3> jgarzik: FUCK THAT
112 2012-11-17 05:25:58 <Diablo-D3> jgarzik: fuck lex and fuck yacc
113 2012-11-17 05:26:07 <Diablo-D3> D34TH: <3
114 2012-11-17 05:26:24 <Diablo-D3> jgarzik: I want a modern fucking parser, not some slow ass bloated shit
115 2012-11-17 05:27:24 <Diablo-D3> lemon+re2c can do it faster with lower memory overhead and less generated code
116 2012-11-17 05:29:14 <Diablo-D3> also, I have discovered the most hilarious thing in C that I've never noticed before
117 2012-11-17 05:29:50 <Diablo-D3> sys/types, sys/stat, fcntl are included for open(2)
118 2012-11-17 05:30:02 <Diablo-D3> unistd is included for close(2)
119 2012-11-17 05:30:11 <Diablo-D3> (according to their manpages, anyways)
120 2012-11-17 05:31:09 <Diablo-D3> yup, its true, if I include just those three, close isnt defined
121 2012-11-17 05:33:08 <phantomcircuit> looking at my code now i realize
122 2012-11-17 05:33:13 <phantomcircuit> this calls for xor
123 2012-11-17 05:33:36 <Diablo-D3> phantomcircuit: MOAR XOAR
124 2012-11-17 05:34:43 <D34TH> http://upload.wikimedia.org/wikipedia/commons/thumb/d/d6/Z2%5E4%3B_Cayley_table%3B_binary.svg/498px-Z2%5E4%3B_Cayley_table%3B_binary.svg.png
125 2012-11-17 05:35:26 <Diablo-D3> D34TH: whats that
126 2012-11-17 05:36:17 <D34TH> nimber addition
127 2012-11-17 05:37:57 <phantomcircuit> wait no i can simplify this to a single !
128 2012-11-17 05:38:01 <phantomcircuit> oh well
129 2012-11-17 05:39:10 <cjd> 00:47 < gmaxwell> Luke-Jr: there are a lot of negative opinions of O3 that come from aliasing violations (and now O2, since O2 picked up some of  those optimizations) <-- I do quite a bit of aliasing abuse but I seem to get away with it as long as I know in advance that things are aligned.
130 2012-11-17 05:39:11 <phantomcircuit> the anti ddos code i wrote totally screws up that all the checks actually have three results
131 2012-11-17 05:39:45 <phantomcircuit> new worst peer, equal peers, current worst peer is worse
132 2012-11-17 05:39:55 <phantomcircuit> i ignore that later case currently resulting in strange results
133 2012-11-17 05:40:37 <cjd> The other thing which I have found useful is to use -O3 in development and have lots of tests and lots of buildbots on different machines
134 2012-11-17 05:41:18 <phantomcircuit> -Wall -Werror
135 2012-11-17 05:41:26 <phantomcircuit> oh hey there subtle bugs
136 2012-11-17 05:41:27 <phantomcircuit> :)
137 2012-11-17 05:41:31 <cjd> -pedantic -Wextra
138 2012-11-17 05:41:41 <phantomcircuit> ALL THE FLAGS
139 2012-11-17 05:41:55 <cjd> and then test hard
140 2012-11-17 05:42:10 <Diablo-D3> heh
141 2012-11-17 05:42:12 <cjd> because weird stuff happens when you break the rules a little bit
142 2012-11-17 05:42:12 <Diablo-D3> its funny
143 2012-11-17 05:42:22 <Diablo-D3> I usually specify wall wextra pedantic....
144 2012-11-17 05:42:23 <cjd> ok maybe sometimes a lot :)
145 2012-11-17 05:42:30 <Diablo-D3> then start turning off annoying warnings.
146 2012-11-17 05:42:40 <Diablo-D3> like unused function arg... on main.
147 2012-11-17 05:42:44 <Diablo-D3> seriously, shut the fuck up gcc
148 2012-11-17 05:43:00 <Diablo-D3> I have to specify a valid main prototype, I dont have to use the fucking args you whore.
