1 2011-10-09 00:05:44 <CIA-101> bitcoin: Con Kolivas * r27b05db4a504 cgminer/main.c: Use ADL activity report to tell us if a sick GPU is still busy suggesting it is hard hung and do not attempt to restart it.
  2 2011-10-09 00:25:43 <CIA-101> bitcoin: Con Kolivas * r81418d62c7e4 cgminer/main.c: Update displayed messages.
  3 2011-10-09 00:25:45 <CIA-101> bitcoin: Con Kolivas * rf9308f8fcab5 cgminer/main.c: Disable work division for CPU mining which can lead to repeated work being performed.
  4 2011-10-09 00:26:53 <shadders> can someone confirm this is current target in little endian?
  5 2011-10-09 00:26:54 <shadders> "target": "00000000000000000000000000000000000000000000005dee09000000000000"
  6 2011-10-09 00:27:50 <lianj> http://blockexplorer.com/q/hextarget
  7 2011-10-09 00:29:40 <shadders> that's in BE?
  8 2011-10-09 00:30:17 <shadders> I'm fairly sure it is but it's critical I don't fuck this up.
  9 2011-10-09 00:36:06 <upb> i'd guess what you pasted is big endian :)
 10 2011-10-09 00:36:41 <upb> otherwise it'd be like 0xff^6
 11 2011-10-09 00:36:59 <shadders> oh right duh... the target is meant to be low not high...
 12 2011-10-09 00:37:03 <shadders> thx
 13 2011-10-09 00:37:13 <diki> shadders:
 14 2011-10-09 00:37:15 <diki> are you shad?
 15 2011-10-09 00:37:18 <mizerydearia> Are there any active bitcoin-geeks/enthusiasts/devs(including other bitcoin-related projects) here?
 16 2011-10-09 00:37:22 <diki> in the forums i mean
 17 2011-10-09 00:37:24 <shadders> probably just saved a few pools quite a few BTC :)
 18 2011-10-09 00:37:32 <shadders> diki: yes
 19 2011-10-09 00:37:44 <diki> can you please add the x-roll-ntime header to poolservj?
 20 2011-10-09 00:37:53 <diki> you have rollntime in the config
 21 2011-10-09 00:37:57 <diki> but no header
 22 2011-10-09 00:38:28 <upb> nah its supposed to be a big number to be easy
 23 2011-10-09 00:38:34 <upb> and small number ot be hard :)
 24 2011-10-09 00:38:44 <upb> or not ?:P
 25 2011-10-09 00:38:56 <shadders> not really a priority... psj is handling loads fine atm without it.  I do plan to do it though.
 26 2011-10-09 00:38:57 <upb> fairly sure that makes sense
 27 2011-10-09 00:39:18 <shadders> upb: yes that makes sense...
 28 2011-10-09 00:39:30 <diki> i also heard you found a flaw in merged mining?
 29 2011-10-09 00:41:56 <shadders> not really a flaw... just a problem that all the daemons have, duplicate works... I think some of the pools have modded the JK patches to fix it now though.
 30 2011-10-09 00:42:34 <diki> duplicate works?
 31 2011-10-09 00:42:39 <diki> can you shed more light about that?
 32 2011-10-09 00:43:24 <shadders> yeah daemon can get itself in a state where it fails to update extraNonce so it only changes the work it delivers 1/sec or when it gets a new transaction
 33 2011-10-09 00:45:26 <shadders> there's a post in the PSJ thread where I talk about it...
 34 2011-10-09 00:45:46 <shadders> it's a pretty evil bug
 35 2011-10-09 01:15:40 <CIA-101> bitcoin: Con Kolivas * ra41db9d436aa cgminer/ (NEWS README configure.ac): Bump version to 2.0.6 and update docs.
 36 2011-10-09 02:56:26 <coderrr> AlexWaters, ping
 37 2011-10-09 06:53:04 <Diablo-D3> http://www.youtube.com/watch?v=XzbTlY1uOJk
 38 2011-10-09 07:46:46 <CIA-101> poolserverj: shadders * e82d3117677c r162 /poolserverj-main/src/main/java/com/shadworld/poolserver/db/shares/BlackHoleSharesDBFlushEngine.java: - add blackhole db share flush engine which does nothing. A quick work around for disabling writing shares to db.
 39 2011-10-09 07:46:47 <CIA-101> poolserverj: shadders * 749d972a7709 r163 / (35 files in 10 dirs): just a small commit
 40 2011-10-09 08:04:38 <diki> just to ask
 41 2011-10-09 08:04:47 <diki> but does anybody know how many addresses are currently in the blockchain?
 42 2011-10-09 08:33:42 <diki> at the moment i think its 2,263,406 unique addresses in the blockchain
 43 2011-10-09 09:00:34 <diki> or maybe not
 44 2011-10-09 09:00:48 <nathan7> =o
 45 2011-10-09 09:01:55 <diki> i get 151,060 unique addresses, thats not even funny
 46 2011-10-09 09:02:35 <diki> so basically, (now)>2,263,406 addresses in the chain, but this number includes dupes, while 151,060 is the unique addresses?
 47 2011-10-09 09:02:45 <diki> lolwut?
 48 2011-10-09 09:02:56 <diki> so for 2 years, just 150k addresses made?
 49 2011-10-09 09:03:48 <diki> can someone confirm this?
 50 2011-10-09 09:06:20 <da2ce7> diki, not made, but have transactions.
 51 2011-10-09 09:06:33 <da2ce7> many more addresses are made, but not used
 52 2011-10-09 09:08:41 <diki> ok, that too
 53 2011-10-09 09:08:49 <diki> but really only 150k??
 54 2011-10-09 09:12:51 <diki> so 2,2 million addresses used in transactions, but only 150k are unique, i am kinda surprised, at this rate..19 quindecillion is indeed far away
 55 2011-10-09 09:15:03 <tcatm> even if it was 100 million unique addresses, 19 quindecillion would be very far away
 56 2011-10-09 09:18:30 <diki> still as the owner of bitcoincharts can you verify its indeed 150k?
 57 2011-10-09 09:19:30 <tcatm> not easily
 58 2011-10-09 09:19:58 <tcatm> how did you calculate that number?
 59 2011-10-09 09:20:47 <diki> by counting the pubkeys in my database
 60 2011-10-09 09:21:03 <diki> well, counting the unique public keys
 61 2011-10-09 09:29:33 <tcatm> 1977193
 62 2011-10-09 09:31:49 <diki> really, you get that number?
