1 2013-06-10 00:05:32 darkee has joined
2 2013-06-10 00:13:13 brson has quit (Quit: leaving)
3 2013-06-10 00:14:33 Internet13 has quit (Quit: Leaving)
4 2013-06-10 00:16:03 Internet13 has joined
5 2013-06-10 00:16:46 viperhr has quit (Ping timeout: 240 seconds)
6 2013-06-10 00:20:50 atweiden has joined
7 2013-06-10 00:22:50 CodeShark has joined
8 2013-06-10 00:24:19 owowo has joined
9 2013-06-10 00:26:22 da2ce7 has joined
10 2013-06-10 00:26:30 jaequery has joined
11 2013-06-10 00:26:56 jaequery has quit (Max SendQ exceeded)
12 2013-06-10 00:27:49 jaequery has joined
13 2013-06-10 00:28:06 jaequery has quit (Client Quit)
14 2013-06-10 00:28:28 msvb-lab has quit (Quit: msvb-lab)
15 2013-06-10 00:30:25 _W_ has quit (Read error: Connection reset by peer)
16 2013-06-10 00:30:51 B0g4r7_ has quit (Read error: Connection reset by peer)
17 2013-06-10 00:31:05 B0g4r7_ has joined
18 2013-06-10 00:33:53 _W_ has joined
19 2013-06-10 00:33:53 one_zero has joined
20 2013-06-10 00:37:00 MobPhone has joined
21 2013-06-10 00:48:47 macboz has joined
22 2013-06-10 00:49:34 GordonG3kko has quit (Remote host closed the connection)
23 2013-06-10 00:52:38 <shesek> is the signature length fixed?
24 2013-06-10 00:52:53 GordonG3kko has joined
25 2013-06-10 01:00:35 ixaeon has quit (Ping timeout: 264 seconds)
26 2013-06-10 01:01:56 <shesek> with hashTransactionForSignature for a p2sh input, should the "connectedScript" be the p2sh hashed script (HASH_160 ... OP_EQUAL) or the actual script?
27 2013-06-10 01:03:01 <shesek> I'm not sure if its called the same bitcoind's source, but both bitcoinj bitcoinjs call it hashTransactionForSignature
28 2013-06-10 01:03:09 <shesek> * same in
29 2013-06-10 01:04:28 brson has joined
30 2013-06-10 01:09:01 ixaeon has joined
31 2013-06-10 01:09:09 ixaeon has quit (Max SendQ exceeded)
32 2013-06-10 01:15:58 fourlights has joined
33 2013-06-10 01:16:31 fourlights_ has quit (Ping timeout: 245 seconds)
34 2013-06-10 01:19:43 GordonG3kko has quit (Remote host closed the connection)
35 2013-06-10 01:22:53 GordonG3kko has joined
36 2013-06-10 01:23:26 <TheUni> sipa: no chance of running a gitian attempt, then?
37 2013-06-10 01:24:33 johnsoft has quit (Ping timeout: 276 seconds)
38 2013-06-10 01:25:59 <warren> TheUni: you need someone to do a gitian attempt?
39 2013-06-10 01:26:29 <TheUni> warren: yea. there's absolutely zero chance it will work (i'm stabbing in the dark), so i was hoping to find someone willing to help massage it a bit
40 2013-06-10 01:27:36 michagogo has quit (Quit: goodnight)
41 2013-06-10 01:27:39 <warren> TheUni: point me at what you want tested/fixed
42 2013-06-10 01:27:56 <TheUni> so far i've rewritten the win32 dependencies to be more sane. i've tested those locally and believe they're in working order. any chance i could convince you to run that bit as a start?
43 2013-06-10 01:28:33 <warren> master + your patches?
44 2013-06-10 01:28:41 <warren> are they in a pull req?
45 2013-06-10 01:28:57 <TheUni> yes. i can point you to my branch in a min, just need to look over it once more
46 2013-06-10 01:29:04 <warren> ok
47 2013-06-10 01:29:14 <TheUni> thanks, this is a big help
48 2013-06-10 01:30:52 <warren> TheUni: you on mac?