149 2012-11-17 05:43:05 <cjd> yeah, same.. I only turn them off if there is a valid reason to really need not to have it
150 2012-11-17 05:45:18 <phantomcircuit> Diablo-D3, iirc there is a macro to disable that for specific functions
151 2012-11-17 05:46:11 <Diablo-D3> yeah but that uglies code
152 2012-11-17 05:48:16 <phantomcircuit> ok code now works
153 2012-11-17 05:48:22 <phantomcircuit> sillyness
154 2012-11-17 05:54:35 <phantomcircuit> ok so now you can connect to that bitcoind even when i have hundreds of inbound connections
155 2012-11-17 05:54:47 <phantomcircuit> but it only works if im attacking for the same netblock
156 2012-11-17 05:55:02 <phantomcircuit> protection becomes erratic at best when they're coming from multiple ips
157 2012-11-17 06:00:01 <cjd> https://bugzilla.redhat.com/show_bug.cgi?id=877605
158 2012-11-17 06:05:12 <phantomcircuit> getting hundreds of connections through tor is ridiculously slow
159 2012-11-17 06:05:20 <phantomcircuit> but im too lazy to apply patches to change that
160 2012-11-17 06:45:40 <phantomcircuit> ok screw it custom tor build time
161 2012-11-17 07:16:25 <phantomcircuit> yeah so the jist of this is
162 2012-11-17 07:16:29 <phantomcircuit> the attack still works
163 2012-11-17 07:16:36 <phantomcircuit> but it's way wayyyy more annoying to carry out
164 2012-11-17 07:16:45 <Diablo-D3> http://loopj.com/2012/11/16/nvd3-erased-from-existence/
165 2012-11-17 08:15:04 <Diablo-D3> er guys
166 2012-11-17 08:15:10 <Diablo-D3> whats the default namecoin data and rpc port?
167 2012-11-17 08:15:15 <Diablo-D3> I thought it was 9333 and 9332
168 2012-11-17 08:22:09 <Diablo-D3> huh apparently its 8334 and 8336
169 2012-11-17 08:28:24 <xenland> what time does Bitcoin go by?
170 2012-11-17 08:28:44 <Diablo-D3> utc I think
171 2012-11-17 08:28:50 <xenland> k thanks diablo
172 2012-11-17 08:29:08 <Diablo-D3> logs I think are in your local time, though
173 2012-11-17 08:31:48 <xenland> And your UTC comment was referring to the accepted block timestamp timezone? (As i understand it there is a time zone sync in the protocal it self)
174 2012-11-17 08:32:20 <Diablo-D3> yes
175 2012-11-17 08:32:32 <jgarzik> whee
176 2012-11-17 08:32:33 <xenland> cool
177 2012-11-17 08:32:40 <jgarzik> transcribing EvalScript() is much faster
178 2012-11-17 08:32:42 <Diablo-D3> at least, every representation Ive seen of the timestamp has been UTC
179 2012-11-17 08:32:44 <jgarzik> when going C++ -> C
180 2012-11-17 08:33:30 <Luke-Jr> jgarzik: are you just porting satoshi's code to every language you can? :P
181 2012-11-17 08:33:37 <xenland> Any plans on releasing official documentation on Bitcoin library?
182 2012-11-17 08:33:59 <xenland> jgarzik: make your next client be made in PHP :P
183 2012-11-17 08:34:04 <Luke-Jr> xenland: if you mean libbitcoin, the guy behind that isn't on IRC anymore
184 2012-11-17 08:34:07 <Diablo-D3> lol
185 2012-11-17 08:34:28 <xenland> Libbitcoin huh?
186 2012-11-17 08:34:54 <jgarzik> Luke-Jr: just C and python ;p
187 2012-11-17 08:34:56 <jgarzik> xenland: bleh :)
188 2012-11-17 08:35:01 <xenland> heh
189 2012-11-17 08:35:21 <xenland> luke-jr: I thought libbitcoin was for converting the block chain to Database or no?
190 2012-11-17 08:35:33 <Luke-Jr> xenland: I don't know, not familiar with it.