 63 2011-10-09 09:31:58 <diki> hmm
 64 2011-10-09 09:32:03 <diki> i guess something is wrong with abe
 65 2011-10-09 09:32:12 <diki> abe is an opensource blockexplorer
 66 2011-10-09 09:32:32 <diki> but for unique keys or total used in transactions?
 67 2011-10-09 09:33:00 <tcatm> unique addresses
 68 2011-10-09 09:33:14 <diki> ok, i guess i have to redo everything...
 69 2011-10-09 09:45:09 <CIA-101> poolserverj: shadders * 652bf2f3611b r164 /poolserverj-main/src/main/java/com/shadworld/poolserver/ (5 files in 3 dirs):
 70 2011-10-09 11:36:01 <diki> well i am dumbstruck
 71 2011-10-09 11:36:09 <diki> i still get 150k unique addresses
 72 2011-10-09 11:41:25 <diki> and i waited an hour for my query to execute
 73 2011-10-09 11:43:14 <makomk> diki: addresses or public keys?
 74 2011-10-09 11:44:19 <diki> its public keys converted to addresses
 75 2011-10-09 11:44:41 <diki> 1 public key should be 1 btc address(or whatever chain)
 76 2011-10-09 11:44:53 <diki> so X keys is X bitcoin addresses still
 77 2011-10-09 11:44:58 <makomk> Not all public keys for addresses in the blockchain are publicly available.
 78 2011-10-09 11:45:07 <diki> meaning?
 79 2011-10-09 11:45:51 <makomk> Meaning you'll get a smaller number if you count the number of public keys than if you count the number of addresses including public keys converted to addresses.
 80 2011-10-09 11:46:17 <diki> uh my db has only public keys
 81 2011-10-09 11:46:23 <diki> so yeah
 82 2011-10-09 11:47:33 <makomk> You're using ABE, right?
 83 2011-10-09 11:47:47 <diki> yes
 84 2011-10-09 11:48:18 <makomk> I'm pretty sure that has a table of addresses somewhere, but I can't remember where...
 85 2011-10-09 11:48:29 <diki> i am using 0.7
 86 2011-10-09 11:49:34 <diki> abe 0.7 pre
 87 2011-10-09 11:49:57 <diki> and i've checked all 22 tables, no addresses just public keys and public key hashes
 88 2011-10-09 12:02:11 <diki> makomk:did you remember?
 89 2011-10-09 12:02:56 <diki> actually there is a table with a field called txout_scriptPubKey
 90 2011-10-09 12:03:09 <diki> but i cant parse that since i cant remove this script from the pubkey
 91 2011-10-09 12:04:26 <sipa> remove the script from the pubkey?
 92 2011-10-09 12:04:55 <diki> i cant know how many characters it is
 93 2011-10-09 12:05:40 <diki> the first addresses in chain have this 4104 before the pubkey
 94 2011-10-09 12:05:53 <diki> after a few more thousand rows its 76a914
 95 2011-10-09 12:06:13 <sipa> you need to pattern-match the script to get the pubkey
 96 2011-10-09 12:06:29 <sipa> there's only a few script templates in use
 97 2011-10-09 12:06:37 <diki> a few?
 98 2011-10-09 12:06:40 <diki> any link to them?
 99 2011-10-09 12:06:48 <diki> i'd love to str_replace them
100 2011-10-09 12:07:02 <sipa> replace them by what?
101 2011-10-09 12:07:13 <diki> with what or by what?
102 2011-10-09 12:07:40 <diki> if i know what i should replace id replace it with nothing i.e it will be removed
103 2011-10-09 12:07:40 <upb> str_replace your way out of a php spaghetty
104 2011-10-09 12:07:50 <sipa> https://en.bitcoin.it/wiki/Script#Scripts
105 2011-10-09 12:08:06 <sipa> only "Standard Transaction to Bitcoin address" and "Standard Generation / transaction to IP address" are used by the default client
106 2011-10-09 12:09:26 <sipa> obviously you can't extract a pubkey from a txout script for a to-bitcoin-address transaction
107 2011-10-09 12:09:35 <sipa> so i'd suggest to convert everything to addresses
108 2011-10-09 12:09:38 <sipa> and then count
109 2011-10-09 12:10:40 <makomk> diki: public key hashes == addresses, more ore less.
110 2011-10-09 12:11:28 <diki> right, and i have 150k unique, tcatm said 1.9 million
111 2011-10-09 12:11:42 <diki> so i basically have 150k unique public keys
112 2011-10-09 12:12:48 <diki> sipa, i dont need transactions, i just need the bitcoin address
113 2011-10-09 12:13:06 <diki> and i mean ALL of them(but the unique ones, no dupes)
114 2011-10-09 12:16:54 <diki> well, i have a very very small amount of bitcoins, but if someone can help me with this, i'd be willing to pay him
115 2011-10-09 13:31:50 <lfm> diki 4104 is a "length coded constant
116 2011-10-09 13:36:05 <lfm> diki actually the 41 is code for a 21 byte constant and the 04 is a standard prefix for the pub key hash
117 2011-10-09 13:47:13 <lfm> diki actually the 41 is code for a 65 byte constant and the 04 is a standard prefix for the pub key hash
118 2011-10-09 13:49:32 <lfm> and script code from 1 to 75 is a code for a constant of that length. opcode 76 means thje next byte is a length code and opcode 77 means a two byte length code, 78 is a 3 byte length code and 79 is a 4 byte length code so you can have constants of any size up to 4gb
119 2011-10-09 13:50:25 <lfm> the 75 byte constant is a 04 byte which never changes for pub keys and 64 bytes of pub key
120 2011-10-09 13:50:54 <lfm> 75 -> 65
121 2011-10-09 14:55:44 <ThomasV> if somebody uses an e-wallet and send me some bitcoins, in which case can I use their sending address in order to refund them?
122 2011-10-09 14:56:14 <ThomasV> I guess never, but I want to be sure
123 2011-10-09 14:58:06 <lfm> well you could always do that but you should tell them to watch for the refund
124 2011-10-09 14:58:22 <imsaguy> no
125 2011-10-09 14:58:39 <imsaguy> Many e-wallet's don't use outgoing addresses that match with incoming.