49 2013-06-10 01:31:18 <TheUni> on a macbook, i live in a shell on my linux workstation though
50 2013-06-10 01:31:24 <warren> TheUni: one of my contractors is working on virtualbox-based gitian that should run on mac
51 2013-06-10 01:31:54 * nsh still doesn't understand why gitian is so hard
52 2013-06-10 01:31:58 <nsh> (not tried)
53 2013-06-10 01:32:41 <TheUni> warren: gitian makes me want to drive blunt objects through my head. i have hopes/plans to make it unnecessary. I have no desire to go anywhere near it unless i have to
54 2013-06-10 01:32:48 melvster has quit (Ping timeout: 264 seconds)
55 2013-06-10 01:32:58 <nsh> would have maybe thought it would be easy to set up a VM for deterministic builds once and then it would be done
56 2013-06-10 01:33:19 <nsh> that's kinda one of the benefits of virtualization
57 2013-06-10 01:33:22 <warren> TheUni: The unpatched ancient deps part scares me.
58 2013-06-10 01:33:32 <TheUni> unpatched ancient deps?
59 2013-06-10 01:33:45 <warren> TheUni: for example, the static linked openssl from 2009
60 2013-06-10 01:33:56 <TheUni> what about it?
61 2013-06-10 01:34:08 <nsh> that it's from 2009?
62 2013-06-10 01:34:15 * warren looks
63 2013-06-10 01:34:56 <TheUni> warren: how exactly does gitian improve that situation at all?
64 2013-06-10 01:35:22 BurtyB has quit (Read error: Connection reset by peer)
65 2013-06-10 01:35:49 BurtyB has joined
66 2013-06-10 01:35:53 <warren> TheUni: It doesn't.
67 2013-06-10 01:36:33 <TheUni> exactly. seems to me it fails badly at the very few things it's designed to do, and makes the project nearly impossibly difficult to build for release
68 2013-06-10 01:36:39 <warren> ok, I was wrong, it's unpatched from May 2012
69 2013-06-10 01:36:49 <warren> I might have been looking at a different dep
70 2013-06-10 01:36:49 johnsoft has joined
71 2013-06-10 01:41:24 bitanarchy has quit (Quit: Leaving)
72 2013-06-10 01:44:37 <devrandom> TheUni: if you have a better idea for creating trusted builds, I would be happy to hear :)
73 2013-06-10 01:45:09 <TheUni> devrandom: i'm working on it one piece at a time
74 2013-06-10 01:45:21 <warren> devrandom: hey, could you please process your pull requests =)
75 2013-06-10 01:45:43 <TheUni> devrandom: basically i intend to continue to use gitian as-is, i think that part is fine. problem is that it's doing things that (imo) aren't its job
76 2013-06-10 01:46:58 <devrandom> warren: pull #40?
77 2013-06-10 01:47:09 <devrandom> TheUni: say more
78 2013-06-10 01:47:13 agnostic98 has quit (Remote host closed the connection)
79 2013-06-10 01:47:19 <warren> devrandom: not urgent, but yes.
80 2013-06-10 01:47:28 <TheUni> devrandom: i'll say it with a POC :)
81 2013-06-10 01:47:46 <devrandom> okay
82 2013-06-10 01:47:48 <devrandom> warren: merged
83 2013-06-10 01:48:06 <warren> thx
84 2013-06-10 01:48:44 <TheUni> devrandom: i have no problem discussing, it just tends to lead towards bikeshedding when there's nothing concrete to show. so i'd prefer to hold off until then
85 2013-06-10 01:49:34 <warren> TheUni: bikeshedding: I'll happily hand over my gitian rewrite that makes deterministic in fedora vm's.
86 2013-06-10 01:49:58 agnostic98 has joined
87 2013-06-10 01:49:58 agnostic98 has quit (Remote host closed the connection)
88 2013-06-10 01:50:31 agnostic98 has joined
89 2013-06-10 01:50:39 <devrandom> TheUni: okay, up to you
90 2013-06-10 01:50:50 <devrandom> bbl
91 2013-06-10 01:51:45 patcon has joined
92 2013-06-10 01:52:15 <shesek> when preparing a transaction for signature, what script should I use for p2sh inputs? the original hashed output script (HASH_160 ... OP_EQUAL) or the original pre-hashed script?