191 2012-11-17 08:35:48 <xenland> oh okay, thanks anyways
192 2012-11-17 08:36:17 <cjd> jgarzik: +1
193 2012-11-17 08:36:22 <cjd> C ftw
194 2012-11-17 08:36:42 <cjd> I have a couple small functions from bitcoin which I ported to C and they are wildly faster
195 2012-11-17 09:23:25 <phantomcircuit> xenland, it's a full client...
196 2012-11-17 09:23:55 <xenland> phantomcircuit: thanks for clearing that up
197 2012-11-17 09:31:58 <Diablo-D3> argh
198 2012-11-17 09:32:05 <Diablo-D3> whats the all account for listtransactions
199 2012-11-17 09:32:35 <xenland> listtransactions '' 9999999999
200 2012-11-17 09:32:37 <xenland> ?
201 2012-11-17 09:33:12 <Diablo-D3> yeah thats it
202 2012-11-17 09:36:01 <phantomcircuit> Diablo-D3, *+
203 2012-11-17 09:36:02 <phantomcircuit> er
204 2012-11-17 09:36:03 <phantomcircuit> *
205 2012-11-17 09:36:07 <phantomcircuit> the default account is ''
206 2012-11-17 09:36:14 <Diablo-D3> well Im only using default
207 2012-11-17 09:36:15 <Diablo-D3> so heh
208 2012-11-17 12:49:47 <Ferroh> This guy says that running bitcoin -checkblocks=0 took 10 minutes
209 2012-11-17 12:49:51 <Ferroh> should that not take many hours?
210 2012-11-17 12:55:05 <Ferroh> Does -checkblocks=0 do the same verification that bitcoind normally does at startup?
211 2012-11-17 12:57:17 <sipa> it IS the check bitcoin does at startup
212 2012-11-17 12:57:27 <Ferroh> So why is it taking this guy only 10 minutes?
213 2012-11-17 12:57:27 <sipa> but normally only the last 2500 blocks
214 2012-11-17 12:57:41 <Ferroh> well checkblocks=0 supposedly checks the entire chain
215 2012-11-17 12:57:45 <sipa> yes
216 2012-11-17 12:57:56 <sipa> but far from the same level of validation as is done when connecting
217 2012-11-17 12:57:59 <sipa> it checks blocks
218 2012-11-17 12:58:08 <sipa> it doesn't check the consistency of the index
219 2012-11-17 12:58:20 <Ferroh> ah ok
220 2012-11-17 12:59:44 <Ferroh> Is there a command to force checking the index for benchmarking purposes?
221 2012-11-17 13:00:29 <sipa> 0.8 will have -reindex
222 2012-11-17 13:00:41 <sipa> which does the same as -loadblock, but using the existing block files
223 2012-11-17 13:00:58 <Ferroh> Well, I was trying to prove to this person that he is disk bound (not network bound).
224 2012-11-17 13:01:05 <Ferroh> So the idea was to do it in 0.7
225 2012-11-17 13:01:14 <Ferroh> but it's not really a big deal, I suppose.
226 2012-11-17 13:01:50 <sipa> you can move the blk000?.dat files away, delete blkindex.dat, and then start with -loadblock=/path/to/blk0001.dat -loadblock=/path/to/blk0002.dat
227 2012-11-17 13:02:25 <Ferroh> awesome, thanks sipa!
228 2012-11-17 13:02:34 <sipa> also, in 0.8 it will almost certainly be network/cpu bound, and not disk bound
229 2012-11-17 13:02:54 <Ferroh> I know.
230 2012-11-17 13:02:57 <Ferroh> http://www.reddit.com/r/Bitcoin/comments/13bsv4/this_is_frustrating_this_might_be_the_10th_time/
231 2012-11-17 15:13:34 <BitDev> hi all, can some one help me? In bitcoin protocol there are "Variable length integer" it can be 1, 3, 5 and 9 bytes length... how can i know which length it must be?
