1 2011-07-25 00:17:34 <hippich> howdy, if i want to start completely separate bitcoin-based network, what should I update in original client? I found: the whole genesis block, data dir, address version, message header version, port.
  2 2011-07-25 00:17:50 <hippich> remove few inplace checks for valid block chain.
  3 2011-07-25 00:22:55 <doublec> hippich: this might help https://github.com/sacarlson/MultiCoin/blob/master/create_new_genisis_block.txt
  4 2011-07-25 00:24:00 <hippich> doublec, thanx!
  5 2011-07-25 00:24:07 <doublec> np
  6 2011-07-25 00:24:09 <sacarlson> hippich: you need to run MultiCoin to create new chains
  7 2011-07-25 00:25:06 <sacarlson> hippich: also checkout #multicoin on freenode
  8 2011-07-25 00:27:30 <hippich> sacarlson, thanx. i wanted to do experiment (it will involve too much of things to change in original code), but wanted to make sure it will not give any problem to existing network.
  9 2011-07-25 00:28:27 <sacarlson> hippich: MultiCoin works on all networks with just config changes but your free to modify it to make more changes if needed
 10 2011-07-25 00:29:02 <hippich> yeah. i want to change a way reward is applied, coins represented, etc %)
 11 2011-07-25 00:29:06 <sacarlson> hippich: it made for expermentation and advancement
 12 2011-07-25 00:29:59 <sacarlson> hippich: I"m not sure what you want but the standard config has verible inflation settings
 13 2011-07-25 00:30:28 <hippich> yeah, the whole thing helps a lot!
 14 2011-07-25 00:30:39 <hippich> thanx for it!
 15 2011-07-25 01:02:50 <coderrr> see this yet? http://news.ycombinator.com/item?id=2800790
 16 2011-07-25 01:04:56 <jrmithdobbs> someone put a huge BITCOIN IS NOT ANONYMOUS flashing banner on bitcoin.org already
 17 2011-07-25 01:05:27 <senseles> people still wouldnt see it
 18 2011-07-25 01:05:51 <jrmithdobbs> so this guy wasted a lot of time proving something that was never claimed
 19 2011-07-25 01:05:53 <jrmithdobbs> good job i guess
 20 2011-07-25 01:07:35 <coderrr> but theres lots of pretty pictures so it must be important
 21 2011-07-25 01:08:00 <Akiron> so in Bitcoin, the blockchain is stored as a linked list of CBlockIndex objects, correct?
 22 2011-07-25 01:08:40 <jrmithdobbs> now
 23 2011-07-25 01:08:50 <jrmithdobbs> if he published the code he used to generate those pretty pictures
 24 2011-07-25 01:08:55 <jrmithdobbs> *that* would be worth talking about ;p
 25 2011-07-25 01:09:06 <jrmithdobbs> but i'm sure it's just shit that calls blockexplorer.com ;p
 26 2011-07-25 01:21:28 <luke-jr> jrmithdobbs: no, I think proving Bitcoin isn't anonymous is a good thing
 27 2011-07-25 01:21:35 <luke-jr> even when it was never claimed
 28 2011-07-25 01:21:36 <jrmithdobbs> but we already knew it wasn't
 29 2011-07-25 01:21:51 <jrmithdobbs> he didn't prove anything non-obvious to anyone familiar
 30 2011-07-25 01:21:57 <jrmithdobbs> with the actual design/impl
 31 2011-07-25 01:21:58 <luke-jr> jgarzik went out of his way to make it pretty public that it wasn't anonymous, and STILL people didn't believe him
 32 2011-07-25 01:22:24 <jrmithdobbs> luke-jr: and you think those people are the kinds that read academic-style papers?
 33 2011-07-25 01:22:27 <jrmithdobbs> heh
 34 2011-07-25 01:22:44 <luke-jr> :p
 35 2011-07-25 01:23:25 <coderrr> yea wikileaks said it was anonymous and ppl trust wikileaks way more than any of u crazy btc devs :P
 36 2011-07-25 01:23:37 <luke-jr> wtf is mtgox doing
 37 2011-07-25 01:24:51 <iddo> someone is buying...
 38 2011-07-25 01:25:02 <moa7> btc has turned into a huge honey-pot trap ...
 39 2011-07-25 01:27:25 <jrmithdobbs> luke-jr: ?
 40 2011-07-25 01:27:44 <luke-jr> jrmithdobbs: MtGox price is skyrocketting
 41 2011-07-25 01:27:45 <pierce> nothing is anonymous, the whole debate is just annoying
 42 2011-07-25 01:28:15 <jrmithdobbs> tor could be traced if anyone put enough effort into it
 43 2011-07-25 01:28:21 <jrmithdobbs> for instance
 44 2011-07-25 01:28:38 <luke-jr> yep
 45 2011-07-25 01:28:58 <jrmithdobbs> and tor is MUCH more "anonymous" than btc ever could be
 46 2011-07-25 01:29:23 <pierce> ya, but tor is more anonymous than letting your ISP see all your traffic, and bitcoin is more anonymous than paying for things with a credit card etc
 47 2011-07-25 01:29:43 <jrmithdobbs> actually
 48 2011-07-25 01:29:52 <jrmithdobbs> i'd say bitcoin is less anonymous than prepaid cards
 49 2011-07-25 01:30:36 <pierce> depends on how you buy them
 50 2011-07-25 01:30:44 <jrmithdobbs> right
 51 2011-07-25 01:30:47 <pierce> same with any monetary exchange
 52 2011-07-25 01:31:36 <pierce> IMO bitcoins can be used in safer and easier to control ways than anything that touches USD
 53 2011-07-25 01:32:06 <jrmithdobbs> depends on context but in general that's probably true
 54 2011-07-25 01:32:13 <jrmithdobbs> just because there's no governing body involved
 55 2011-07-25 01:33:20 <pierce> to some people, anonymous means it doesn't require state ID and a signature.  The whole debate really boils down to definitions, not technical correctness.
 56 2011-07-25 01:35:52 <moa7> there is a legal proof standard also ....
 57 2011-07-25 01:36:40 <pierce> has anyone thought of switching from sha256 to something more memory intensive like bcrypt to make specialized mining hardware more expensive and more on par with current desktop systems?
 58 2011-07-25 01:37:08 <moa7> plausible deniability is sufficient for lots of purposes
 59 2011-07-25 01:38:35 <coderrr> pierce, i think bcrypt is just computationally expensive, not memory intensive
 60 2011-07-25 01:38:43 <coderrr> scrypt is both
 61 2011-07-25 01:39:05 <pierce> coderrr: okay then, your idea :-)
 62 2011-07-25 01:39:58 <pierce> just getting annoyed about huge mining monopolies buying the silence of anyone who makes any advancements
 63 2011-07-25 01:40:31 <AlonzoTG> ugh.
 64 2011-07-25 01:40:57 <AlonzoTG> the data structure for "wallet_entry" is not described in the protocol docs yet it is part of the protocol.
 65 2011-07-25 01:42:44 <whomp> are bitcoin donations to non-profits tax deductible?
 66 2011-07-25 01:43:17 <luke-jr> I don't see why not
 67 2011-07-25 01:43:21 <phantomcircuit> i cant see why not
 68 2011-07-25 01:43:21 <pierce> whomp: you can try
 69 2011-07-25 01:43:23 <luke-jr> but IANAL
 70 2011-07-25 01:44:58 <pierce> whomp: tax law isn't written in C, english has the magical feature of being very context sensitive, so whoever has the most expensive lawyer can deduce whatever they want from the law, but no, that particular case has never been tested to my knowledge.
 71 2011-07-25 01:45:29 <phantomcircuit> lol
 72 2011-07-25 01:45:39 <phantomcircuit> implying that c is clear and absolute
 73 2011-07-25 01:46:01 <pierce> meh, just trying to avoid getting kickbanned if I had said java :-)
 74 2011-07-25 01:46:01 <[Tycho]> ?? is perfect
 75 2011-07-25 03:08:04 <cuddlefish> aha
 76 2011-07-25 03:08:13 <cuddlefish> to people getting "DATABASE ERROR: Run recovery"
 77 2011-07-25 03:08:17 <cuddlefish> I found the 'recovery'!
 78 2011-07-25 03:08:18 <cuddlefish> http://pybsddb.sourceforge.net/utility/db_recover.html
 79 2011-07-25 05:09:09 <gjs278> when poclbm crashes a core of my gpu, is there anyway to keep the system up besides rebooting
 80 2011-07-25 05:09:22 <gjs278> I can tell one of the 5970 cores stopped hashing
 81 2011-07-25 05:09:37 <gjs278> and if I send poclbm I'll probably have a full crash occur
 82 2011-07-25 05:09:41 <gjs278> end*
 83 2011-07-25 05:13:05 <gjs278> I need a tool that can like down the core gracefully so I don't get screen freezes every 15 seconds due to 1/3 cores being offline
 84 2011-07-25 05:20:55 <Joric> how may i update blockchain on a gae application if it doesnt support sockets? is there any http compatible source to download blocks using urlfetch?
 85 2011-07-25 05:21:04 <cuddlefish> Can't locate object method "USE" via package "THIS" (perhaps you forgot to load "THIS"?) at - line 1.
 86 2011-07-25 05:21:07 <cuddlefish> gaha
 87 2011-07-25 05:21:11 <cuddlefish> Joric: Nope.
 88 2011-07-25 05:21:29 <cuddlefish> Joric: can you look stuff up in blockexplorer?
 89 2011-07-25 05:21:36 <Joric> hell, i even could upload them from my own desktop i just need some http wrapper for that
 90 2011-07-25 05:22:46 <Joric> maybe blockexplorer, really
 91 2011-07-25 05:22:56 <Joric> for new blocks
 92 2011-07-25 05:23:20 <Joric> poor blockexplorer
 93 2011-07-25 05:23:43 <Joric> http://blockexplorer.com/rawblock/00000000000004b2db9d01a39b84752cb05e2692897d66f6632acb15e41fbaa5
 94 2011-07-25 05:24:11 <Joric> is it the whole block or it missing something
 95 2011-07-25 05:25:43 <Joric> i've just seen some analythics want to try something similar
 96 2011-07-25 05:25:44 <Joric> http://anonymity-in-bitcoin.blogspot.com/2011/07/bitcoin-is-not-anonymous.html
 97 2011-07-25 05:29:00 <Joric> not sure how realtime it could be, depends of the proper indexing
 98 2011-07-25 05:30:21 <cuddlefish> Joric: it's the binary data
 99 2011-07-25 05:42:02 <Soak> nobody for help me here? http://forum.bitcoin.org/index.php?topic=31368 (How to estimate next difficulty?)
100 2011-07-25 05:42:56 <cuddlefish> Soak: ;;bc,nextdiff
101 2011-07-25 05:43:02 <cuddlefish> ;;bc,nextdiff
102 2011-07-25 05:43:03 <gribble> Error: "bc,nextdiff" is not a valid command.
103 2011-07-25 05:43:08 <cuddlefish> ;;bc,statc
104 2011-07-25 05:43:09 <cuddlefish> ;;bc,stats
105 2011-07-25 05:43:09 <gribble> Error: "bc,statc" is not a valid command.