126 2011-10-09 14:58:57 <lfm> the outgoing will still work as an incomming
127 2011-10-09 14:59:02 <imsaguy> yes
128 2011-10-09 14:59:10 <imsaguy> but the ewallet may never credit the person
129 2011-10-09 14:59:16 <lfm> the right person will get the coins
130 2011-10-09 14:59:20 <imsaguy> the ewallet provider would get the coins
131 2011-10-09 14:59:22 <ThomasV> yes thats the point
132 2011-10-09 14:59:42 <lfm> oh ya, if they use a bank then it could be a problem
133 2011-10-09 15:00:01 <ThomasV> but this seems to be a limitation of the bitcoin daemon, not of the protocol itself
134 2011-10-09 15:00:24 <lfm> but if theyre running their own bitcoin or bitcoind it will work
135 2011-10-09 15:00:48 <ThomasV> lfm: sure, but that's not my question :-)
136 2011-10-09 15:00:55 <Cusipzzz> the ewallet service would get the coins, but if the coins may be credited to someone else or not credited at all depending on the output used
137 2011-10-09 15:00:58 <imsaguy> not even a limitation of the bitcoin daemon
138 2011-10-09 15:01:13 <imsaguy> the ewallet provider uses different outputs, so that the coins are aged
139 2011-10-09 15:01:18 <imsaguy> it helps avoid transaction fees
140 2011-10-09 15:01:18 <lfm> ya ok if you dont know if they're using a bank or not then you need to ask them for a new refund address
141 2011-10-09 15:02:10 <ThomasV> imsaguy: huh? what does it have to do with tx fees?
142 2011-10-09 15:02:35 <lfm> the bank can avoid fees by "aging" coins in certain ways
143 2011-10-09 15:03:01 <ThomasV> why?
144 2011-10-09 15:03:07 <imsaguy> the more confirmations a coin has, the higher their 'score' when sent.
145 2011-10-09 15:03:18 <imsaguy> its a fundamental of bitcoin
146 2011-10-09 15:03:18 <ThomasV> oh ok
147 2011-10-09 15:03:38 <lfm> cuz there are fees accessed for circulating the coins too quicl;y. it is a smap throttle tecnique
148 2011-10-09 15:03:40 <imsaguy> so by using 'older' coins to send, they're more likely to go through
149 2011-10-09 15:03:47 <lfm> spam
150 2011-10-09 15:03:57 <ThomasV> it's a miners policy
151 2011-10-09 15:03:59 <imsaguy> because* assessed*
152 2011-10-09 15:04:08 <imsaguy> ThomasV, its part of the daemon
153 2011-10-09 15:04:21 <lfm> not really fundamental. ita a policy level rather than protocol level thing
154 2011-10-09 15:04:35 <ThomasV> yes, but miners could decide to tinker with that
155 2011-10-09 15:04:42 <imsaguy> only the pools that do that
156 2011-10-09 15:04:46 <imsaguy> self miners won't
157 2011-10-09 15:04:47 <lfm> ya they could and some do.
158 2011-10-09 15:05:10 <imsaguy> AFAIK, luke-jr's the only bigger pool that does
159 2011-10-09 15:06:07 <makomk> That reminds me...
160 2011-10-09 15:06:08 <lfm> imsaguy self miners sometimes fiddle with those policies too. the thing is the standard bitcoin kinda enforces it pre-emptivley so people without good knowledge of the source code prolly cant change it.
161 2011-10-09 15:06:19 <ThomasV> what's the policy of luke's pool?
162 2011-10-09 15:06:56 <lfm> I think luke includes pretty nearly any legal transaction
163 2011-10-09 15:07:34 <imsaguy> no
164 2011-10-09 15:07:36 <imsaguy> they charge by size
165 2011-10-09 15:08:50 <imsaguy> "Will only include transactions in its blocks if the sender pays a fee of at least 0.1 TBC (0.00004096 BTC) per 512 bytes"
166 2011-10-09 15:09:10 <imsaguy> in most cases, not an issue
167 2011-10-09 15:09:34 <imsaguy> and what they don't pick up, slush/deepbit will
168 2011-10-09 15:09:40 <ThomasV> I am interested in the "return to sender" issue because it would allow to implement a street performer protocol service
169 2011-10-09 15:10:09 <imsaguy> it requires the ewallet provider to adapt their software
170 2011-10-09 15:10:16 <lfm> you pretty much gotta ask for a return address of unknown senders
171 2011-10-09 15:10:24 <imsaguy> yep
172 2011-10-09 15:10:46 <ThomasV> yes, but that makes it more complicated
173 2011-10-09 15:11:05 <lfm> ya, well I guess it is more complicated then. sorry
174 2011-10-09 15:11:39 <ThomasV> maybe I can just issue a big fat warning for people using ewallets
175 2011-10-09 15:11:44 <lfm> why would a street performer/begger return donations anyway?
176 2011-10-09 15:12:09 <ThomasV> lfm: google street performer protocol
177 2011-10-09 15:12:20 <ThomasV> it works with a threshold
178 2011-10-09 15:12:23 <lfm> I dunno what that is
179 2011-10-09 15:12:36 <lfm> bah, just keep the excess
180 2011-10-09 15:13:10 <ThomasV> no, it works like this: if the threshold is not reached, everyone gets refunded
181 2011-10-09 15:13:45 <Cusipzzz> just ask for return address on the donation form or w/e
182 2011-10-09 15:13:56 <lfm> ok well you gotta ask for a refund address then. if the doner decines then have a default forward to a designated charity
183 2011-10-09 15:15:46 <ThomasV> no, I will make the refund address optional, for those who care
184 2011-10-09 15:16:39 <ThomasV> lets keep it simple for those who do not use ewallets
185 2011-10-09 15:16:46 <lfm> ya if the doner declines then forward it to a default donation address, some other charity or something
186 2011-10-09 15:17:09 <ThomasV> lfm: http://en.wikipedia.org/wiki/Threshold_pledge_system
187 2011-10-09 15:17:21 <lfm> Oh, if they leave it blank and use a bank, then the bank will prolly get it
188 2011-10-09 15:18:44 <lfm> ThomasV:  you know how to verify an address for validity? ie chech the checksum?
189 2011-10-09 15:18:56 <ThomasV> hmm, how many ewallets have implemented signmessage ? lol
190 2011-10-09 15:19:24 <ThomasV> lfm: sure, why?
191 2011-10-09 15:19:51 <lfm> ok, youll need to do that I guess
192 2011-10-09 15:20:04 <ThomasV> to do what?
193 2011-10-09 15:20:15 <lfm> check the address checksum
194 2011-10-09 15:20:18 <ThomasV> to verify that the address is valid ?