232 2012-11-17 15:15:24 <etotheipi_> BitDev: if the int is between 0 and 252, it's 1 byte (just the number itself), if it's 253-65535, the first byte is 253 and then the next two bytes are the number
233 2012-11-17 15:15:29 <t7> prices up again today :|
234 2012-11-17 15:15:44 <t7> im gonna sell
235 2012-11-17 15:18:07 <etotheipi_> BitDev, the same goes for 4-byte ints (start with 254), and 8-byte ints (start with 255)
236 2012-11-17 15:18:42 <BitDev> ow, thnx i get it )
237 2012-11-17 15:19:10 <BitDev> strange way, but not bad )
238 2012-11-17 15:19:24 <BitDev> thnx again :)
239 2012-11-17 15:48:51 <Ferroh> t7: no they aren't. The price has been about the same for like 12 hours.
240 2012-11-17 15:49:01 <Ferroh> https://ferroh.com/charts/2day_small?0
241 2012-11-17 15:49:16 <Ferroh> !ticker
242 2012-11-17 15:49:17 <gribble> BTCUSD ticker | Best bid: 11.61001, Best ask: 11.64990, Bid-ask spread: 0.03989, Last trade: 11.61001, 24 hour volume: 26364.78328069, 24 hour low: 11.42000, 24 hour high: 11.80000, 24 hour vwap: 11.66077
243 2012-11-17 15:53:15 <sipa> omg up 0.38 USD!
244 2012-11-17 15:55:04 <D34TH> quick, BUY
245 2012-11-17 16:15:30 <JDuke128> !ticker
246 2012-11-17 16:15:31 <gribble> BTCUSD ticker | Best bid: 11.64963, Best ask: 11.64970, Bid-ask spread: 0.00007, Last trade: 11.64963, 24 hour volume: 25076.23740745, 24 hour low: 11.42000, 24 hour high: 11.80000, 24 hour vwap: 11.65622
247 2012-11-17 17:29:56 <jgarzik> ok
248 2012-11-17 17:30:08 <jgarzik> time to finish implementing script ops in picocoin
249 2012-11-17 17:30:29 <jgarzik> last night got a good start.  I think total coding time for EvalScript()-in-C will probably be under 4 hours for initial implementation.
250 2012-11-17 17:54:21 <jgarzik> hrm
251 2012-11-17 17:54:55 <jgarzik> testnet3 chain might be missing OP_2ROT test?
252 2012-11-17 17:55:06 <jgarzik> I don't see an implementation in pynode
253 2012-11-17 18:20:40 <TD> good evening
254 2012-11-17 18:21:23 <amiller> good afternoon
255 2012-11-17 18:36:10 <daybyter> Hi!
256 2012-11-17 18:38:07 <daybyter> Anyone using the mtgox API and has also problems with the return value being sometimes an object and sometimes an array?
257 2012-11-17 20:49:51 <jgarzik> why in the world was OP_2MUL and OP_2DIV disabled?
258 2012-11-17 20:51:52 <edcba> to learn ppl how to mul using add ?
259 2012-11-17 20:58:41 <cjd> meh why use openssl bignum for everything, I have yet to see a place in bitcoin which really requires an MPI
260 2012-11-17 21:02:23 <cjd> https://github.com/jgarzik/picocoin/blob/master/lib/hexcode.c#L12
261 2012-11-17 21:02:41 <cjd> static const unsigned char hexdigit_val[256] = {   <-- [256] with only 16 entries?
262 2012-11-17 21:03:24 <jgarzik> cjd: RE bignum, because otherwise a bignum implementation must be manually coded
263 2012-11-17 21:04:06 <jgarzik> cjd: RE hexdigit, keeps code nice and compact.
264 2012-11-17 21:05:29 <cjd> sure but why  hexdigit_val[256] if only the first 16 values are populated?