106 2011-07-25 05:43:11 <gribble> Current Blocks: 137924 | Current Difficulty: 1690906.2047244 | Next Difficulty At Block: 139103 | Next Difficulty In: 1179 blocks | Next Difficulty In About: 1 week, 0 days, 18 hours, 1 minute, and 12 seconds | Next Difficulty Estimate: 1787546.53911810
107 2011-07-25 05:43:20 <cuddlefish> there we go
108 2011-07-25 05:45:01 <Soak> yes but I want for my website
109 2011-07-25 05:45:13 <Soak> so I can't do that by IRC :x
110 2011-07-25 05:47:56 <Joric> Soak, http://blockexplorer.com/q
111 2011-07-25 05:48:10 <Joric> particulary, http://blockexplorer.com/q/estimate
112 2011-07-25 05:48:28 <Joric> bot uses this, i believe
113 2011-07-25 05:48:46 <Soak> thank you :D
114 2011-07-25 05:49:11 <Joric> poor, poor blockexplorer
115 2011-07-25 05:50:23 <Joric> though there's at least one official mirror (powered by mtgox live) http://109.123.116.245
116 2011-07-25 05:50:39 <Joric> doesn't work :]
117 2011-07-25 05:53:47 <Joric> doesn't look like it uses dns load balancing either
118 2011-07-25 05:57:47 <Joric> trying to host my own blockexplorer :) node.js compiles forever
119 2011-07-25 06:03:56 <kinlo> is blockexplorer opensource?
120 2011-07-25 06:04:22 <kinlo> Soak: *when* difficutly changes, is easy, but for the rest I'd reccomend you just read the source code
121 2011-07-25 06:05:46 <Joric> kinlo, nope, i'm trying to use https://github.com/bitcoinjs/node-bitcoin-explorer
122 2011-07-25 06:06:41 <kinlo> mmmz
123 2011-07-25 06:11:44 <doublec> Joric: there's also https://github.com/jtobey/bitcoin-abe
124 2011-07-25 07:06:39 <prof7bit> yesterday I have found this thread: http://forum.bitcoin.org/index.php?topic=29416.0 this is the first time I have seen someone trying to really document some of the complicated parts. If I had some bitcoins I would donate to this guy for his efforts.
125 2011-07-25 07:35:13 <Joric> whoa http://bitcoinjs.org/
126 2011-07-25 07:38:38 <Joric> "fork me on github" :) didn't know forking become so popular
127 2011-07-25 07:39:04 <ersi> should rename github to forkhub
128 2011-07-25 07:39:30 <random_cat> for me, you ugly, crumbling bits
129 2011-07-25 07:39:40 <random_cat> s/for/fork/
130 2011-07-25 08:16:32 <coderrr> gmaxwell, you still interested in CLI support for 'send from address' ?
131 2011-07-25 08:30:11 <gmaxwell> coderrr: Yes
132 2011-07-25 08:34:22 <coderrr> gmaxwell, I'm thinking this:             "sendtoaddress <bitcoinaddress>[:<sendfromaddress1>[;<sendfromaddress2>[;...]]] <amount> [comment] [comment-to]", does that sound alright?
133 2011-07-25 08:36:22 <gmaxwell> That sounds reasonable.
134 2011-07-25 08:36:54 <Joric> why not commas
135 2011-07-25 08:37:39 <coderrr> Joric, I already use semicolons to separate addresses in the gui
136 2011-07-25 08:37:47 <coderrr> so otehr than that, no reason
137 2011-07-25 08:42:55 <coderrr> hrm yea, semicolons on cmdline is kinda bad though, if anyone actually used this im sure someone will screw up and not escape one of them
138 2011-07-25 08:43:22 <TD> you know
139 2011-07-25 08:43:45 <TD> i can't think of any reasons to have complex logic in scriptSig
140 2011-07-25 08:43:53 <Eliel> yes, semicolon is bad idea for command line :)
141 2011-07-25 08:43:57 <TD> it could have just been called "initialStack"
142 2011-07-25 08:44:54 <Joric> https://en.bitcoin.it/wiki/URI_Scheme
143 2011-07-25 08:45:04 <Joric> Each segment can contain parameters separated from it using a semicolon (";")
144 2011-07-25 08:45:55 <Joric> hehe not exactly that
145 2011-07-25 08:46:34 <Joric> http://bitcoinstats.com/irc/logs/2011/01/28/7
146 2011-07-25 08:46:46 <Joric> "jgarzik: avoid comma-separated list" :]
147 2011-07-25 08:47:51 <coderrr> oh well doesnt matter this will never be in the official client anyway
148 2011-07-25 08:49:41 <cjdelisle> Just as a straw pole, would people be in to the idea of a telnet controlled daemon where you could use any gui you wanted?
149 2011-07-25 08:50:31 <da2ce7> coderrr, did you check out my proposal: https://en.bitcoin.it/wiki/NewSendCoins
150 2011-07-25 08:51:28 <coderrr> da2ce7, u get any feedback on it ?
151 2011-07-25 08:51:43 <da2ce7> not really... but I have since come up with a better solution...
152 2011-07-25 08:51:49 <da2ce7> it is really quite elegant.
153 2011-07-25 08:52:00 <Joric> i'd prefer an embedded http server :]
154 2011-07-25 08:52:42 <Joric> damn bitcoin-p2p won't install
155 2011-07-25 08:52:43 <coderrr> does boost have a construct for finally blocks ?
156 2011-07-25 08:52:56 <Joric> says No such file or directory: '/atlas/www/node-bitcoin-p2p/build-cc'
157 2011-07-25 09:13:58 <TD> http://anonymity-in-bitcoin.blogspot.com/2011/07/bitcoin-is-not-anonymous.html
158 2011-07-25 09:14:02 <TD> very cool analysis
159 2011-07-25 09:16:24 <Joric> TD, did you try to write a bitcoin client on python?
160 2011-07-25 09:16:29 <TD> no, why?
161 2011-07-25 09:16:50 <da2ce7> codemojo, Joric, http://code.bulix.org/k1i0ul-80318
162 2011-07-25 09:16:58 <da2ce7> what do you think... good concept?
163 2011-07-25 09:16:58 <Joric> there's another google guy, https://github.com/samrushing/caesure
164 2011-07-25 09:17:44 <Joric> transaction std's? :)
165 2011-07-25 09:18:19 <Joric> da2ce7, what's that?
166 2011-07-25 09:18:36 <TD> what makes you think sam rushing is a googler?
167 2011-07-25 09:18:46 <da2ce7> it is a for bitcoind, it is a proposed API for making transactions.
168 2011-07-25 09:18:57 <da2ce7> we can add to it later, so we can add inputs.
169 2011-07-25 09:19:04 <Joric> TD, http://www.nightmare.com/~rushing/
170 2011-07-25 09:19:27 <Joric> though google 2000, then cisco
171 2011-07-25 09:19:46 <TD> he worked there wants 11 years ago :)
172 2011-07-25 09:19:49 <TD> once
173 2011-07-25 09:20:06 <TD> i don't think python is a good language for a bitcoin implementation, personally
174 2011-07-25 09:20:11 <TD> but hey, if you're just doing it for fun, whatever
175 2011-07-25 09:23:39 <bonsaikitten> TD: python can be quite nice, if you know how to use it ... most of the criticism comes from people that want to use it like C++ and then notice that it doesn't work well
176 2011-07-25 09:27:14 <Joric> on the other hand, java can't be nice at all
177 2011-07-25 09:27:28 <phedny> what's wrong with Java?
178 2011-07-25 09:28:07 <phantomcircuit> phedny, oracle will sue you
179 2011-07-25 09:28:17 <phedny> good point
180 2011-07-25 09:29:27 <TD> bonsaikitten: oh it's a nice language. the problems are mostly efficiency and correctness
181 2011-07-25 09:29:33 <Zagitta> TD: that analasys is kinda BS
182 2011-07-25 09:30:18 <phantomcircuit> lol @ efficiency
183 2011-07-25 09:30:39 <bonsaikitten> TD: correctness? how is that more a problem than in other languages?
184 2011-07-25 09:31:11 <bonsaikitten> and I haven't noticed any efficiency problems yet ... just write lazy code that does the minimal amount of work needed
185 2011-07-25 09:31:30 <da2ce7> updated version of make new tx, what do you think? http://code.bulix.org/u01vhw-80319
186 2011-07-25 09:32:34 <da2ce7> ah, there is a small error in it.
187 2011-07-25 09:33:03 <phantomcircuit> bonsaikitten, python code doing identical stuff is usually about 4x slower than java which is about 2x slower than c
188 2011-07-25 09:33:20 <phantomcircuit> however it is *much* easier to use the proper ds in python
189 2011-07-25 09:33:22 <phantomcircuit> so
190 2011-07-25 09:33:24 <phantomcircuit> whatevah
191 2011-07-25 09:34:23 <Joric> da2ce7, why camelcase? also "Transction"
192 2011-07-25 09:34:32 <bonsaikitten> phantomcircuit: hmm. from exposure to average java code I must say that *good* java can be reasonably fast, but on average it's about as fast as a glacier
193 2011-07-25 09:34:32 <phedny> Java speed depends a lot on programming style .. as with other languages of course .. but it performance is critical, an ugly programming style can be as efficient as C
194 2011-07-25 09:34:54 <phedny> espacially object creation and desctruction is expensive
195 2011-07-25 09:34:59 <phantomcircuit> bonsaikitten, yes which is why i said code that does identical things
196 2011-07-25 09:35:02 <bonsaikitten> and sadly most people who attempt to write C manage to make it slower than proper py because they do too much work
197 2011-07-25 09:35:06 <phantomcircuit> same algorithm structure
198 2011-07-25 09:35:14 <bonsaikitten> phantomcircuit: that's not going to work out
199 2011-07-25 09:35:39 <bonsaikitten> x = [ 2**y for y in range(1024)] <-- try doing that in Java
200 2011-07-25 09:35:46 <phantomcircuit> bonsaikitten, write a for loop using xrange and print "hello world" it will be a shitton slower
201 2011-07-25 09:35:52 <phantomcircuit> facepalm
202 2011-07-25 09:35:59 <phantomcircuit> bonsaikitten, you're just echoing what i said
203 2011-07-25 09:36:13 <bonsaikitten> well, here's the funny thing. you can't easily do it in Java
204 2011-07-25 09:36:19 <phantomcircuit> bonsaikitten, you're just echoing what i said
205 2011-07-25 09:36:30 <bonsaikitten> phantomcircuit: you seem to be stuck in a temporal loop
206 2011-07-25 09:36:31 <phantomcircuit> I HEAR AN ECHOO ECHOechoechech
207 2011-07-25 09:36:42 <bonsaikitten> next time you come around I'll smack you in the face to kick you out of the loop
208 2011-07-25 09:37:18 <phantomcircuit> bonsaikitten, wow really threatening me? nice
209 2011-07-25 09:37:21 <bonsaikitten> same algorithm can still have different runtime complexity, so while it *looks* the same the runtime ends up O(n^2) vs. O(n^6) or something funny like that
210 2011-07-25 09:37:49 <bonsaikitten> and then people say foo is slow, because ... in foo this *is* slow if you do it this way, instead of the smart way
211 2011-07-25 09:37:55 <phantomcircuit> yeah im obviously talking about algorithms with identical runtime complexity taking longer in python than java or c
212 2011-07-25 09:37:59 <bonsaikitten> phantomcircuit: no, just trying to stop looping
213 2011-07-25 09:38:05 <phantomcircuit> the interpretter for python is fucktarded slow
214 2011-07-25 09:38:19 <bonsaikitten> unladen-swallow ? :)
215 2011-07-25 09:38:39 <phantomcircuit> lol
216 2011-07-25 09:38:45 <Joric> i'm using pure python ecdsa currently, works fine
217 2011-07-25 09:38:49 <bonsaikitten> at least py is portable ... java claims to be, but I had enough fun trying to make things work
218 2011-07-25 09:38:49 <phantomcircuit> looked abandoned last i checked
219 2011-07-25 09:38:53 <bonsaikitten> yep
220 2011-07-25 09:39:07 <bonsaikitten> but that shows that it's mostly CPython being very simple-minded
221 2011-07-25 09:39:26 <phantomcircuit> and they were using horribly gcc hacks to make some things faster
222 2011-07-25 09:39:33 <bonsaikitten> PyPy might be fun
223 2011-07-25 09:39:45 <phantomcircuit> pypy is nice and is actually faster in most instances
224 2011-07-25 09:39:51 <bonsaikitten> gcc? I thought it was using the llvm infrastructure?