93 2013-06-10 01:52:40 <TheUni> warren: first 2 basic changes are here: https://github.com/theuni/bitcoin/tree/gitian-cleanup
94 2013-06-10 01:52:47 <TheUni> warren: mind giving that a go?
95 2013-06-10 01:53:27 <warren> k
96 2013-06-10 01:53:46 <warren> just gitian win32?
97 2013-06-10 01:54:11 <TheUni> not even that, it'll fail. just win32-deps and win32-boost
98 2013-06-10 01:54:21 duckybsd has quit (Remote host closed the connection)
99 2013-06-10 01:55:00 <warren> oh ok
100 2013-06-10 01:55:26 <TheUni> i need the output from those in order to do win32. i have a simulated output here, but i'd prefer to work with real generated zips
101 2013-06-10 01:55:46 <warren> logs too?
102 2013-06-10 01:56:06 <TheUni> if they actually manage to build, zips should be enough i think
103 2013-06-10 02:00:20 atweiden has quit (Ping timeout: 252 seconds)
104 2013-06-10 02:01:01 <shesek> anyone? I'm really close to finish this really cool project :)
105 2013-06-10 02:01:10 tyn has quit (Ping timeout: 246 seconds)
106 2013-06-10 02:01:17 atweiden has joined
107 2013-06-10 02:02:03 defunctzombie is now known as defunctzombie_zz
108 2013-06-10 02:02:31 BeeDeePee has quit (Ping timeout: 256 seconds)
109 2013-06-10 02:03:07 <warren> building
110 2013-06-10 02:05:37 ixaeon has joined
111 2013-06-10 02:05:42 atweiden has quit (Ping timeout: 246 seconds)
112 2013-06-10 02:05:44 ixaeon has quit (Max SendQ exceeded)
113 2013-06-10 02:07:39 BeeDeePee has joined
114 2013-06-10 02:12:17 ralphtheninja has quit (Ping timeout: 240 seconds)
115 2013-06-10 02:14:23 malexmedia has quit (Ping timeout: 264 seconds)
116 2013-06-10 02:18:02 atweiden has joined
117 2013-06-10 02:21:30 Marc____ has joined
118 2013-06-10 02:22:57 <warren> TheUni: boost worked, trying oter
119 2013-06-10 02:23:14 <TheUni> great, thanks
120 2013-06-10 02:23:33 <TheUni> warren: is it possible for you to link me to the zips when done?
121 2013-06-10 02:23:36 <warren> yes
122 2013-06-10 02:25:11 atweiden has quit (Ping timeout: 264 seconds)
123 2013-06-10 02:25:35 atweiden has joined
124 2013-06-10 02:26:35 treaki__ has joined
125 2013-06-10 02:27:06 treaki_ has quit (Ping timeout: 248 seconds)
126 2013-06-10 02:31:26 hsmiths has quit (Ping timeout: 248 seconds)
127 2013-06-10 02:32:19 macboz has quit (Ping timeout: 246 seconds)
128 2013-06-10 02:35:02 hsmiths has joined
129 2013-06-10 02:36:06 macboz has joined
130 2013-06-10 02:36:59 Eiii has joined
131 2013-06-10 02:36:59 Eiii has quit (Changing host)
132 2013-06-10 02:36:59 Eiii has joined
133 2013-06-10 02:37:47 <TheUni> warren: appeared to be successful?
134 2013-06-10 02:41:23 <TheUni> welp, it's enough for a successful bitcoin build, so that's good enough to move on
135 2013-06-10 02:48:08 brson has quit (Quit: leaving)
136 2013-06-10 02:48:24 hnz has quit (Ping timeout: 264 seconds)
137 2013-06-10 02:50:30 Jamesz has joined
138 2013-06-10 02:51:54 hnz has joined
139 2013-06-10 02:52:18 JZavala has quit (Ping timeout: 276 seconds)
140 2013-06-10 02:57:33 <TheUni> warren: branch updated to build win32 if you feel like giving it a shot. this gitian branch needs to build against my autotools branch: https://github.com/theuni/bitcoin/tree/autotools-rfc-1
141 2013-06-10 02:58:10 fiesh has quit (Ping timeout: 255 seconds)
142 2013-06-10 02:58:21 <warren> I'm out of time. I'll try tomorrow if nobody else did.