195 2011-10-09 15:20:36 <lfm> cuz you might not be using it right away
196 2011-10-09 15:20:40 <ThomasV> you cannot send to an invalid address
197 2011-10-09 15:20:50 <lfm> right
198 2011-10-09 15:21:37 <ThomasV> lfm: I use abe, it has a function that checks that. but I would love to have it in javascript
199 2011-10-09 15:22:03 <lfm> you cant verify the address by sending to it tho. you need to verify it when it is entered
200 2011-10-09 15:23:51 <ThomasV> lfm: abe has a decode_check_address() function, fyi
201 2011-10-09 15:24:16 <lfm> ok that could work, I dont know abe
202 2011-10-09 15:24:42 <ThomasV> well, it does that and also little bobby tables
203 2011-10-09 15:25:04 <ThomasV> because addresses cannot contain quotes
204 2011-10-09 15:25:06 <ThomasV> :-)
205 2011-10-09 15:25:28 <lfm> ok, I never heard of a bobby table either
206 2011-10-09 15:25:43 <ThomasV> lol
207 2011-10-09 15:26:01 <ThomasV> you must be from another planet
208 2011-10-09 15:26:04 <lfm> ya addresses are pure alphanumeric, no special chars at all
209 2011-10-09 15:26:30 <ThomasV> lfm: http://xkcd.com/327/
210 2011-10-09 15:27:21 <terrytibbs> lol
211 2011-10-09 15:27:36 <CIA-101> bitcoin: globalcitizen  * r7b465b6 / doc/build-unix.txt : Collate generic dependency information. Segregate distribution-specific (Ubuntu/Debian) information. Add Gentoo information. - http://git.io/iazBmQ
212 2011-10-09 15:27:37 <CIA-101> bitcoin: globalcitizen  * r4adf78b / doc/build-unix.txt : Improved Gentoo instructions. - http://git.io/dSr0hg
213 2011-10-09 15:27:38 <CIA-101> bitcoin: Gavin Andresen  * rddd9b39 / doc/build-unix.txt :
214 2011-10-09 15:30:29 <lfm> xkcd IS another planet. they have real skinny limbs there.
215 2011-10-09 15:31:27 <ThomasV> not like in the usa
216 2011-10-09 15:31:57 <lfm> ya, more like ethiopia but hi tech
217 2011-10-09 15:32:02 <luke-jr> gavinandresen: those Gentoo instructions suck :p
218 2011-10-09 15:32:17 <gavinandresen> luke-jr: patches welcome
219 2011-10-09 15:32:20 <luke-jr> gavinandresen: is it supposed to be "install from source" instructions, or simple "install" instructions?
220 2011-10-09 15:32:26 <CIA-101> bitcoin: Gavin Andresen  * r4db9705 / src/bitcoinrpc.cpp :
221 2011-10-09 15:32:39 <luke-jr> gavinandresen: for the latter, it's "layman -a bitcoin && emerge bitcoind"
222 2011-10-09 15:33:02 <gavinandresen> luke-jr: ... but talk to the committer before submitting a competing patch, please
223 2011-10-09 15:33:14 <luke-jr> gavinandresen: never heard of him
224 2011-10-09 15:33:29 <luke-jr> he's not involved in #Bitcoin-Gentoo afaik
225 2011-10-09 15:36:57 <luke-jr> gavinandresen: would you merge a patch to make makefile.unix dynamic-like by default, but allow 'make LINK=static'?
226 2011-10-09 15:38:14 <luke-jr> gavinandresen: also, I think Boost 1.41 is the minimum these days for bitcoind
227 2011-10-09 15:38:21 <gavinandresen> luke-jr: yes, as long as you update doc/release-process.txt to remind us to build LINK=static so we don't accidently send out a broken, doesn't-include-dynamic-libraries relase
228 2011-10-09 15:39:08 <gavinandresen> I'm figuring out how to ship a Mac bitcoin-qt that has all the dynamic dependencies in it right now....
229 2011-10-09 15:49:01 <ThomasV> gavinandresen: is the menubar of bitcoin-qt in its final version? I don't see the point of what's in the 'File' menu. if it's keyboard access, why not use tabs?
230 2011-10-09 15:49:38 <gavinandresen> ThomasV: "final" ?   All of bitcoin is beta software....
231 2011-10-09 15:49:53 <ThomasV> I mean, for 0.5
232 2011-10-09 15:50:44 <gavinandresen> Probably.  I'm concentrating on build- and release-process stuff, I'd rather not include any more major functional changes
233 2011-10-09 15:51:22 <terrytibbs> "Beta" as in "Google-Beta"...
234 2011-10-09 15:51:41 <ThomasV> but do you agree with me on the weirdness of this menu?
235 2011-10-09 15:51:56 <gavinandresen> No opinion, frankly I hardly ever run the GUI
236 2011-10-09 15:52:07 <gavinandresen> (I'm a command-line kind of guy)
237 2011-10-09 15:52:08 <ThomasV> well, that's a problem
238 2011-10-09 15:52:27 <luke-jr> gavinandresen: don't forget coinbaser
239 2011-10-09 15:52:39 <gavinandresen> coinbaser isn't going into 0.5
240 2011-10-09 15:52:50 <imsaguy> lol
241 2011-10-09 15:53:06 <gavinandresen> ... at least not the -coinbaser <cmd> part of it...
242 2011-10-09 15:53:20 <luke-jr> gavinandresen: why not?
243 2011-10-09 15:53:36 <gavinandresen> ... and probably not the setauxwork part of it, either.  Those are major functional changes, and I just said why I don't want to include any more of those
244 2011-10-09 15:53:36 <luke-jr> you had *no* objection to that when I originally submitted it
245 2011-10-09 15:53:51 <luke-jr> gavinandresen: they were ready and tested and had support BEFORE you said that
246 2011-10-09 15:53:57 <gavinandresen> okey doke
247 2011-10-09 15:54:32 <gavinandresen> I apologize for not putting the "no more changes" stick in the ground explicitly earlier
248 2011-10-09 15:54:52 <gavinandresen> ... but I want to get the new GUI out the door.  I did announce the release schedule a couple of weeks ago
249 2011-10-09 15:55:12 <luke-jr> they were ready/tested/etc before you merged other major changes
250 2011-10-09 15:55:16 <freewil> gavinandresen, where can one find the release schedule
251 2011-10-09 15:55:22 <luke-jr> anyhow, like I said, I'm unlikely to maintain it for a 0.6
252 2011-10-09 15:55:26 <luke-jr> so take it or lose it
253 2011-10-09 15:55:30 <ThomasV> and everybody's gonna laugh about the 'File' menu of the gui
254 2011-10-09 15:55:31 <luke-jr> only hurts your users, not me
255 2011-10-09 15:55:35 <gavinandresen> freewil: subscribe to the bitcoin-dev mailing list
256 2011-10-09 15:55:47 <gavinandresen> ThomasV: take it up with laanwj
257 2011-10-09 15:56:10 <ThomasV> gavinandresen: never saw him. is he on irc?