225 2011-07-25 09:40:32 <phantomcircuit> yes it does use the llvm architecture
226 2011-07-25 09:40:39 <phantomcircuit> *and* a metric ton of hackery
227 2011-07-25 09:40:56 <phantomcircuit> to make opcode lookup O(1) they used references to labels
228 2011-07-25 09:41:06 <phantomcircuit> which only just barely makes sense
229 2011-07-25 09:41:55 <bonsaikitten> Eliel: slight barrier of entry there, but if you can make it work that's ok
230 2011-07-25 09:42:08 <Eliel> yes, the barrier of entry was... huge
231 2011-07-25 09:42:17 <Eliel> but IMO, totally worth it
232 2011-07-25 09:42:21 <bonsaikitten> phantomcircuit: I guess you find struct padding offensive too, even if it can make ~30% performance difference
233 2011-07-25 09:42:30 <molecular> does anyone use his mtgox yubikey for ssh auth?
234 2011-07-25 09:42:38 <molecular> oh, wrong channel, sry
235 2011-07-25 09:43:19 <phantomcircuit> bonsaikitten, opcode lookup even in worst case was searching tree which fit entirely in a single cache line
236 2011-07-25 09:43:42 <phantomcircuit> they could have used an O(n^2) algorithm and it would have made almost no difference
237 2011-07-25 09:44:00 <da2ce7> Joric, updated 3rd time: https://en.bitcoin.it/wiki/Da2ce7:New_Transaction
238 2011-07-25 09:44:01 <bonsaikitten> hmm. must have had a good reason for that hackery then
239 2011-07-25 09:44:30 <phantomcircuit> yes i spoke with the guy who did it
240 2011-07-25 09:44:39 <phantomcircuit> he said basically "it cant make it slower!"
241 2011-07-25 09:44:56 <bonsaikitten> eew.
242 2011-07-25 09:45:00 <phantomcircuit> lol
243 2011-07-25 09:45:09 <phantomcircuit> profiling? FUCK THAT
244 2011-07-25 09:45:35 <ersi> FUCK IT
245 2011-07-25 09:45:38 <ersi> We'll do it LIVE!
246 2011-07-25 09:46:03 <phantomcircuit> not to mention switch case on x86 platforms is already O(1) for sequential enums
247 2011-07-25 09:47:31 <TD> ah, the joy of language flamewars
248 2011-07-25 09:48:18 <TuxBlackEdo> papa bear will never get old
249 2011-07-25 09:49:24 <phantomcircuit> bonsaikitten, CPython is a shitty implementation, but it's the only universal implementation
250 2011-07-25 09:49:25 <TuxBlackEdo> FUCKING THING SUCKS
251 2011-07-25 09:49:54 <Zeiris> What's better than CPython?
252 2011-07-25 09:50:11 <phantomcircuit> Zeiris, pypy stackless python
253 2011-07-25 09:50:15 <phantomcircuit> shit IPython
254 2011-07-25 09:50:46 <phantomcircuit> wait fuck what's it called
255 2011-07-25 09:51:08 <phantomcircuit> jython is faster
256 2011-07-25 09:51:18 <phantomcircuit> but is hilariously unstable in comparison
257 2011-07-25 09:55:05 <TuxBlackEdo> http://www.youtube.com/watch?v=n3Bdlmi60Cw
258 2011-07-25 09:56:11 <diki> hey guys, do you know that Windows 7 loads faster than bitcoin?
259 2011-07-25 09:56:13 <diki> just a fact
260 2011-07-25 09:56:56 <GMP> im positively surprised with the amount of effort spent on making python fast, Psyco for example, its just crazy!
261 2011-07-25 09:57:19 <diki> oh wait, my 600mhz phone boots up faster
262 2011-07-25 09:57:23 <diki> bitcoin is a nice OS
263 2011-07-25 09:57:31 <cjdelisle> tehehe
264 2011-07-25 09:57:33 <diki> too much stuff to load in memory
265 2011-07-25 09:57:45 <diki> must be vert far ahead than win7
266 2011-07-25 09:57:48 <diki> *very
267 2011-07-25 09:58:09 <cjdelisle> Are you referring to block chain download?
268 2011-07-25 09:58:21 <cjdelisle> or just plain startup?
269 2011-07-25 09:58:21 <diki> i am referring to the slow startup
270 2011-07-25 09:58:25 <cjdelisle> ouch
271 2011-07-25 09:58:51 <TuxBlackEdo> you probably need to uninstall your fake anti virus and malware
272 2011-07-25 09:59:20 <cjdelisle> What I see in bitcoin is a combination of really brilliant code and really horrendus code.
273 2011-07-25 09:59:58 <cjdelisle> Unfortunately I am no help since I'm basicly sunk with all the nuances of C++/boost
274 2011-07-25 10:00:04 <da2ce7> cjdelisle, both by satoshi?
275 2011-07-25 10:00:54 <cjdelisle> Assuming satoshi was not a team, I would have to suggest that he was opperating under extreme time pressure.
276 2011-07-25 10:01:31 <da2ce7> so not a bad programer, but jsut a super rushed programer.
277 2011-07-25 10:02:32 <cjdelisle> There are things in btc that leave me in awe. There are also things which I would never commit myself (but then that's a problem with me, I value correctness over ever actually getting something done)
278 2011-07-25 10:03:45 <cjdelisle> I know C and I can guess at most of it. When you see a hardcoded ip address, that's not hard to know what that means...
279 2011-07-25 10:06:08 <da2ce7> lol
280 2011-07-25 10:06:10 <da2ce7> ya
281 2011-07-25 10:08:42 <UukGoblin> cjdelisle, looks to me like he focused on important stuff, and left the details for later generations
282 2011-07-25 10:10:37 <cjdelisle> Indeed. I am a little bit disheartened at the apparent lack of support some of the, now very wealthy, early adopters are putting in to improving the code.
283 2011-07-25 10:11:47 <cjdelisle> Perhaps I am all wet here but it feels to me like some of this stuff should be first class, well documented well tested code by now.
284 2011-07-25 10:16:07 <prof7bit> what is the purpose of OP_VERIFY? Isn't it enough that true or false is left on the stack after the script exited?
285 2011-07-25 10:20:01 <UukGoblin> cjdelisle, well, I for one, am not wealthy enough yet
286 2011-07-25 10:20:10 <UukGoblin> I sold about 20k bitcoins while they were below $1
287 2011-07-25 10:20:15 <UukGoblin> way below
288 2011-07-25 10:20:44 <UukGoblin> but! if the bitcoin gets up to $100 I promise to start contributing some good stuff
289 2011-07-25 10:21:02 <UukGoblin> (after I build a house that is)
290 2011-07-25 10:21:39 <cjdelisle> Yea. It's just kind of a problem, nobody like me really is incentivised to donate code because people will profit and people is not me.
291 2011-07-25 10:22:13 <cjdelisle> Not to be greedy about it but it's going to hold things back.
292 2011-07-25 10:22:16 <UukGoblin> well, not really... if the client keeps breaking and discouraging non-tecchy people, bitcoin value will drop and I'll lose money...
293 2011-07-25 10:22:49 <UukGoblin> but I just don't have enough time at the moment, having a full time job, distance learning university and a rig to maintain...
294 2011-07-25 10:22:56 <TuxBlackEdo> shit my skin is peeling now, i had some bad sunburn like 3 days ago
295 2011-07-25 10:23:10 <cjdelisle> /nod
296 2011-07-25 10:23:57 <cjdelisle> I don't think there's much risk of it losing value *unless* there is a competitor.
297 2011-07-25 10:24:22 <UukGoblin> namecoin is a competitor
298 2011-07-25 10:24:38 <TuxBlackEdo> not in 3 months
299 2011-07-25 10:24:51 <Eliel> cjdelisle: I've noticed the tendency for people who have bitcoins to donate to people who work on things they like.
300 2011-07-25 10:25:03 <UukGoblin> TuxBlackEdo, what's in 3 months?
301 2011-07-25 10:25:33 <cjdelisle> Yea, not a very strong one IMO. It has nothing that BTC doesn't have except a dns tld that nobody can access.
302 2011-07-25 10:25:35 <TuxBlackEdo> UukHoblin: block 24,000 is the proposed block to switch to merged mining and base namecoin off of proof of work from the bitcoin blockchain
303 2011-07-25 10:26:01 <UukGoblin> TuxBlackEdo, well, it'll still be a competing currency after that
304 2011-07-25 10:26:09 <TuxBlackEdo> http://dot-bit.org/MergedMining
305 2011-07-25 10:26:19 <TuxBlackEdo> not a competing blockchain though
306 2011-07-25 10:26:34 <cjdelisle> yes a competing blockchain
307 2011-07-25 10:26:39 <TuxBlackEdo> its like saying mtgox usd will be a competing currency
308 2011-07-25 10:26:51 <UukGoblin> well the way I see it, namecoin can do all that bitcoin can, and more (i.e. the domain stuff)
309 2011-07-25 10:26:58 <Zagitta> cjdelisle: indeed, they could probably found a whole dev team
310 2011-07-25 10:26:59 <UukGoblin> so why shouldn't we just start using namecoin instead
311 2011-07-25 10:27:23 <cjdelisle> Because btc was here first?
312 2011-07-25 10:27:28 <TuxBlackEdo> UukGoblin: no idea
313 2011-07-25 10:27:29 <UukGoblin> heh
314 2011-07-25 10:28:22 <cjdelisle> IDK maybe namecoin will become valuable but first it will need some demand for it's .bit tld, I don't see them having that ATM.
315 2011-07-25 10:28:29 <TD> bitcoin has momentum
316 2011-07-25 10:28:31 <TD> namecoin doesn't
317 2011-07-25 10:28:34 <TuxBlackEdo> after block 24,000 the only way to mine for namecoin is to mine for bitcoin and namecoin together
318 2011-07-25 10:28:39 <cjdelisle> maybe a dns tld for tor or i2p? idk
319 2011-07-25 10:28:45 <TD> TuxBlackEdo: no, it can still be mined independently
320 2011-07-25 10:29:02 <cjdelisle> yea, btc has "momentum"
321 2011-07-25 10:29:13 <TuxBlackEdo> apparently the proof of work to get namecoins will be based off the bitcoin blockchain
322 2011-07-25 10:29:24 <TD> i understand how it works
323 2011-07-25 10:29:28 <TD> i "wrote the book" on the algorithm
324 2011-07-25 10:29:35 <TuxBlackEdo> never mind then lol
325 2011-07-25 10:29:42 <TD> you can mine on namecoin post-change without mining on bitcoin too
326 2011-07-25 10:29:42 <UukGoblin> well in 3 months, namecoin will get loads of momentum, because it'll suddenly halve the difficulty for all miners
327 2011-07-25 10:29:48 <TD> most people will probably choose both though. why not ?