143 2013-06-10 02:59:07 <TheUni> no worries. thanks.
144 2013-06-10 03:01:44 ixaeon has joined
145 2013-06-10 03:02:05 fiesh has joined
146 2013-06-10 03:02:40 atweiden has quit (Quit: Leaving)
147 2013-06-10 03:03:54 malaimo has quit (Ping timeout: 248 seconds)
148 2013-06-10 03:05:08 ethanbuchman has joined
149 2013-06-10 03:06:04 malaimo has joined
150 2013-06-10 03:06:38 <ethanbuchman> suppose i were foolish enough to try to understand the bit coin source code. where do i even begin?!
151 2013-06-10 03:07:10 <Luke-Jr> Earth
152 2013-06-10 03:07:40 ethanbuchman is now known as work2heat
153 2013-06-10 03:07:43 <work2heat> thats helpful
154 2013-06-10 03:08:20 Subo1978_ has quit (Remote host closed the connection)
155 2013-06-10 03:08:26 <work2heat> seriously tho. I'm a young/new programmer. I mostly write in python. but id love to be able to work through all the open source code out there. i just never know where to start...
156 2013-06-10 03:09:03 Subo1978 has joined
157 2013-06-10 03:10:21 <TheUni> work2heat: come up with a problem, large or small. work til you fix it
158 2013-06-10 03:11:23 <TheUni> sometimes that means doodling a new icon and figuring out where to stick it. sometimes it means understanding the entire codebase to come up with the correct one-line fix
159 2013-06-10 03:12:03 <work2heat> but i don't even know where things are. i scroll through the bitcoin github time to time thinking "now I'll understand something!", but alas, I have yet to. is there documentation somewhere or what? how are others expected to start working it?
160 2013-06-10 03:12:23 <TheUni> git grep
161 2013-06-10 03:13:21 <work2heat> touche!
162 2013-06-10 03:13:30 <Luke-Jr> work2heat: the RPC code is pretty nicely modularized now
163 2013-06-10 03:13:33 <Luke-Jr> rpc*.cpp
164 2013-06-10 03:13:42 <work2heat> whats that
165 2013-06-10 03:14:00 <TheUni> work2heat: to be clear, i'm an outsider as well. but i've been involved in several big projects, and that's how i start
166 2013-06-10 03:14:22 <work2heat> "remote procedure call"?
167 2013-06-10 03:14:28 <Luke-Jr> work2heat: seriously, I doubt you'll find anyone who knows it who is willing to hold your hand. LOOK. IT'S RIGHT IN FRONT OF YOU.
168 2013-06-10 03:14:46 <work2heat> just by greppin for whatever u feel like working on huh
169 2013-06-10 03:14:59 <work2heat> not looking for hand holding. just pointers, tips,
170 2013-06-10 03:15:15 <Luke-Jr> I gave you a pointer.
171 2013-06-10 03:16:24 <work2heat> perhaps all that python has made me ignorant to pointers ;)
172 2013-06-10 03:16:28 <work2heat> :P
173 2013-06-10 03:21:32 agricocb has joined
174 2013-06-10 03:22:59 slyda has joined
175 2013-06-10 03:27:52 shesek has quit (Ping timeout: 255 seconds)
176 2013-06-10 03:32:17 zer0def has quit (Ping timeout: 240 seconds)
177 2013-06-10 03:35:23 Zoop_ has quit (Ping timeout: 264 seconds)
178 2013-06-10 03:36:23 nus has quit (Read error: Connection reset by peer)
179 2013-06-10 03:37:04 nus has joined
180 2013-06-10 03:37:39 Zoop_ has joined
181 2013-06-10 03:39:36 whiterabbit has joined
182 2013-06-10 03:40:22 wrabbit has quit (Ping timeout: 240 seconds)
183 2013-06-10 03:40:25 whiterabbit is now known as wrabbit
184 2013-06-10 03:40:45 EasyAt_ is now known as EasyAt
185 2013-06-10 03:41:02 EasyAt has quit (Changing host)
186 2013-06-10 03:41:02 EasyAt has joined
187 2013-06-10 03:43:02 work2heat has quit (Quit: work2heat)
188 2013-06-10 03:47:36 shesek has joined
189 2013-06-10 03:51:24 phma has quit (Remote host closed the connection)
190 2013-06-10 03:51:49 phma has joined
191 2013-06-10 03:56:38 patcon has quit (Remote host closed the connection)
192 2013-06-10 03:56:38 grau has joined
193 2013-06-10 03:56:43 <owowo> why is my bitcoind showing me "offline, has not been successfully broadcast yet" ?