258 2011-10-09 15:56:23 <gavinandresen> ThomasV: he's on github
259 2011-10-09 15:56:33 <ThomasV> yeah, for sure, but I am not
260 2011-10-09 15:56:35 <gavinandresen> ThomasV: ... and I believe he reads the forums
261 2011-10-09 15:56:44 <ThomasV> lol
262 2011-10-09 15:57:12 <ThomasV> and am I really the only one who finds that the menu is weird ?
263 2011-10-09 15:57:37 <ThomasV> (silence)
264 2011-10-09 16:02:31 <ThomasV> in memory of steve jobs, please take the time to launch the GUI and ask yourself the following question: what's the point of the 'File' menu?
265 2011-10-09 16:02:52 <noagendamarket> lol
266 2011-10-09 16:03:14 <noagendamarket> I stopped and restarted bitcoin to see if it would work properly
267 2011-10-09 16:04:35 <ThomasV> this menubar is the sign of an unfinished product. Steve would have died earlier if he had lived long enough to see it... err... forget about the logic
268 2011-10-09 16:08:24 <wumpus> ThomasV: bitcoin is an unfinished product, it's not at 1.0 yet
269 2011-10-09 16:08:47 <ThomasV> yeah
270 2011-10-09 16:09:17 <JFK911> ThomasV: Thanks for making the case that this software is too late
271 2011-10-09 16:10:01 <ThomasV> too late, as in "should have been released earlier" ?
272 2011-10-09 16:10:06 <wumpus> and yes the file menu makes no sense, feel free to submit a patch
273 2011-10-09 16:10:56 <ThomasV> wumpus: I feel free enough to say it without submitting a patch
274 2011-10-09 16:11:23 <wumpus> yes, but your complaint will get fixed a lot sooner if you can submit a patch
275 2011-10-09 16:11:57 <ThomasV> it's not a complaint about a bug that needs to be fixed. it's about a poor decision that needs to be questionned
276 2011-10-09 16:12:11 <freewil> it probably wasnt even a decision
277 2011-10-09 16:12:21 <freewil> it probably was just whatever works
278 2011-10-09 16:12:21 <wumpus> it is a leftover from before the ui was tabbed
279 2011-10-09 16:13:03 <ThomasV> it is not tabbed ; it it was tabbed, it would be possible to navigate between tabs with keyboard
280 2011-10-09 16:13:17 <ThomasV> it would be much better if it was tabbed, btw
281 2011-10-09 16:13:18 <wumpus> I'm not sure that we should have a file menu at all, then again, it will be neccesary again when we want to add/remove loading external wallets
282 2011-10-09 16:13:41 <wumpus> so what keyboard shortcuts for that?
283 2011-10-09 16:14:35 <ThomasV> depends on your window manager. on my box I switch tabs with ctrl-pageUp/Down
284 2011-10-09 16:14:49 <wumpus> I don't understand
285 2011-10-09 16:15:12 <wumpus> tabs are something within the application, the window manager has nothing to do with it
286 2011-10-09 16:15:16 <ThomasV> are you asking how to navigate between tabs ?
287 2011-10-09 16:15:30 <wumpus> I mean like firefox tabs
288 2011-10-09 16:15:47 <ThomasV> oh, I am not sure, I think you can change that with gnome, but I might be wrong
289 2011-10-09 16:16:10 <ThomasV> with FF I switch with ctrl-PageUp/Down
290 2011-10-09 16:16:11 <wumpus> alt-1 .. alt-<N>  would work I guess
291 2011-10-09 16:16:18 <wumpus> like gnome-terminal
292 2011-10-09 16:16:42 <ThomasV> oh yes that works too in FF; I did not know that one
293 2011-10-09 16:16:48 <wumpus> anyway, it's something that needs to be explicitly added to an application, it's not like you get that for free from the window manager or something
294 2011-10-09 16:16:59 <ThomasV> anyway, the fact is that bitcion-qt is NOT using tabs atm
295 2011-10-09 16:17:13 <wumpus> well it is using tabs, but navigation is mouse-only
296 2011-10-09 16:17:34 <ThomasV> wumpus: I suspect you get it for free from qt
297 2011-10-09 16:17:43 <ThomasV> if you use proper tabs
298 2011-10-09 16:17:52 <wumpus> sigh, no you don't
299 2011-10-09 16:18:10 <wumpus> do it yourself if you know it better
300 2011-10-09 16:18:19 <ThomasV> no time
301 2011-10-09 16:18:56 <wumpus> you seem to have a lot of time to chat here
302 2011-10-09 16:19:06 <ThomasV> yes
303 2011-10-09 16:19:15 <wumpus> so that's no excuse
304 2011-10-09 16:20:00 <ThomasV> heh, I am taking time pr provide feedback, so please try to be constructive
305 2011-10-09 16:20:44 <ThomasV> and you agree that the file menu is silly
306 2011-10-09 16:21:19 <wumpus> it's great that you provide feedback, but I don't like the way that you do it.. if you want keyboard navigation just ask for it, don't come with nonsense like 'it is not tabbed!!!'
307 2011-10-09 16:21:43 <wumpus> it's a project in development and keyboard nav is on my todo list
308 2011-10-09 16:22:12 <ThomasV> ok, fair enough, but I was thinking that keyboard navigation comes automatically with qt
309 2011-10-09 16:22:17 <ThomasV> sorry for that
310 2011-10-09 16:22:47 <wumpus> yes you can set keyboard shortcuts for any action, but I haven't configured them for everything
311 2011-10-09 16:22:52 <wumpus> ok np
312 2011-10-09 16:24:54 <ThomasV> note that with gnome-terminal, you can change tabs with alt-<n> and ctrl-PageUp/Down too ; I just checked
313 2011-10-09 16:25:45 <wumpus> ok, makes sense to use that too, then
314 2011-10-09 16:26:11 <luke-jr> https://github.com/bitcoin/bitcoin/pull/576
315 2011-10-09 16:27:22 <ThomasV> wumpus: btw, if you allow me another remark, it is very uncommon to use the letter E for the File->Exit menu. In general, the 'x' of Exit is the shortcut
316 2011-10-09 16:27:53 <wumpus> so, what do you propose doing with the file menu?