328 2011-07-25 10:29:58 <TD> it's not so hard to keep up with the network
329 2011-07-25 10:30:06 <UukGoblin> yeah
330 2011-07-25 10:30:24 <TuxBlackEdo> yeah once merged mining starts it would be stupid to just mine for namecoin or just bitcoin
331 2011-07-25 10:30:28 <cjdelisle> meh momentum is more about getting on the news...
332 2011-07-25 10:30:40 <TuxBlackEdo> like silk road
333 2011-07-25 10:30:47 <UukGoblin> cjdelisle, news follow momentum
334 2011-07-25 10:30:52 <TuxBlackEdo> isnt that what got the price up to $30usd/btc anyhow
335 2011-07-25 10:30:54 <UukGoblin> and create more at the same time
336 2011-07-25 10:30:56 <cjdelisle> yea, silkroad is a really interesting case study.
337 2011-07-25 10:31:31 <cjdelisle> I don't know who started it but I doubt they really made any money off it, perhaps they were an early adopter and they did, idk.
338 2011-07-25 10:31:35 <TuxBlackEdo> we need to start an anonymous tor site that sells escrots in the major cities, even if it is a fake one it might generate enough news interest for us to cash out :)
339 2011-07-25 10:31:52 <TuxBlackEdo> just kidding... or am i?
340 2011-07-25 10:32:35 <bonsaikitten> TuxBlackEdo: limited market
341 2011-07-25 10:32:39 <UukGoblin> TuxBlackEdo, and then start a gay one just for political correctness and to get even more news
342 2011-07-25 10:33:05 <cjdelisle> AFAICT the btc community mostly gave silkroad nothing but scorn for "giving btc a bad name" but on the other hand if it wern't for that "bad name", btc would have no name and it would still be <1$.
343 2011-07-25 10:33:09 <TuxBlackEdo> UukHoblin: you are a genius
344 2011-07-25 10:33:20 <bonsaikitten> I mean ... if I were interested I'd just look at any of the local aggregator sites and ask for a bill including VAT ... living in a sane country is fun
345 2011-07-25 10:33:28 <TuxBlackEdo> i agree cjdelisle
346 2011-07-25 10:34:37 <cjdelisle> So if I'm an inventor and I can invent something which will drive btc to 100$, people are going to make huge $$ off of my invention but I will make 0.
347 2011-07-25 10:34:56 <cjdelisle> This is going to drive the next big thing away.
348 2011-07-25 10:35:29 <cjdelisle> And I don't really see an easy way around it unless only early adopters invent things which use btc.
349 2011-07-25 10:37:36 <TuxBlackEdo> early adopters probably sold a lot of btc thats what drove the price down to $13
350 2011-07-25 10:39:02 <TuxBlackEdo> if everyone just bought and never sold, the price would never have a down tick
351 2011-07-25 10:41:57 <TuxBlackEdo> i mean all the new interest in bitcoin drove the price down, there was no reason for the price of bitcoind to go down from $30 to $13
352 2011-07-25 10:46:29 <TD> cjdelisle: bitcoin would have hit the press without silk road sooner or later anyway
353 2011-07-25 10:46:45 <TD> SR is definitely a net negative for the project, unfortunately
354 2011-07-25 10:47:13 <TuxBlackEdo> it hasnt really hit the press yet
355 2011-07-25 10:47:47 <Joric> check out this video http://money.cnn.com/video/technology/2011/07/18/t_bitcoin_currency.cnnmoney/
356 2011-07-25 10:48:05 <TuxBlackEdo> meze grill video?
357 2011-07-25 10:48:28 <Joric> yes
358 2011-07-25 10:49:33 <TuxBlackEdo> that guy bruce wagner runs the bitcoin show
359 2011-07-25 10:49:58 <Joric> bruce wagner just ate 7.035 bitcoins :)
360 2011-07-25 10:51:23 <Joric> thats quite a lunch
361 2011-07-25 11:05:09 <TuxBlackEdo> oh shit
362 2011-07-25 11:05:13 <TuxBlackEdo> bitcoin in the news
363 2011-07-25 11:05:23 <TuxBlackEdo> bitcoin got slashdotted because its insecure
364 2011-07-25 11:06:06 <pierce> damn hackers
365 2011-07-25 11:06:20 <b4epoche> link
366 2011-07-25 11:06:26 <TuxBlackEdo> slashdot.org
367 2011-07-25 11:06:27 <TuxBlackEdo> top one
368 2011-07-25 11:06:56 <b4epoche> not sure that's such bad news
369 2011-07-25 11:07:23 <BlueMatt> thats good news
370 2011-07-25 11:07:25 <TuxBlackEdo> yeah its great news
371 2011-07-25 11:07:30 <BlueMatt> points out what we've all known
372 2011-07-25 11:07:41 <BlueMatt> people just keep saying no, no thats wrong
373 2011-07-25 11:07:46 <TuxBlackEdo> yeah awesome i want stories like this every day
374 2011-07-25 11:07:48 <gmaxwell> Yea, I wish it pointed it out with stronger evidence.
375 2011-07-25 11:08:00 <gmaxwell> E.g. I wish they'd actually done the work to identify someone.
376 2011-07-25 11:09:03 <pierce> My favorite part of that story is when they still don't know who the thief is, just a couple other things that they did.
377 2011-07-25 11:09:29 <da2ce7> gmaxwell, make transaction bitcoind syntax proposal: https://en.bitcoin.it/wiki/Da2ce7:New_Transaction
378 2011-07-25 11:09:36 <Zagitta> gmaxwell: am i missunderstanding something here? because as far as i can see the only thing they prove is that the services people use with bitcoin makes them tracable NOT bitcoin itself
379 2011-07-25 11:09:41 <da2ce7> do you like the general idea?
380 2011-07-25 11:10:38 <pierce> Zagitta: ya, and they associate people posting on bitcoin forums with public donation addresses in their signatures etc
381 2011-07-25 11:11:39 <Zagitta> pierce: indeed so wth is the articles point?
382 2011-07-25 11:11:55 <pierce> grant money?  maybe they are selling something?
383 2011-07-25 11:12:04 <gmaxwell> Zagitta: bitcoin itself links togeather a lot of activities.
384 2011-07-25 11:12:18 <gmaxwell> Bitcoin itself also potentially attaches them to you, but they didn't look into that.
385 2011-07-25 11:12:49 <gmaxwell> (e.g. someone with a lot of monitoring nodes might be able to guess your IP with high confidence if you send frequently.
386 2011-07-25 11:12:53 <gmaxwell> )
387 2011-07-25 11:13:08 <Zagitta> if you don't use different addresses yeah
388 2011-07-25 11:13:19 <gmaxwell> No, even if you use different addresses.
389 2011-07-25 11:13:34 <gmaxwell> Because bitcoin itself frequently links addresses, which is part of what they were writing about.
390 2011-07-25 11:14:29 <gmaxwell> da2ce7: there should be export/import/sign operations.
391 2011-07-25 11:14:37 <Zagitta> well that was kinda hard to decifer from the article
392 2011-07-25 11:15:30 <b4epoche> "We contract all vertices whose corresponding public-keys belong to the same user. "  How?
393 2011-07-25 11:16:05 <gmaxwell> da2ce7: e.g. I should be able to prepare a txn on a node with the blockchain but no private keys.
394 2011-07-25 11:16:29 <gmaxwell> da2ce7: then export it, take it to an offline node with no block chain, sign it, export it, import it, and send
395 2011-07-25 11:16:42 <da2ce7> gmaxwell, ah ok.  but you want bitcoin to 'autofill' when there is missing info also.
396 2011-07-25 11:16:46 <gmaxwell> b4epoche: observing them used as common inputs.
397 2011-07-25 11:16:47 <b4epoche> or are they just saying they contract vertices for public-keys they know belong to the same user?
398 2011-07-25 11:17:01 <_W_> as long as you keep separate wallets (so you're not fooled by the client into mixing anonymized and non-anonymized coins), and make sure to anonymize before you spend on anything that can be tied to you, bitcoin is anonymous
399 2011-07-25 11:17:12 <gmaxwell> b4epoche: yes.
400 2011-07-25 11:18:13 <_W_> would be kind of neat if the client could track various "taint" lables on coins
401 2011-07-25 11:18:23 <gmaxwell> _W_: there is a patch for this!
402 2011-07-25 11:18:35 <_W_> excellent!
403 2011-07-25 11:18:55 <da2ce7> so I guess if we use states, we can use the 'building' state, where we can make the trancstion... when we change it to a 'ready' state when the client checkd that it has all the private keys to preform the operation.
404 2011-07-25 11:22:53 <gmaxwell> The other application of being able to export a signed but incomplete transaction is so you can pass it off for someone else to announce it for improved anonymity.
405 2011-07-25 11:23:31 <da2ce7> :)
406 2011-07-25 11:23:34 <da2ce7> yep adding them now :)
407 2011-07-25 11:23:56 <gmaxwell> s/incomplete/unannounced/
408 2011-07-25 11:24:21 <da2ce7> so we have 'build' where the transaction is being built... 'ready' when the client checks if it has all the private keys + enougth coins. 'done' when it is signed' and 'submitted' when the note publishes it.
409 2011-07-25 11:24:33 <da2ce7> *node
410 2011-07-25 11:25:04 <da2ce7> should be able to inport/export the transaction at any stage
411 2011-07-25 11:25:17 <gmaxwell> right, build, then export, move to ready node. Sign. move the done to a connected node,and submit it.
412 2011-07-25 11:29:38 <gmaxwell> da2ce7: there should probably be some special cases for addoutput
413 2011-07-25 11:29:50 <gmaxwell> e.g. addoutput keypool rest
414 2011-07-25 11:29:59 <gmaxwell> (to take change)
415 2011-07-25 11:30:24 <gmaxwell> also perhaps 'addoutput fee' even though it's not really an output.
416 2011-07-25 11:31:47 <da2ce7> ah... we can have an overide addoutput (pubkey) will be the output for any leftover btc
417 2011-07-25 11:32:21 <gmaxwell> right and the ability to have it autofill the pubkey from the pool.
418 2011-07-25 11:32:40 <Zagitta> random thought: wouldn't it be possible to build something like TOR with bitcoin transactions? i mean where you can send coins to a p2p network where the coins take a random route spreading out etc and then arrive at a specified address(es)?
419 2011-07-25 11:32:41 <da2ce7> yep, just as we dont specify ammounts on the inputs.
420 2011-07-25 11:32:58 <da2ce7> Zagitta, connect to a hidden service
421 2011-07-25 11:33:12 <da2ce7> bitcoin works great over tor
422 2011-07-25 11:33:32 <da2ce7> also checkout the BtcFn project... bitcoin on freenet.