194 2013-06-10 03:56:49 <owowo> *bitcoin-qt
195 2013-06-10 04:00:43 <Diablo-D3> because you're not connected to other peers?
196 2013-06-10 04:01:32 <owowo> 8 active connections...
197 2013-06-10 04:01:36 da2ce7 has quit (Ping timeout: 264 seconds)
198 2013-06-10 04:01:57 buddyrandom has joined
199 2013-06-10 04:02:28 fanquake has joined
200 2013-06-10 04:02:38 erska has quit (Ping timeout: 248 seconds)
201 2013-06-10 04:03:02 TheSeven has quit (Disconnected by services)
202 2013-06-10 04:03:11 [7] has joined
203 2013-06-10 04:05:16 ixaeon has left ()
204 2013-06-10 04:05:17 <Diablo-D3> owowo: were you connected and up to date when you sent that tx?
205 2013-06-10 04:06:52 <owowo> connected yes, but not up to date, but not being up to date never was a problem, The tx just showed 0 conf until it was up to date, but it was at least broadcast
206 2013-06-10 04:07:20 slyda has quit (Ping timeout: 252 seconds)
207 2013-06-10 04:07:35 <owowo> it's not even showing ob blockexplorer
208 2013-06-10 04:09:24 debiantoruser has quit (Ping timeout: 264 seconds)
209 2013-06-10 04:10:30 erska has joined
210 2013-06-10 04:10:53 debiantoruser has joined
211 2013-06-10 04:11:18 slyda has joined
212 2013-06-10 04:13:30 patcon has joined
213 2013-06-10 04:16:22 k9quaint has joined
214 2013-06-10 04:16:30 skeledrew has quit (Read error: Connection reset by peer)
215 2013-06-10 04:18:50 <owowo> nvm it got broadcast
216 2013-06-10 04:20:01 FabianB has joined
217 2013-06-10 04:20:01 FabianB has quit (Changing host)
218 2013-06-10 04:20:01 FabianB has joined
219 2013-06-10 04:20:58 FabianB_ has quit (Ping timeout: 255 seconds)
220 2013-06-10 04:23:35 nx201_ has quit (Ping timeout: 256 seconds)
221 2013-06-10 04:24:48 Thepok has joined
222 2013-06-10 04:25:55 buddyrandom has quit (Quit: Leaving)
223 2013-06-10 04:27:31 slyda has quit ()
224 2013-06-10 04:27:36 <phantomcircuit> ;;bc,blocks
225 2013-06-10 04:27:36 <gribble> 240693
226 2013-06-10 04:30:57 enquirer has joined
227 2013-06-10 04:32:01 shesek has quit (Ping timeout: 246 seconds)
228 2013-06-10 04:33:22 macboz_ has joined
229 2013-06-10 04:35:55 <Vinnie_win> PHEW price coming back
230 2013-06-10 04:35:58 <Vinnie_win> fucking spam emails
231 2013-06-10 04:36:06 macboz has quit (Ping timeout: 245 seconds)
232 2013-06-10 04:37:49 <owowo> exactly, mate ;o)
233 2013-06-10 04:44:28 sark has quit (!~sark@37-252-106-157.ip.skylogicnet.com|Read error: Connection timed out)
234 2013-06-10 04:45:01 RazielZ has joined
235 2013-06-10 04:45:58 sark has joined
236 2013-06-10 04:46:14 Namworld has quit ()
237 2013-06-10 04:49:47 Julius129 has quit ()
238 2013-06-10 04:49:59 Eiii has quit (Remote host closed the connection)
239 2013-06-10 04:51:18 Eiii has joined
240 2013-06-10 04:51:18 Eiii has quit (Changing host)
241 2013-06-10 04:51:18 Eiii has joined
242 2013-06-10 04:53:50 jaequery has joined
243 2013-06-10 04:56:18 shesek has joined