317 2011-10-09 16:28:28 <ThomasV> remove the radio buttons, keep exit but change the shortcut to x
318 2011-10-09 16:28:42 <wumpus> I don't think nuking it completely is a good idea either, but having only Exit makes no sense either
319 2011-10-09 16:28:50 <ThomasV> or if you don't like 'exit', use 'quit'
320 2011-10-09 16:29:25 <wumpus> yes it seems qt applications generlaly use 'quit' (Ctrl-Q)
321 2011-10-09 16:29:48 <ThomasV> well, 'help' too has only one item ; you can merge them if you don't like singletons
322 2011-10-09 16:30:04 <ThomasV> for me singletons are fine
323 2011-10-09 16:30:14 <wumpus> singletons suck :P
324 2011-10-09 16:30:32 <CIA-101> bitcoin: various bitcoind_build_improvements * rb4d058..903a25 bitcoind-personal/ (36 files in 20 dirs): (9 commits)
325 2011-10-09 16:30:48 <ThomasV> as you said earlier, you expect new menu items to be merged later, so they will not remain singletons
326 2011-10-09 16:30:53 <wumpus> then again, bitcoin might add file-related options later, so I'll keep it in
327 2011-10-09 16:34:59 <wumpus> interesting, Chromium doesn't even have Quit in its file menu
328 2011-10-09 16:35:17 <wumpus> Firefox has Ctrl-Q/Quit
329 2011-10-09 16:36:38 <wumpus> Qt Creator has Ctrl-Q/Exit
330 2011-10-09 16:36:39 <ThomasV> chromium sucks
331 2011-10-09 16:36:39 <wumpus> lol
332 2011-10-09 16:37:08 <wumpus> no browser wars please
333 2011-10-09 16:37:12 <ThomasV> ok
334 2011-10-09 16:37:15 <ThomasV> :-)
335 2011-10-09 16:37:38 <cjdelisle> chromium sucks but at least it sucks quickly
336 2011-10-09 16:37:51 <lianj> haha
337 2011-10-09 16:37:53 <cjdelisle> firefox sucks and it takes an insultingly long time to do it
338 2011-10-09 16:38:47 <wumpus> I mean, if you want a browser that sucks get IE6 :-)
339 2011-10-09 16:39:07 <ThomasV> btw, please allow me another remark/complaint/rant/whatever you think it should be called. why does the daemon have a 'stop' command and no 'start' command that would spawn a daemon?
340 2011-10-09 16:39:28 <ThomasV> hmm I think this is for gavin but he left
341 2011-10-09 16:39:51 <wumpus> because you can only send commands to it if it is already started
342 2011-10-09 16:40:11 <wumpus> sending a start command would be pointless (unless you want self-replicating daemons0
343 2011-10-09 16:40:32 <ThomasV> no, most daemons are spawned with a start command
344 2011-10-09 16:40:52 <wumpus> though it would be fun to add a start command and just ignore it :-)
345 2011-10-09 16:41:00 <wumpus> no no no , you're confused with /etc/init.d scripts
346 2011-10-09 16:41:18 <wumpus> those have start and stop
347 2011-10-09 16:41:29 <wumpus> and are a launcher construct around the daemon itself
348 2011-10-09 16:41:32 <ThomasV> yes, but why is it different?
349 2011-10-09 16:41:52 <wumpus> because it is. Look at apached command line arguments for example.
350 2011-10-09 16:42:22 <wumpus> you know what's the problem and why you're confused?
351 2011-10-09 16:42:31 <wumpus> because bitcoind is *both* rpc client and daemon
352 2011-10-09 16:42:32 <ThomasV> no?
353 2011-10-09 16:42:32 <wumpus> :-)
354 2011-10-09 16:42:48 <ThomasV> I don't get it
355 2011-10-09 16:42:57 <wumpus> that's the problem
356 2011-10-09 16:43:24 <wumpus> most servers have a daemon executable and a process to send it commands
357 2011-10-09 16:43:50 <wumpus> bitcoin combines them ,and is a client or server dependent on the command line arguments
358 2011-10-09 16:43:51 <ThomasV> yes, like apache2ctl
359 2011-10-09 16:43:55 <wumpus> right
360 2011-10-09 16:44:15 <wumpus> bitcoind is apached *and* apache2ctl
361 2011-10-09 16:44:39 <wumpus> I've already complained about that on the mailing list one time, but it's kept like this to not break compatibility
362 2011-10-09 16:44:43 <ThomasV> I understand that, but that does not explain why it would be bad to use a 'start' syntax
363 2011-10-09 16:45:04 <wumpus> I actually had to convince people that it's not a good idea to make bitcoin-qt work as command-line RPC client :')
364 2011-10-09 16:45:53 <ThomasV> oh, because some devs wanted to have a server + a client in separate processes ?
365 2011-10-09 16:46:10 <wumpus> adding start syntax would be weird because adding a  command to the command line means send it to the RPC service
366 2011-10-09 16:46:28 <wumpus> otherwie you have to parse some commands on the command line and if no match send it to the RPC service
367 2011-10-09 16:46:39 <ThomasV> well, it would be an exception
368 2011-10-09 16:46:44 <wumpus> that's even more confused than it is now
369 2011-10-09 16:46:49 <wumpus> please let's not go that way
370 2011-10-09 16:47:26 <ThomasV> have you ever used python's daemon library ?
371 2011-10-09 16:47:28 <wumpus> let's clean up the source and separate concerns instead of conflating them even more
372 2011-10-09 16:47:43 <freewil> you just have to understand that when you send a command like 'stop' you are sending a rpc command, thus the dameon must already be started
373 2011-10-09 16:47:44 <ThomasV> heh
374 2011-10-09 16:47:44 <wumpus> no
375 2011-10-09 16:47:51 <wumpus> freewil: +1
376 2011-10-09 16:49:10 <ThomasV> and RPC shall remain pure!