423 2011-07-25 11:34:42 <Zagitta> ah yeah i guess that could work too but this was a way to avoid being able to trace coins back to your public key, at least it would require a ton of work
424 2011-07-25 11:36:03 <gmaxwell> Zagitta: well, it would close of using the bitcoin network itself for the deanonymization, but people can still go after endpoints.
425 2011-07-25 11:51:50 <UukGoblin> hrm, someone should hack gource to work on bitcoin addresses instead of vcs commits
426 2011-07-25 12:01:49 <Joric> http://minus.com/lcnfxe afrikaans block
427 2011-07-25 12:02:56 <UukGoblin> lol
428 2011-07-25 12:03:38 <Joric> url http://blockexplorer.com/rawblock/000000000000027737bfd1370f3f28b63baf06e54ca646eeb595255cb27087e7
429 2011-07-25 12:04:22 <TuxBlackEdo> wow i just bought some 100% pure aloe vera on my sunburned and peeling back, and as soon as i put it on i started feeling intense pain and started screaming, i think i woke the whole house up just now :-/
430 2011-07-25 12:04:52 <UukGoblin> Joric, the translation sucks though
431 2011-07-25 12:05:39 <UukGoblin> TuxBlackEdo, wuss
432 2011-07-25 12:05:50 <TuxBlackEdo> i know :(
433 2011-07-25 12:06:29 <TuxBlackEdo> http://blockexplorer.com/block/000000000000027737bfd1370f3f28b63baf06e54ca646eeb595255cb27087e7
434 2011-07-25 12:06:32 <TuxBlackEdo> this one?
435 2011-07-25 12:06:35 <TuxBlackEdo> what is with this block?
436 2011-07-25 12:07:12 <UukGoblin> probably has an afrikaans word somewhere in the hashes
437 2011-07-25 12:07:45 <Joric> baaababa :)
438 2011-07-25 12:08:11 <theorbtwo> Hm.  A transaction with the same source address multiple times?  Odd.
439 2011-07-25 12:09:32 <UukGoblin> theorbtwo, nah, not really... multiple inputs from the same source
440 2011-07-25 12:10:02 <theorbtwo> UukGoblin: Er, why not just add them directly, and only list that source one time?
441 2011-07-25 12:10:26 <UukGoblin> theorbtwo, because input != source
442 2011-07-25 12:10:34 <cjdelisle> multiple people payed that address and now that address can't just pay once, it has to nad over all of the inputs seperately.
443 2011-07-25 12:10:42 <cjdelisle> *hand
444 2011-07-25 12:11:04 <UukGoblin> input != source address I should say. Yeah, cjdelisle described it better
445 2011-07-25 12:29:21 <Eliel> TuxBlackEdo: there's an interesting transaction in that block... the TxIn part has the same address several times with different amounts. Is that how it works when the same address has received several transactions into it?
446 2011-07-25 12:31:15 <Eliel> ... oh nevermind, I should've read the conversation before asking :)
447 2011-07-25 12:35:39 <da2ce7> gmaxwell, updated: https://en.bitcoin.it/wiki/Da2ce7:New_Transaction
448 2011-07-25 12:36:45 <gmaxwell> da2ce7: whats the purpose of the name if you always reference by number after that?
449 2011-07-25 12:36:53 <da2ce7> for accounting
450 2011-07-25 12:37:06 <da2ce7> like "to paying gmaxwell"
451 2011-07-25 12:37:18 <da2ce7> or "move to safe wallet"
452 2011-07-25 12:38:43 <gmaxwell> da2ce7: Perhaps there should be another option for input where it you specify by value not identiy, so you can crete txn where you don't care about the source?
453 2011-07-25 12:39:49 <da2ce7> but you never gain inputs by value... you gain them by source.
454 2011-07-25 12:40:37 <gmaxwell> da2ce7: right, but say I want to pay you 1 btc, I don't care where it comes from.
455 2011-07-25 12:40:42 <gmaxwell> addinput txn 1.0
456 2011-07-25 12:40:49 <gmaxwell> addoutput you 1.0
457 2011-07-25 12:40:59 <gmaxwell> addoutput
458 2011-07-25 12:41:19 <da2ce7> no you would go addoutput me 1.0  then bitcoin will atomaticaly find a input that suits. in the 'reddytransaction {txnumber}' stage
459 2011-07-25 12:41:29 <gmaxwell> ahh. okay!
460 2011-07-25 12:41:32 <gmaxwell> That works too.
461 2011-07-25 12:42:01 <da2ce7> it will also automaticly add a change output if nessccary also
462 2011-07-25 12:42:34 <gmaxwell> yep. Thats good in that it should reduce people footgunning themselves by forgeting change and turning eveything into fees.
463 2011-07-25 12:42:52 <da2ce7> lool 4000 btc fee
464 2011-07-25 12:43:50 <da2ce7> it should be very lazy... where you only need to spcify one output, and the reddytransaction will fill-out everything else for you.
465 2011-07-25 12:44:54 <jusan> hi *
466 2011-07-25 12:45:06 <da2ce7> gday jusan
467 2011-07-25 12:45:18 <gmaxwell> da2ce7: yup.
468 2011-07-25 12:46:32 <UukGoblin> da2ce7, what does this newtransaction do?
469 2011-07-25 12:46:41 <UukGoblin> and where's 4k btc fee?
470 2011-07-25 12:47:00 <da2ce7> lets you have as little or as much controll on making a tx as you like
471 2011-07-25 12:47:03 <gmaxwell> da2ce7: shownewtransaction  might show you the sums of inputs and outputs too? (in the same way that it shows you the fee)
472 2011-07-25 12:47:24 <da2ce7> hmm yeah... good idea.
473 2011-07-25 12:50:14 <da2ce7> updated: https://en.bitcoin.it/wiki/Da2ce7:New_Transaction
474 2011-07-25 12:52:34 <gmaxwell> "Starbucks"kkkkkkkk~.
475 2011-07-25 12:52:46 <gmaxwell> hah sorry.
476 2011-07-25 12:52:59 <ersi> kekeke~
477 2011-07-25 12:53:03 <gmaxwell> da2ce7: I suppose optional means "maximum" ?
478 2011-07-25 12:53:29 <gmaxwell> What does suggested mean? (on the add outputs for fees)
479 2011-07-25 12:54:12 <da2ce7> sugested is minimum.
480 2011-07-25 12:54:42 <da2ce7> optional is  "maximum", if needed.
481 2011-07-25 12:54:49 <da2ce7> forced, is always that ammount.
482 2011-07-25 12:54:54 <gmaxwell> Makes sense to me.
483 2011-07-25 12:56:53 <gmaxwell> da2ce7: newtransaction {name} {txid}   Make a new txn, copying a prior one.
484 2011-07-25 12:57:15 <da2ce7> ah yes. :)
485 2011-07-25 12:57:55 <da2ce7> and newtransaction {txnumber} {name} for a dupicate
486 2011-07-25 12:57:55 <gmaxwell> Will sign run reddytransaction if it hasn't been done yet?
487 2011-07-25 12:58:12 <da2ce7> no it will be just complain.
488 2011-07-25 12:58:42 <gmaxwell> and shouldn't that be "readytransaction"
489 2011-07-25 12:58:52 <da2ce7> we prob should make a 'completetransaction' that automaticaly ready, signs, and submits.
490 2011-07-25 12:59:10 <da2ce7> sign transaction only proceses transactions in the 'ready' stage.
491 2011-07-25 12:59:30 <gmaxwell> da2ce7: hm. I think most of the time (when I don't care about fee/change details) I'd want to just sign then submit.
492 2011-07-25 12:59:33 <da2ce7> it will throgh an error if you are tring to sign something that is in the 'build' stage
493 2011-07-25 12:59:48 <da2ce7> so I'll add completetransaction
494 2011-07-25 13:00:10 <da2ce7> that will do the, ready > sign > submit in-one
495 2011-07-25 13:00:13 <gmaxwell> But what when I want to sigh but not submit yet (e.g. because I'm going to export and move it?)
496 2011-07-25 13:00:40 <gmaxwell> e.g. in the case where you're using an offline secure wallet, it would be maintaining the keypool.
497 2011-07-25 13:00:48 <gmaxwell> So you'd want it to do the final addition of change outputs.
498 2011-07-25 13:00:57 <gmaxwell> I'm nitpicking here.
499 2011-07-25 13:01:00 <da2ce7> hmm... I suppose we can add an override to sign that accepts calle ready if needed.
500 2011-07-25 13:01:05 <da2ce7> *calls
501 2011-07-25 13:01:20 <da2ce7> but it certanly shouldn't be the default behaviour.
502 2011-07-25 13:01:27 <gmaxwell> yea.. a sign "force"
503 2011-07-25 13:01:52 <gmaxwell> right, normally you'd want to review before you sign, so I agree with the seperate ready step.
504 2011-07-25 13:03:40 <Eliel> if there's an offline keypool/signing system, there ought to be a protocol where it also includes a new public key to be monitored at times. Not a transaction but message to the network connected system that there's a new address to keep an eye out for.
505 2011-07-25 13:04:04 <Eliel> or maybe just an address, not the public key
506 2011-07-25 13:04:10 <gmaxwell> ;;later tell sipa that the walletimport/export patch ought to let you just import public keys without the private keys too.
507 2011-07-25 13:04:10 <gribble> The operation succeeded.
508 2011-07-25 13:04:39 <gmaxwell> You'll need the public key in order to prepare the transaction.
509 2011-07-25 13:04:56 <Eliel> ah, then that too
510 2011-07-25 13:05:21 <gmaxwell> ;;later tell sipa (or maybe support importing addresses without the public key too)
511 2011-07-25 13:05:21 <gribble> The operation succeeded.
512 2011-07-25 13:05:31 <gmaxwell> I guess the signing operation can still supply the public key.
513 2011-07-25 13:06:32 <da2ce7> https://en.bitcoin.it/wiki/Da2ce7:New_Transaction updated
514 2011-07-25 13:09:07 <gmaxwell> da2ce7: you should be able to give sign/submit/complete many txids, and perhaps even "all"
515 2011-07-25 13:09:32 <gmaxwell> the idea being you could prep a bunch of txn in a encrypted wallet.. the only step needing the wallet to be unlocked is sign.
516 2011-07-25 13:09:48 <da2ce7> gmaxwell, you should be using a script for that, I think
517 2011-07-25 13:10:09 <gmaxwell> Perhaps.
518 2011-07-25 13:10:17 <gmaxwell> Fair enough I guess.
519 2011-07-25 13:10:57 <da2ce7> it also add secuiry issues.
520 2011-07-25 13:11:14 <da2ce7> cause you could have a rouge tx that you have forgotten about.
521 2011-07-25 13:12:31 <gmaxwell> well a multi-add e.g. txid1,txid2 doesn't have that issue, but you're right wrt add.
522 2011-07-25 13:12:33 <jrmithdobbs> ya, i don't think there should be an "all"
523 2011-07-25 13:12:34 <darrenburns> anyone 10 btc fir 12ppusd each
524 2011-07-25 13:12:47 <sacarlson> gmaxwell: you doing offline transactions?
525 2011-07-25 13:13:09 <jrmithdobbs> but i think their should be able to give several txids to everything
526 2011-07-25 13:13:12 <gmaxwell> sacarlson: da2ce7 just wrote up an rpc/cli interface for this.
527 2011-07-25 13:13:29 <gmaxwell> jrmithdobbs: well, it can be version 2.