265 2012-11-17 21:07:45 <cjd> oh not 16, actually 22 because of the capitals
266 2012-11-17 21:15:35 <sipa> "meh"
267 2012-11-17 21:18:03 <cjd> base58 is one of the more annoying ones because there's really no way to calculate it without using bignums but in general I think bignum adds complexity disproportionate to it's benefits
268 2012-11-17 21:21:12 <jgarzik> cjd: indeed... it's not the first 16 values
269 2012-11-17 21:21:29 <jgarzik> cjd: thus [256]
270 2012-11-17 21:21:51 <cjd> hrm maybe I have to read that more closely
271 2012-11-17 21:22:23 <jgarzik> cjd: picocoin uses [] array initializers, to initialize values in the middle of the array
272 2012-11-17 21:22:28 <jgarzik> cjd: the rest are zeroed by the compiler
273 2012-11-17 21:22:37 <cjd> ahh ic
274 2012-11-17 21:22:39 <cjd> that makes sense
275 2012-11-17 21:23:04 <cjd> sorry for noise
276 2012-11-17 21:23:19 <jgarzik> cjd: RE bignum...  from picocoin's perspective, it would add even more complexity to picocoin, to manually reimplement BIGNUM ;p
277 2012-11-17 21:23:30 <jgarzik> if you link with libcrypto, you get bignum anyway
278 2012-11-17 21:23:40 <jgarzik> pointless to avoid, if using sha1/sha256/ripemd160/...
279 2012-11-17 21:24:09 <jgarzik> so yeah, I'd remove BIGNUM usage, if we were not already using sha*, ripemd, and ECDSA from libcrypto
280 2012-11-17 21:24:25 <jgarzik> ECDSA requires BIGNUM internally
281 2012-11-17 21:24:32 <cjd> sort of
282 2012-11-17 21:24:50 <cjd> I was just looking at CBitcoin's base58 implementation and it's kind of sad
283 2012-11-17 21:25:11 <cjd> unless you really want to keep base58 inside of 1 portable .h file, openssl is easier
284 2012-11-17 21:25:16 <jgarzik> cjd: CBitcoin is kind of sad, in general
285 2012-11-17 21:25:43 <jgarzik> cjd: It's the authors "I'm learning C, networking, bitcoin and ADTs" project
286 2012-11-17 21:26:13 <cjd> well.. don't buck "I'm learning C" projects, my project is one of them
287 2012-11-17 21:27:21 <cjd> If I can extract things that I need from it then I'm happy.. and if he does the base58 stuff manually then it is a great candidate for a Base58.h file
288 2012-11-17 21:30:22 <jgarzik> cjd: true
289 2012-11-17 21:30:44 <jgarzik> cjd: He just rubs me the wrong way, because his forum thread, literally, claims CBitcoin to be the 'future of bitcoin'
290 2012-11-17 21:30:58 <jgarzik> and he's taking money from gullible people, towards that end
291 2012-11-17 21:31:05 <cjd> yeah, /me doesn't have a forum account :)
292 2012-11-17 21:49:39 <jgarzik> w00t
293 2012-11-17 21:49:49 <jgarzik> picocoin script eval: done!
294 2012-11-17 21:50:04 <cjd> \\o/
295 2012-11-17 21:50:05 <jgarzik> time to start adding script tests
296 2012-11-17 22:08:44 <yellowhat> if somebody is interested in a talk "bitcoin for java developers" with some interesting listener questions and slightly covering bitcoinJ: http://www.youtube.com/watch?v=aiQWRVrsW_Q
297 2012-11-17 23:22:00 <xenland> eh BitcoinJ is just too cluster mucked to learn or use from
298 2012-11-17 23:44:54 <KingSolamin> wtb bitcoins pm me if interested in selling
299 2012-11-17 23:46:22 <Luke-Jr> KingSolamin: wrong channel
300 2012-11-17 23:46:23 <coyo> good evening. i have a question. is there a more elaborate p2p network (possibly based on udp or more elaborate techniques) than using a hardcoded list of initial bitcoin seed nodes, using tcp (which isnt really a great transport layer protocol for this purpose), or using an unqualified broadcast architecture
301 2012-11-17 23:48:40 <cjd> coyo: no matter how smart your p2p code is, it needs to know *someone* so they can introduce it to all of the others
302 2012-11-17 23:49:28 <cjd> and bitcoin is weird, some of it's messages would fit in UDP packets just fine but others are as much as hundreds of kilobytes
303 2012-11-17 23:56:07 <coyo> cjd, that's good to know.
304 2012-11-17 23:56:43 <cjd> also that original node should be trusted
305 2012-11-17 23:57:07 <cjd> because if they are evil they can keep you seperate from the rest of the network