377 2011-10-09 16:49:24 <wumpus> freewil: it's another mental step though
378 2011-10-09 16:49:41 <freewil> what is
379 2011-10-09 16:50:25 <wumpus> freewil: that it sometimes launches a new daemon and sometimes send a command to a running daemon
380 2011-10-09 16:51:03 <freewil> im confused, im not saying we should add start to bitcoind as a command
381 2011-10-09 16:51:27 <wumpus> I know, I'm just talking about the current state
382 2011-10-09 16:51:30 <ThomasV> well, it's not so important
383 2011-10-09 16:51:36 <wumpus> which is why ThomasV got confused
384 2011-10-09 16:52:16 <freewil> it is rather confusing
385 2011-10-09 16:52:20 <wumpus> it would be better if it were separate commands
386 2011-10-09 16:52:26 <freewil> agreed
387 2011-10-09 16:52:34 <ThomasV> no it would not
388 2011-10-09 16:52:39 <freewil> separate files/processes
389 2011-10-09 16:53:06 <freewil> bitcoind acts both as a rpc client and a rpc server
390 2011-10-09 16:53:13 <wumpus> bitcoin-rpc-client <host:port> <command> <args...>
391 2011-10-09 16:53:15 <wumpus> or something like that
392 2011-10-09 16:53:43 <ThomasV> well, keep it as is, please, no two commands
393 2011-10-09 16:54:05 <JFK911> i have a need for the rpc commands to be localized
394 2011-10-09 16:54:17 <JFK911> my partner doesn't program in english
395 2011-10-09 16:54:20 <ThomasV> lol
396 2011-10-09 16:54:23 <wumpus> rpc client is something entirely conceptually different from the daemon
397 2011-10-09 16:54:25 <JFK911> is there unicode support?
398 2011-10-09 16:54:27 <luke-jr> JFK911: that's his problem
399 2011-10-09 16:54:30 <wumpus> lol JFK911
400 2011-10-09 16:54:33 <luke-jr> JFK911: programming is English
401 2011-10-09 16:54:46 <JFK911> it's lisp
402 2011-10-09 16:54:51 <wumpus> yes there is unicode support
403 2011-10-09 16:55:00 <wumpus> if your terminal supports it
404 2011-10-09 16:55:03 <wumpus> :P
405 2011-10-09 16:55:24 <wumpus> but localizing API commands, lol
406 2011-10-09 16:55:29 <wumpus> I think that'd be a first
407 2011-10-09 16:55:35 <ThomasV> heh
408 2011-10-09 16:56:13 <ThomasV> in german verbs have separable particles, that could be fun
409 2011-10-09 16:56:22 <JFK911> someone probably already did that with preprocessor
410 2011-10-09 16:56:34 <wumpus> just define constants in your favourite language that map to the english commands...
411 2011-10-09 16:56:35 <JFK911> ThomasV: well all the programming ive seen in german uses infinitive form verbs
412 2011-10-09 16:57:18 <ThomasV> JFK911: that's a pity!
413 2011-10-09 16:57:21 <JFK911> eg. a process might be called muschi_lecken
414 2011-10-09 17:07:48 <wumpus> ThomasV: https://github.com/laanwj/bitcoin-qt/commit/d934e7e3ddd7d919e58f76705439cdb15c745951
415 2011-10-09 17:08:22 <ThomasV> nice
416 2011-10-09 17:25:33 <ThomasV> wumpus: btw, are you planning to provide a gui frontend to signmessage/verifymessage ?
417 2011-10-09 17:26:05 <wumpus> I don't know
418 2011-10-09 17:26:22 <ThomasV> I think it would be useful
419 2011-10-09 17:27:04 <wumpus> so it's kind of PGP cleartext signing but using the bitcoin private keys?
420 2011-10-09 17:27:17 <ThomasV> yes
421 2011-10-09 17:27:30 <wumpus> it sounds dangerous somehow.. what if someone makes you sign a valid transaction?
422 2011-10-09 17:27:44 <gmaxwell> They can't. That risk was thought of. :)
423 2011-10-09 17:28:09 <wumpus> how is that prevented?
424 2011-10-09 17:28:27 <gmaxwell> And extra run of the hash function.
425 2011-10-09 17:29:00 <wumpus> cool
426 2011-10-09 17:29:05 <ThomasV> nice
427 2011-10-09 17:30:15 <wumpus> I have no problem with the feature itself then
428 2011-10-09 17:34:53 <AlexWaters> coderrr: pong
429 2011-10-09 18:48:50 <FellowTraveler> hi all.
430 2011-10-09 19:22:03 <odysseus654> okay, trying another random channel.  is this the place to ask for cmdline parameters for using db_dump for wallet.dat recovery?
431 2011-10-09 19:28:44 <odysseus654> ok, pretty quiet in here.  I really don't want to have to yank BitCoin into a debugger to figure out what's going on here :-( oh well
432 2011-10-09 20:09:45 <diki> ;seen
433 2011-10-09 20:09:52 <diki> ,,seen
434 2011-10-09 20:09:52 <gribble> (seen [<channel>] <nick>) -- Returns the last time <nick> was seen and what <nick> was last seen saying. <channel> is only necessary if the message isn't sent on the channel itself. <nick> may contain * as a wildcard.
435 2011-10-09 20:09:56 <diki> ,,seen conman
436 2011-10-09 20:09:57 <gribble> (seen [<channel>] <nick>) -- Returns the last time <nick> was seen and what <nick> was last seen saying. <channel> is only necessary if the message isn't sent on the channel itself. <nick> may contain * as a wildcard.
437 2011-10-09 20:10:05 <diki> ,,seen #bitcoin-mining conman
438 2011-10-09 20:10:06 <gribble> (seen [<channel>] <nick>) -- Returns the last time <nick> was seen and what <nick> was last seen saying. <channel> is only necessary if the message isn't sent on the channel itself. <nick> may contain * as a wildcard.
439 2011-10-09 20:10:11 <diki> ,,seen bitcoin-mining conman
440 2011-10-09 20:10:12 <gribble> (seen [<channel>] <nick>) -- Returns the last time <nick> was seen and what <nick> was last seen saying. <channel> is only necessary if the message isn't sent on the channel itself. <nick> may contain * as a wildcard.
441 2011-10-09 20:10:16 <diki> wtf work
442 2011-10-09 20:10:27 <diki> ,,seen [bitcoin-mining] conman
443 2011-10-09 20:10:28 <gribble> (seen [<channel>] <nick>) -- Returns the last time <nick> was seen and what <nick> was last seen saying. <channel> is only necessary if the message isn't sent on the channel itself. <nick> may contain * as a wildcard.
444 2011-10-09 20:10:32 <diki> ,,seen [#bitcoin-mining] conman
445 2011-10-09 20:10:33 <gribble> (seen [<channel>] <nick>) -- Returns the last time <nick> was seen and what <nick> was last seen saying. <channel> is only necessary if the message isn't sent on the channel itself. <nick> may contain * as a wildcard.
446 2011-10-09 20:15:25 <TuxBlackEdo> wtf
447 2011-10-09 20:15:43 <gribble> (seen [<channel>] <nick>) -- Returns the last time <nick> was seen and what <nick> was last seen saying. <channel> is only necessary if the message isn't sent on the channel itself. <nick> may contain * as a wildcard.