528 2011-07-25 13:13:29 <jrmithdobbs> s/their/you/
529 2011-07-25 13:13:49 <jrmithdobbs> fair enough
530 2011-07-25 13:13:52 <da2ce7> jrmithdobbs, what do you mean?
531 2011-07-25 13:13:54 <jrmithdobbs> that's some good stuff though
532 2011-07-25 13:14:18 <jrmithdobbs> da2ce7: i mean that other than the sign step you should be able to perform an operation on multiple txids at once like gmaxwell was saying
533 2011-07-25 13:14:30 <da2ce7> ah
534 2011-07-25 13:14:32 <[twisti]> i accidentally sent some money with an old client and no fee. will the money ever get moved ? can i cancel and resend it with a fee somehow ?
535 2011-07-25 13:14:50 <jrmithdobbs> da2ce7: but you have a valid point about forgotten txns and so i don't think there should be an option to manipulate "all pending"
536 2011-07-25 13:15:11 <jrmithdobbs> da2ce7: actually, so long as there's not an all target i think even multiple txnids in the sign step is fine
537 2011-07-25 13:15:32 <da2ce7> jrmithdobbs, fair point, I'll add an override for a list
538 2011-07-25 13:15:51 <da2ce7> [twisti], probbely, just may take a loooong time
539 2011-07-25 13:15:56 <da2ce7> depending on the size of th tx
540 2011-07-25 13:16:18 <da2ce7> if it is large... you may need to beg one of the big miners to let your tx in...
541 2011-07-25 13:17:41 <[twisti]> no idea how to tell
542 2011-07-25 13:18:12 <da2ce7> sure, it depends if your tx is being relayed or not
543 2011-07-25 13:18:29 <da2ce7> umm... load up bitcoin in -debug mode and doubble click on the tx
544 2011-07-25 13:18:46 <sacarlson> gmaxwell: da2ce7: I don't think I read back far enuf but I like this offline transaction method https://github.com/piotrnar/bitcoin/tree/importexporttx
545 2011-07-25 13:19:22 <gmaxwell> da2ce7: shownewtransaction should show you the size in bytes for readytxn
546 2011-07-25 13:19:33 <da2ce7> yep :)
547 2011-07-25 13:19:38 <da2ce7> that would be sooo usefull
548 2011-07-25 13:20:29 <gmaxwell> sacarlson: IIRC it doesn't let you seperate signing, so you can't have a sneakernet wallet without the blockchain for signing.
549 2011-07-25 13:20:34 <prof7bit> quick question: is it planned to implement a wallet export function that would dump the wallet contents into something like an easy to parse text file?
550 2011-07-25 13:20:45 <gmaxwell> prof7bit: bitcointools
551 2011-07-25 13:21:13 <gmaxwell> There is sipa's key import and export code too, which I think will go in for 0.4, but it just does keys.
552 2011-07-25 13:21:17 <sacarlson> gmaxwell: I think it's mean to transaction to yourself or trusted entity that won't double spend
553 2011-07-25 13:21:27 <prof7bit> is this part of the official distribution or would i have to convince the users to tust yet another thrid party
554 2011-07-25 13:21:42 <gmaxwell> prof7bit: bitcointools is written by gavin
555 2011-07-25 13:22:03 <sacarlson> gmaxwell: but for offline I guess there is no perfect solution anyway
556 2011-07-25 13:22:21 <gmaxwell> sacarlson: da2ce7's sketch looks pretty good.
557 2011-07-25 13:22:45 <sacarlson> gmaxwell: where's the link or code?
558 2011-07-25 13:22:51 <prof7bit> because I am planning to use a different wallet file format (not depend on berkeley db) and need an *easy* way for the users to transfer the wallet into my format
559 2011-07-25 13:22:54 <jrmithdobbs> gmaxwell: no it doesn't just do keys
560 2011-07-25 13:22:57 <gmaxwell> sacarlson: https://en.bitcoin.it/wiki/Da2ce7:New_Transaction
561 2011-07-25 13:23:02 <jrmithdobbs> gmaxwell: sipas code will dump the txn info too
562 2011-07-25 13:23:08 <jrmithdobbs> gmaxwell: just doesn't by default
563 2011-07-25 13:23:11 <gmaxwell> jrmithdobbs: oh, then he added that more recently
564 2011-07-25 13:23:22 <jrmithdobbs> it's been in it since i started testing it
565 2011-07-25 13:23:36 <gmaxwell> ah, then I must have missed it
566 2011-07-25 13:23:51 <jrmithdobbs> ya it's hidden away in the help output/code and not really obvious
567 2011-07-25 13:23:52 <jrmithdobbs> heh
568 2011-07-25 13:24:32 <gmaxwell> in any case, the key thing needed for offline wallets is making signing a seperate step... since the offline wallet can't be expected to know about inputs.
569 2011-07-25 13:25:31 <gmaxwell> Unless you're going to carry the blockchain to it by hand, which makes it not all that offline anymore and isn't viable e.g. for a inexpensive iron-wallet hardware device
570 2011-07-25 13:26:05 <sacarlson> gmaxwell: I'm not sure but that sounds like the escrow transaction combined with offline transaction, am I wrong?
571 2011-07-25 13:26:17 <prof7bit> is there a binary distribution of libdb for windows anywhere from some official place (just the needed DLL(s), not 45MB of other related stuff in a monster installer from oracle)?
572 2011-07-25 13:26:23 <gmaxwell> sacarlson: er? no.
573 2011-07-25 13:26:43 <sacarlson> gmaxwell: no I"m right?
574 2011-07-25 13:27:20 <gmaxwell> sacarlson: say I have a little bluetooth electronic device. Looks like a tamagotchi. It contains my private keys.
575 2011-07-25 13:27:34 <prof7bit> i couldn't find anything other than the oracle website
576 2011-07-25 13:27:40 <gmaxwell> My possibly compromised desktop knows my addresses.
577 2011-07-25 13:28:15 <prof7bit> and i can't believe the compiled berkeley db is 30 MB huge?
578 2011-07-25 13:28:52 <gmaxwell> On my desktop I prep a transaction. then send the txn to the tamagotchi. The tamagotchi shows me the transaction and asks my permission to sign it. Then it gets sent back to my desktop which publishes it.
579 2011-07-25 13:29:52 <sacarlson> gmaxwell: oh this is cool it's an offline transaction that can be edited before transmision
580 2011-07-25 13:30:44 <sacarlson> gmaxwell: not sure you couldn't just break things up into smaler transaction to do the same but hard to say
581 2011-07-25 13:31:45 <sacarlson> gmaxwell: escrows are similar they don't realy get submited until they get signed
582 2011-07-25 13:32:07 <da2ce7> jrmithdobbs, added lists for inport/export and processing
583 2011-07-25 13:32:09 <da2ce7> https://en.bitcoin.it/wiki/Da2ce7:New_Transaction
584 2011-07-25 13:32:41 <da2ce7> I've decided that it is ok to have submittransaction -a, that submits all pending 'done' transactions
585 2011-07-25 13:33:29 <gmaxwell> sacarlson: well, they're signed by multiple parties.
586 2011-07-25 13:33:40 <sacarlson> da2ce7: does this code exist or is it just a fantasy?
587 2011-07-25 13:33:49 <gmaxwell> da2ce7: thats good. an exaple usecase there is to prep a bunch of stuff, then switch to tor and submit.
588 2011-07-25 13:33:53 <da2ce7> sacarlson, just brainstorming
589 2011-07-25 13:34:01 <sacarlson> da2ce7: ok cool
590 2011-07-25 13:34:10 <TD> gmaxwell: the hard part there is, if your desktop is compromised, how do you know the address you see is owned by who you think it is
591 2011-07-25 13:34:18 <TD> bitcoin needs more infrastructure to solve this
592 2011-07-25 13:34:27 <TD> moving away from base58 addresses to more human-friendly identifiers
593 2011-07-25 13:34:52 <sacarlson> gmaxwell: you decide in an escrow or now called multisign transaction how many need to sign it,  it could be only one needed
594 2011-07-25 13:35:01 <gmaxwell> TD: solving that is generally hard. Even namecoin doesn't really solve it because lite clients can't easily exist right now.
595 2011-07-25 13:35:09 <TD> yeah
596 2011-07-25 13:35:20 <da2ce7> sacarlson, we could easly add an overide for a muilt sign system
597 2011-07-25 13:35:20 <TD> what we need to do is have secure hardware (or at least, more secure hardware) that knows how to challenge a remote server directly
598 2011-07-25 13:35:32 <TD> eg bitcoin:1Aa535.....?domain=amazon.com&value=20.00
599 2011-07-25 13:35:34 <gmaxwell> sacarlson: the patch as I reviewed it couldn't do that. (e.g. it couldn't export a txn with no sigs at all)
600 2011-07-25 13:35:47 <TD> the receiver would challenge amazon.com to sign a nonce.
601 2011-07-25 13:35:52 <TD> with that public key
602 2011-07-25 13:36:09 <sacarlson> gmaxwell: I said one not none
603 2011-07-25 13:36:12 <TD> if it passes, the tamagotchi can show "amazon.com" instead of an opaque address. you can have confidence you're really paying them
604 2011-07-25 13:36:14 <prof7bit> nobody knows whether there exists a binary only distribution of berkeley db for windows? (It should be around 1MB judging from the size of libdb for linux)
605 2011-07-25 13:36:20 <gmaxwell> sacarlson: ...
606 2011-07-25 13:36:49 <gmaxwell> sacarlson: if there is only one required then it can't be external to the device making the transaction.
607 2011-07-25 13:37:05 <sacarlson> gmaxwell: as example store money in an escrow to yourself with one signer
608 2011-07-25 13:37:06 <gmaxwell> sacarlson: e.g. there is no way to prep the txn on one device and sign it on another when there is only one sig.
609 2011-07-25 13:37:27 <gmaxwell> sacarlson: then that device has to have the blockchain
610 2011-07-25 13:37:48 <gmaxwell> What you're suggesting would be pretty easy to add, I think, and a good idea, but it doesn't do it yet.
611 2011-07-25 13:38:07 <sacarlson> gmaxwell: it already exists what I just said
612 2011-07-25 13:38:21 <gmaxwell> It does not. I just looked.
613 2011-07-25 13:38:42 <jrmithdobbs> prof7bit: check oracle.com (hahahhahaha)
614 2011-07-25 13:38:55 <sacarlson> gmaxwell: you can create two different addreses to yourself on the same walet
615 2011-07-25 13:39:05 <prof7bit> 30MB cant be real. libdb.so is 1MB on my system
616 2011-07-25 13:39:32 <da2ce7> why ain't we useing a super light sql?
617 2011-07-25 13:39:33 <jrmithdobbs> prof7bit: i'm sure it includes all the docs and shit
618 2011-07-25 13:39:42 <sacarlson> gmaxwell: I'll have to create more examples of what can be done with what we already have
619 2011-07-25 13:40:12 <gmaxwell> sacarlson: redeemmultisign will fail when you don't have any of the private keys on the transaction.
620 2011-07-25 13:40:41 <gmaxwell> In order to do this you'd need a redeemmultisign that works when you have none of the keys.
621 2011-07-25 13:40:50 <sacarlson> gmaxwell: I said you will need at least one private key
622 2011-07-25 13:41:13 <gmaxwell> Then it's not at all the same.
623 2011-07-25 13:41:44 <gmaxwell> It would be stupid to require that users convert all their inputs to multisign ones just to secure them.
624 2011-07-25 13:41:47 <prof7bit> It would be no problem for me to dynamically link (loadlibrary) if the user needs to open an original bitcoin wallet (and otherwise not use this dll at all) but I can' tell my users to install 30MB only to activate this additional functionality
625 2011-07-25 13:42:01 <sacarlson> gmaxwell: I can't ague this with you, I will later make you an example case to prove or disprove I'm right or wrong
626 2011-07-25 13:42:43 <prof7bit> and i'm not sure what oracle will do to me if i povide this dll alone for download.
627 2011-07-25 13:43:07 <gmaxwell> libdb is free software.
628 2011-07-25 13:43:09 <sacarlson> gmaxwell: but I have seen it said before to secure funds in an escrow to yourself even if it required two walets but don't think it's needed
629 2011-07-25 13:43:19 <talpan> can someone tell me what i am doing wrong ? $move = $bitcoin->move("1 '' 1"); I get error 500 from bitcoin. I've tripplecheked everything, the account balance is correct and other calls are working just fine
630 2011-07-25 13:43:34 <gmaxwell> sacarlson: I can't parse that.
631 2011-07-25 13:43:55 <gmaxwell> Yes, you can do it using two signatures and geting everyone who sends you funds to sign to both.
632 2011-07-25 13:44:15 <sacarlson> gmaxwell: even if it required two as you seem to think it does people can hold unlimited number of walets and private keys
633 2011-07-25 13:44:37 <gmaxwell> sacarlson: but then you have to have people who send you money do something special.
634 2011-07-25 13:45:23 <sacarlson> gmaxwell: it's just a mater of not needing a walet to hold money, just a private key
635 2011-07-25 13:45:31 <AlonzoTG> =P
636 2011-07-25 13:47:17 <iddo> gmaxwell: why did you say yesterday that we cannot have A sending coin to escrow iff B send coin, where they cannot destroy each other coin, isn't that what the default SIGHASH_ALL mean?
637 2011-07-25 13:49:03 <prof7bit> what version was it again that bitcoin uses? 4.6 or 4.7?
638 2011-07-25 13:49:24 <iddo> so why not A just create tx1 that has two inputs and needs two sigs for output, fill his input, sends tx1 to B, B fills his input, signs it and send to A, and A signs it?
639 2011-07-25 13:50:05 <AlonzoTG> I think the "clear and free memory" code is redundant because heap memory is always cleared by the operating system before being made available to the next process.
640 2011-07-25 13:50:39 <sacarlson> iddo: multi input escrow?  that's what we don't have yet.  that's what I'm waiting for
641 2011-07-25 13:50:55 <iddo> https://en.bitcoin.it/wiki/Contracts says that SIGHASH_ALL means  "I agree to put my money in, if everyone puts their money in and the outputs are this" ?
642 2011-07-25 13:51:05 <sacarlson> iddo: we already have multi out esrow but still only one input
643 2011-07-25 13:51:49 <iddo> sacarlson: i'm trying to understand if it cannot be done
644 2011-07-25 13:52:01 <gmaxwell> iddo: Think about it a bit.
645 2011-07-25 13:52:19 <sacarlson> iddo: yes I know it can be done just no code to do it yet
646 2011-07-25 13:52:26 <gmaxwell> iddo: if you have a TXN set to require A and B to release, then B can say "fuck you I won't release"
647 2011-07-25 13:52:49 <gmaxwell> iddo: whoever goes first on the release can get screwed.
648 2011-07-25 13:53:26 <iddo> gmaxwell: but doesnt sighash_all says that the coin get locked if only both of them signed TXN ?
649 2011-07-25 13:53:26 <talpan> can someone tell me what i am doing wrong ? $move = $bitcoin->move("1 '' 1"); I get error 500 from bitcoin. I've tripplechecked everything, the account balance is correct and other calls are working just fine
650 2011-07-25 13:53:44 <sacarlson> iddo: gmaxwell: I've seen other transactions already with multi input so It's just a mater of modifing grofer's code to allow mulit input as well
651 2011-07-25 13:54:00 <gmaxwell> iddo: you sign to _release_
652 2011-07-25 13:54:26 <gmaxwell> sacarlson: the inputs are held by seperate people.
653 2011-07-25 13:54:42 <gmaxwell> iddo: oh I see what you're saying.
654 2011-07-25 13:54:55 <iddo> why cannot we have TXN that says that the inputs are locked only if two signatures on this TXN are provided ?
655 2011-07-25 13:55:10 <TD> you can
656 2011-07-25 13:55:26 <gmaxwell> Yes, you do it by using inputs from both.
657 2011-07-25 13:56:26 <gmaxwell> E.g. I form a txn that says take one coin from me, one coin from you.  I sign it, pass to you... you sign it. If both don't sign it, it's not useful.
658 2011-07-25 13:56:30 <iddo> so this scenario is possible? sending coin to escrow iff B send coin, where they cannot destroy each other coin unless they also deswtroy their own coin, using the default SIGHASH_ALL flag
659 2011-07-25 13:56:41 <iddo> cool
660 2011-07-25 13:57:23 <gmaxwell> yea, I totally missed doing it on the _input_ side of the transaction.
661 2011-07-25 13:57:36 <TD> iddo: what do you want it for ?
662 2011-07-25 13:57:41 <gmaxwell> da2ce7: got a new feature for your list there.
663 2011-07-25 13:58:02 <sacarlson> td: I want it for multi in multi out escrow
664 2011-07-25 13:58:09 <da2ce7> :)
665 2011-07-25 13:58:28 <TD> can you illustrate that with a real world example?
666 2011-07-25 13:58:47 <gmaxwell> da2ce7: I should be able to send you a _partially_ signed txn e.g. sign should sign all it can, but if you don't have all the input keys it only signs the ones it can.
667 2011-07-25 13:58:55 <sacarlson> td: as in example poker game 10 players put into one escrow funds,  later unknown X player collect winings
668 2011-07-25 13:59:26 <gmaxwell> da2ce7: so e.g. I form a txn that spends 1 of my coins, 1 of your coins.  I prep it with all outputs. sign, export, send to you, you review and sign, then transmit.
669 2011-07-25 13:59:47 <da2ce7> ah ok... so we inport a tx, in the 'partialy ready state, that has some, but not all of the signutures.
670 2011-07-25 13:59:52 <gmaxwell> Yep!
671 2011-07-25 13:59:57 <TD> p2p poker is a hard problem. there's not much point only trying to solve a part of it. do you have a non poker example?
672 2011-07-25 13:59:59 <da2ce7> ok
673 2011-07-25 14:00:28 <iddo> can A create the other input saying "B sends one coin", or he must send it privately to B to do it? (not that it should matter much)
674 2011-07-25 14:00:39 <da2ce7> so will we just need to sign it... don't need to add any more tx's.
675 2011-07-25 14:00:47 <da2ce7> *um... inputs
676 2011-07-25 14:00:48 <sacarlson> td: it's just a good example case poker most people can understand
677 2011-07-25 14:01:11 <da2ce7> we keep on passing arround this 'ready' tx untill all the inputs are signed.
678 2011-07-25 14:01:17 <sacarlson> td: and yet one of the most complex
679 2011-07-25 14:01:28 <gmaxwell> E.g. joe blow wants to release his cure for cancer but wants 100 BTC.  We trust that joe blow isn't cheating.. but I don't want to send him 1 BTC towards that effort if perhaps he'll not raise 100 btc.
680 2011-07-25 14:01:46 <gmaxwell> So I start a txn paying him and add 1btc input, then I keep passing it around until it gets to 100 inputs.
681 2011-07-25 14:02:01 <da2ce7> ahh
682 2011-07-25 14:02:13 <da2ce7> yep... so we can 'sign' a building tx
683 2011-07-25 14:02:27 <gmaxwell> sadly there are complications, e.g. if the txn gets too big the outputs may need to change to add fee. :( but oh well. can't fix that.
684 2011-07-25 14:03:00 <da2ce7> gmaxwell, what about change?
685 2011-07-25 14:03:18 <gmaxwell> da2ce7: all the outputs need to be set when you start signing.
686 2011-07-25 14:03:43 <gmaxwell> as if you change the outputs then the signatures are blown up
687 2011-07-25 14:04:07 <da2ce7> ahh.. but changing an input dosn't hurt it.
688 2011-07-25 14:04:12 <iddo> i'm trying to understand the comment on https://en.bitcoin.it/wiki/Contracts about not being able to sign input scripts, can that be used by malicious parties to add input scripted to partically signed TXNs ?
689 2011-07-25 14:04:19 <gmaxwell> right, the inputs are stubbed off.
690 2011-07-25 14:04:35 <iddo> s/scripted/scripts
691 2011-07-25 14:04:56 <gmaxwell> iddo: yes, people can send inputs. It's hard to see how that can be malicious.
692 2011-07-25 14:05:01 <TD> sorry. i guess i missed an answer
693 2011-07-25 14:05:03 <TD> flaky internet here
694 2011-07-25 14:06:01 <da2ce7> is this using scripting or, the standard pratice?
695 2011-07-25 14:06:13 <iddo> gmaxwell: can you elaborate? SIGHASH_ALL behavior is " "I agree to put my money in, if everyone puts their money in and the outputs are this", so what adding unforseen input scripts do?
696 2011-07-25 14:06:19 <da2ce7> so the outputs are hashed, as a group, then signed.
697 2011-07-25 14:06:49 <sacarlson> td: last think I sent to you was td: it's just a good example case poker most people can understand td: and yet one of the most complex
698 2011-07-25 14:07:16 <TD> ok
699 2011-07-25 14:07:47 <gmaxwell> iddo: nothing interesting as far as I know.
700 2011-07-25 14:08:46 <iddo> i'm trying to undestand how the behavior is specified, if you sign without input scripts then can someone add input scripts while still having your signature valid ?
701 2011-07-25 14:08:46 <sacarlson> after poker is solved only thing to add would be time and ??
702 2011-07-25 14:09:00 <TD> they can do that with the SIGHASH_ANYONECANPAY flag
703 2011-07-25 14:09:06 <TD> by default it is not used
704 2011-07-25 14:09:15 <TD> the inputs are fixed except for the scripts
705 2011-07-25 14:09:19 <da2ce7> gmaxwell, so If I made a standard tx now, and sent it to a friend... that friend could change one of my inputs?
706 2011-07-25 14:10:00 <da2ce7> so that he pays for half of it?
707 2011-07-25 14:10:02 <talpan> can someone tell me what i am doing wrong ? $move = $bitcoin->move("1 '' 1"); I get error 500 from bitcoin. I've tripplechecked everything, the account balance is correct and other calls are working just fine
708 2011-07-25 14:10:06 <TD> the input scripts can be changed. not the other parts of the tx.
709 2011-07-25 14:10:20 <TD> talpan: i don't think anyone knows the answer. try looking in debug.log
710 2011-07-25 14:10:22 <da2ce7> ah ok
711 2011-07-25 14:10:30 <gmaxwell> da2ce7: yes. well, they could add a bunch of extra coin to it, which would become fee, because your inputs and outputs already matched.
712 2011-07-25 14:10:40 <iddo> in the example where A and B send coin to escrow only if both of them provide signature, are there any input scripts there?