448 2011-10-09 20:15:43 <TuxBlackEdo> ,,seen #bitcoin-mining conman
449 2011-10-09 20:15:51 <TuxBlackEdo> ,,seen bitcoin-mining conman
450 2011-10-09 20:15:52 <gribble> (seen [<channel>] <nick>) -- Returns the last time <nick> was seen and what <nick> was last seen saying. <channel> is only necessary if the message isn't sent on the channel itself. <nick> may contain * as a wildcard.
451 2011-10-09 20:19:25 <gmaxwell> ...
452 2011-10-09 21:36:40 <nanotube> TuxBlackEdo: try ;;
453 2011-10-09 21:36:52 <nanotube> TuxBlackEdo: the double-comma is for one-word inline commands. so that's only picking up the 'seen'
454 2011-10-09 22:36:39 <b4epoche_> I'm trying to come up with simple ways to explain how public/private key pairs are used in Bitcoin.
455 2011-10-09 22:37:02 <b4epoche_> public keys (in disguise) are obviously used as addresses
456 2011-10-09 22:37:42 <b4epoche_> is it correct to say that the private keys essentially just confirm who 'owns' an address?
457 2011-10-09 22:47:34 <BTCTrader_> i think so
458 2011-10-09 22:47:45 <BTCTrader_> a dev could confirm
459 2011-10-09 23:16:39 <sipa> b4epoche_: sounds correct to me
460 2011-10-09 23:17:36 <b4epoche_> okay&  it's proving hard to dumb down complex things when you start to realize you don't fully understand the complex thing
461 2011-10-09 23:17:39 <b4epoche_> thx
462 2011-10-09 23:18:15 <sipa> see it as mailboxes: anyone can put something in it, but only one with the key can unlock
463 2011-10-09 23:18:28 <gmaxwell> b4epoche_: I used a model of saying an address is like a safty deposit box, and the private key is the key to the box.
464 2011-10-09 23:18:41 <gmaxwell> Oh, sipa's mailboxes are better.
465 2011-10-09 23:19:12 <sipa> the pubkey/address is the number of the mailbox, the secret key the key
466 2011-10-09 23:19:45 <b4epoche_> oh, like the postal mailboxes...
467 2011-10-09 23:22:04 <b4epoche_> these guys:  http://blog.timesunion.com/opinion/files/2011/08/0822_WVpostal.jpg
468 2011-10-09 23:25:09 <sipa> i was thinking about this: http://community.tradeking.com/upload/0001/0435/apartment_mailboxes.jpg
469 2011-10-09 23:25:49 <b4epoche_> not sure anyone can actually put stuff in those mailboxes.
470 2011-10-09 23:27:19 <shadders> anyone care to help me figure out out construct a merkle branch?
471 2011-10-09 23:27:58 <shadders> Trying to decipher GetMerkleBranch but with all the bitshifts and xors it's a bit confusing
472 2011-10-09 23:28:16 <shadders> Just trying to understand the order to add the nodes...
473 2011-10-09 23:28:44 <shadders> i.e. if I want the branch for node 4 of a 5 node tree.
474 2011-10-09 23:28:44 <sipa> b4epoche_: if youknow the address, why not?
475 2011-10-09 23:29:07 <shadders> would I begin by adding node 3,4 or 4,3?
476 2011-10-09 23:29:42 <sipa> what is node 3,4 or 4,3?
477 2011-10-09 23:29:43 <b4epoche_> usually only the postal worker has a master key to open a panel in the back of those to put mail in
478 2011-10-09 23:30:14 <shadders> (((12)(34))((55)(55)))
479 2011-10-09 23:30:15 <shadders> /      /      /  \n1916406
480 2011-10-09 23:30:30 <gmaxwell> sipa: I was actually thinking of apartment mailboxes too, but I couldn't find a picture of one that you could obviously put mail in.
481 2011-10-09 23:30:43 <gmaxwell> b4epoche_: there are apartment mailboxes where you can, they usually have a thin slot.
482 2011-10-09 23:30:51 <shadders> if it's 3,4 that would imply that the branch for either of a pair would be the same
483 2011-10-09 23:31:06 <b4epoche_> gmaxwell:  ah, never lived in an apartment with those kind
484 2011-10-09 23:31:21 <shadders> sipa: branch is just a list of nodes from what I can see...
485 2011-10-09 23:31:28 <sipa> yes
486 2011-10-09 23:31:44 <sipa> but what is 4,3?
487 2011-10-09 23:32:20 <shadders> so I'm guessing branch for node 4 would be something like: 3,4,12,34,1234,5555,root
488 2011-10-09 23:32:54 <shadders> 4,3 is two entries in the list... 1st 4, 2nd 3
489 2011-10-09 23:33:07 <sipa> ah
490 2011-10-09 23:33:22 <sipa> is the entry itself included?
491 2011-10-09 23:33:45 <shadders> good question
492 2011-10-09 23:35:21 <sipa> i never imagined so
493 2011-10-09 23:35:58 <shadders> I might have to see if I can javafy the code so I can run it and see what it produces...
494 2011-10-09 23:36:02 <sipa> but obviously one needs to know whether the merkle branch elements are lw
495 2011-10-09 23:36:15 <sipa> left or right partners
496 2011-10-09 23:36:30 <shadders> yes
497 2011-10-09 23:37:05 <shadders> the example branch I posted above seems the most complete.  no ambiguity
498 2011-10-09 23:37:55 <shadders> though the next question is, if yr building from node five do you duplicate the missing nodes in the branch...
499 2011-10-09 23:38:22 <shadders> i.e. 5,5,55,55,1234,5555,root
500 2011-10-09 23:38:40 <shadders> if you didn't then again it would be ambiguous
501 2011-10-09 23:40:31 <sipa> the merkle branch in your example is 3, 12, 5555
502 2011-10-09 23:41:49 <shadders> ahh it's only nodes you to calculate the rest.
503 2011-10-09 23:41:58 <sipa> and you need to know the position of what you're checking
504 2011-10-09 23:43:34 <shadders> ok.. so using position you can calculate if each of the provided nodes in the branch represent a left or right
505 2011-10-09 23:44:00 <sipa> indeed
506 2011-10-09 23:45:30 <shadders> k... I think I can work it out from there... will try to implement GetMerkleBranch anyway to see if I can understand the bitshifting convolutions
507 2011-10-09 23:45:39 <shadders> thanks sipa :)