713 2011-07-25 14:10:58 <TD> gmaxwell: i don't think that's possible
714 2011-07-25 14:11:00 <TD> the signatures would break
715 2011-07-25 14:11:52 <TD> iddo: for escrow see the Contracts page. it describes how such txns are structured
716 2011-07-25 14:12:19 <sacarlson> on a bad note my experment in merge mining failed,  but I have not given up yet
717 2011-07-25 14:12:22 <iddo> i'm looking there, sorry that i dont understand the bitcoin protocol well enough yet :(
718 2011-07-25 14:12:48 <aviadbd> Hey guys
719 2011-07-25 14:12:50 <TD> iddo: an escrow transaction involves requiring (in the simple case) 2-of-3 signers
720 2011-07-25 14:13:17 <iddo> TD: yes that example is simple, the output is released by 2 out of 3
721 2011-07-25 14:13:21 <aviadbd> Hmm
722 2011-07-25 14:13:23 <aviadbd> Lost my question.
723 2011-07-25 14:13:27 <aviadbd> :D
724 2011-07-25 14:14:15 <TD> iddo: i'm not sure what you want, sorry
725 2011-07-25 14:14:18 <iddo> TD: but i'm asking about the inputs behavior, so A cannot lock B's coin unless both of their coins get locked and signatures will need to be provided to release them
726 2011-07-25 14:14:36 <da2ce7> so we can sign a 'building' tx.
727 2011-07-25 14:14:39 <sacarlson> td: i'm not sure but I think you could have 2 with just one signer in escrow but I would have to verify
728 2011-07-25 14:14:53 <TD> yes you can have n-of-m
729 2011-07-25 14:14:59 <TD> where n <= m obviously
730 2011-07-25 14:15:34 <TD> iddo: yes, if you want two people to contribute to a pre-agreed set of outputs, you'd create a tx that spends outputs from two different people then have them sign it independently
731 2011-07-25 14:15:36 <sacarlson> td: that's what I was tring to tell them but it make no sence to most that it would be useful
732 2011-07-25 14:16:12 <TD> iddo: this is similar to the "assurance contract" section
733 2011-07-25 14:17:33 <iddo> TD: what does it mean that input scripted arent part of signature? can you insert new input script and still have the signature valid?
734 2011-07-25 14:17:38 <TD> yes
735 2011-07-25 14:17:40 <TD> not new inputs
736 2011-07-25 14:17:45 <TD> you can change the contents of the scripts
737 2011-07-25 14:17:48 <iddo> i'll try to read the assurance contract section more carefully
738 2011-07-25 14:17:58 <TD> in practice, scriptSigs only contain data buffers ... signatures and public keys
739 2011-07-25 14:18:47 <iddo> so how come malicious parties couldn't just insert some malicious input scripts that change the behavior of partially signed txn ?
740 2011-07-25 14:18:55 <TD> what would they change it to?
741 2011-07-25 14:19:17 <klikklak> so whats the best way to test if an fpga is actually doing something? I'm getting getworks, but no announcement of golden nonces
742 2011-07-25 14:20:01 <sacarlson> klikklak: try mine something like testnet see what you can get
743 2011-07-25 14:20:32 <sacarlson> klikklak: if that's too easy try namecoin
744 2011-07-25 14:20:35 <iddo> TD: in the example, A create his input to spend 1 coin, sends privately to B, then B creates his input to spend 1 coin, provides signature and sends back to A, now could A change the input script of B to do something malicious before he signs this txn ?
745 2011-07-25 14:20:46 <TD> like what?
746 2011-07-25 14:21:01 <klikklak> thats just bitcoind -testnet and the set it to localhost:xxx with rpcuser and rpcpassword ?
747 2011-07-25 14:21:03 <iddo> i'm not sure, what inputscripts can do here?
748 2011-07-25 14:21:03 <TD> yes he could change the script. however, i can't think of anything he'd change it to that could be malicious
749 2011-07-25 14:21:19 <TD> they can do the same as output scripts. however, in practice, the only useful thing they can do is push data buffers onto the stack
750 2011-07-25 14:21:39 <TD> there's no point putting any logic in scriptSigs as far as I can tell because the script is run with no input data, ie, you can always evaluate the program ahead of time
751 2011-07-25 14:21:54 <iddo> hmm
752 2011-07-25 14:22:14 <iddo> can you give some simple example of input script that is useful for something?
753 2011-07-25 14:22:35 <iddo> maybe i should read assurance contract
754 2011-07-25 14:22:36 <TD> like i said - the only things you can do with an input script are push signatures and keys. the protocol will let you put arbitrary programs in there
755 2011-07-25 14:22:45 <TD> for example, you can have a CHECKSIG that will verify a signature
756 2011-07-25 14:22:49 <TD> and that'd let you sign the input script
757 2011-07-25 14:23:01 <TD> but, there's no point in doing that
758 2011-07-25 14:23:05 <sacarlson> iddo: have you even looked at groffer's code yet? https://github.com/groffer/bitcoin  it's the only example I have seen in real code of escrow and multisign
759 2011-07-25 14:23:21 <TD> the input script has only one purpose - to set up the stack so the output scripts (scriptPubKey) runs correctly
760 2011-07-25 14:25:38 <TD> btw don't feel bad - that's a point that eluded me for a while too ;)
761 2011-07-25 14:25:39 <klikklak> sacarlson: it just exits the testnet according to debug.log
762 2011-07-25 14:25:53 <TD> i'd like to find a way of having the inputs be more useful
763 2011-07-25 14:25:54 <iddo> so CHECKSIG in input script is needed for the behavior that the coins get locked only if all users sign it?
764 2011-07-25 14:26:03 <TD> because i'd like to build transactions that pass arbitrary signed state to the output program
765 2011-07-25 14:26:08 <TD> i just don't see a way of doing it
766 2011-07-25 14:26:20 <TD> iddo: no. CHECKSIG in input scripts are useless
767 2011-07-25 14:26:26 <TD> nobody has proposed any protocols that involves one
768 2011-07-25 14:26:39 <TD> just think of input "scripts" as containers for bits of data the output script needs to run
769 2011-07-25 14:26:58 <TD> to tie up coins if all users sign, you do exactly what you said before - pass the tx around and collect signatures
770 2011-07-25 14:27:15 <TD> it's obviously the case that input scripts cannot be signed
771 2011-07-25 14:27:16 <sacarlson> klikklak: if you can't mine testnet maybe try weedsnet?
772 2011-07-25 14:27:21 <TD> because otherwise constructing the tx would be impossible
773 2011-07-25 14:27:28 <TD> adding the later signatures would invalidate the earlier ones
774 2011-07-25 14:27:39 <sacarlson> klikklak: does it mine at all?
775 2011-07-25 14:28:23 <sacarlson> klikklak: I"m not an expert in mining, I work on low power nets for a reason,  no miner power
776 2011-07-25 14:28:24 <klikklak> sacarlson: I get getworks from eligius, but the testnet gives errors when trying to get getworks
777 2011-07-25 14:28:58 <sacarlson> klikklak: you running a bitcoind in testnet to run your miner?
778 2011-07-25 14:29:14 <iddo> TD: so why is it so obvious that input script that provide data to output script canont be malicious? is it because the users who sign the TXN verified how the output behaves in this case, so they know that it doesnt involve any data that is controlled by an input script ?
779 2011-07-25 14:29:51 <TD> it could be 'malicious' in the sense of wasting cpu time. that's about it.
780 2011-07-25 14:29:58 <TD> what is the attack you have in mind?
781 2011-07-25 14:30:04 <klikklak> sacarlson: do I need to turn off regular bitcoin?  exact commands were: bitcoind -testnet, then quartus_stp -t mine.tcl
782 2011-07-25 14:30:30 <sacarlson> klikklak: no but you would have to run another bitcoind in testnet
783 2011-07-25 14:30:36 <iddo> i dont have attack in mind because i still didnt understand what input scripts do exactly...
784 2011-07-25 14:31:10 <TD> they /can/ do anything an output script can do. in practice what they /actually/ do is just push data buffers onto the stack
785 2011-07-25 14:31:15 <TD> input and output scripts are run independently
786 2011-07-25 14:31:25 <TD> the only state they share is the primary stack
787 2011-07-25 14:31:28 <sacarlson> klikklak: can you do bitcoind -rpcport 18332  getinfo
788 2011-07-25 14:31:31 <lfm> iddo input scripts supply a signature, the output script then confirms the sig is correct
789 2011-07-25 14:31:35 <TD> thus the output script runs with the stack as input
790 2011-07-25 14:31:47 <TD> input scripts run with nothing as input, thus, there's no point putting logic in them
791 2011-07-25 14:32:01 <TD> at least not with any currently proposed protocols. and i suspect there will never be a use for that
792 2011-07-25 14:32:17 <TD> so there's nothing malicious you can do. the input has to provide data to make the connected output script run
793 2011-07-25 14:32:18 <iddo> hmm
794 2011-07-25 14:32:32 <TD> if it doesn't do that, the tx is invalid. if it does, you can import the value. there are no other possibilities.
795 2011-07-25 14:32:38 <TD> the "script" language is very, very limited
796 2011-07-25 14:33:01 <iddo> ok
797 2011-07-25 14:34:35 <klikklak> sacarlson: thats what the initial bitcoind -testnet does, right?  I get could not connect to server on that port
798 2011-07-25 14:35:15 <lfm> klikklak: ya it takes several seconds for it to start to the point where it will accept connections
799 2011-07-25 14:35:37 <sacarlson> klikklak: well it must not be running or maybe not on that port?
800 2011-07-25 14:36:11 <klikklak> its running alright, I'll take look in debug.log
801 2011-07-25 14:36:15 <lfm> or maybe too soon after starting
802 2011-07-25 14:36:25 <sacarlson> klikklak: lfm: oh ya maybe it just hasn't downloaded all the blocks yet
803 2011-07-25 14:36:43 <lfm> what block number is it at?
804 2011-07-25 14:37:08 <klikklak> well, debug.log has the error for rpcuser/pw (which are set) and it ends with bitcoin exiting
805 2011-07-25 14:37:15 <klikklak> but the process seems to be there
806 2011-07-25 14:37:29 <lfm> ;;bc,blocks
807 2011-07-25 14:37:30 <gribble> 137988
808 2011-07-25 14:37:56 <lfm> klikklak: wrong user/pw?
809 2011-07-25 14:38:43 <sacarlson> klikklak: maybe you have user/pw reversed as I did,  took me 3 days to figure that out  >>>> me not very smart<<<
810 2011-07-25 14:41:29 <lfm> klikklak: could you run bitcoind getinfo ?
811 2011-07-25 14:43:56 <klikklak> lfm: I presume thats safe enough to paste on a pastebin?
812 2011-07-25 14:44:13 <lfm> just tell us what blocks and connections you have
813 2011-07-25 14:45:14 <klikklak> connections 1, and blocks 0
814 2011-07-25 14:45:29 <lfm> ok you havent even gotten started.
815 2011-07-25 14:45:36 <copumpkin> http://anonymity-in-bitcoin.blogspot.com/2011/07/bitcoin-is-not-anonymous.html
816 2011-07-25 14:45:41 <gribble> 137988
817 2011-07-25 14:45:41 <lfm> ;;bc,blocks
818 2011-07-25 14:46:12 <lfm> or if its testnet, is smaller but sure not zero
819 2011-07-25 14:47:42 <klikklak> lfm: hmm. I had namecoin working, lets see if its still intact