244 2013-06-10 04:56:58 debiantoruser has quit (Ping timeout: 248 seconds)
245 2013-06-10 04:58:54 debiantoruser has joined
246 2013-06-10 05:00:19 Marc____ has left ()
247 2013-06-10 05:01:23 porquilho has quit ()
248 2013-06-10 05:01:45 defunctzombie_zz is now known as defunctzombie
249 2013-06-10 05:07:18 defunctzombie is now known as defunctzombie_zz
250 2013-06-10 05:08:34 brwyatt is now known as brwyatt|Away
251 2013-06-10 05:09:08 Eiii has quit ()
252 2013-06-10 05:09:42 judah has joined
253 2013-06-10 05:10:31 ThomasV has joined
254 2013-06-10 05:11:11 phebus has quit (Ping timeout: 252 seconds)
255 2013-06-10 05:11:59 debiantoruser has quit (Ping timeout: 264 seconds)
256 2013-06-10 05:12:23 judah has quit (Client Quit)
257 2013-06-10 05:12:48 phebus has joined
258 2013-06-10 05:13:48 debiantoruser has joined
259 2013-06-10 05:14:44 daktak_ has quit (Quit: No Ping reply in 180 seconds.)
260 2013-06-10 05:17:20 nx201 has joined
261 2013-06-10 05:19:37 judah has joined
262 2013-06-10 05:20:54 fanquake has quit (Quit: fanquake)
263 2013-06-10 05:25:06 fanquake has joined
264 2013-06-10 05:31:22 grau has quit (Remote host closed the connection)
265 2013-06-10 05:32:18 grau has joined
266 2013-06-10 05:33:21 Eiii has joined
267 2013-06-10 05:33:21 Eiii has quit (Changing host)
268 2013-06-10 05:33:21 Eiii has joined
269 2013-06-10 05:41:33 metabyte has quit (Read error: Connection reset by peer)
270 2013-06-10 05:41:54 metabyte has joined
271 2013-06-10 05:42:13 RoboTedd_ has joined
272 2013-06-10 05:42:53 px has joined
273 2013-06-10 05:44:07 RoboTeddy has quit (Ping timeout: 246 seconds)
274 2013-06-10 05:44:38 skeledrew has joined
275 2013-06-10 05:56:31 Eiii has quit (Ping timeout: 245 seconds)
276 2013-06-10 05:58:11 Thepok has quit (Ping timeout: 264 seconds)
277 2013-06-10 06:03:35 zacm has joined
278 2013-06-10 06:05:08 jaequery has quit (Quit: Textual IRC Client: www.textualapp.com)
279 2013-06-10 06:06:18 hnz has quit (Ping timeout: 248 seconds)
280 2013-06-10 06:08:17 <shesek> when preparing a transaction for hashing/signing, what script should I use for p2sh inputs? the original hashed output script (HASH_160 ... OP_EQUAL) or the original pre-hashed script?
281 2013-06-10 06:08:56 <shesek> (I'm referring to bitcoinj[s]'s hashTransactionForSignature connectedScript argument)
282 2013-06-10 06:10:40 jaequery has joined
283 2013-06-10 06:11:52 hnz has joined
284 2013-06-10 06:13:54 macboz_ has quit (Quit: Leaving)
285 2013-06-10 06:15:51 fourlights_ has joined
286 2013-06-10 06:16:38 fourlights has quit (Ping timeout: 252 seconds)
287 2013-06-10 06:17:13 defunctzombie_zz is now known as defunctzombie
288 2013-06-10 06:20:12 shesek has quit (Ping timeout: 264 seconds)
289 2013-06-10 06:26:01 dave4925 has quit (Ping timeout: 256 seconds)
290 2013-06-10 06:26:42 defunctzombie is now known as defunctzombie_zz
291 2013-06-10 06:28:13 nowan_ has joined
292 2013-06-10 06:30:33 tyn has joined
293 2013-06-10 06:31:36 nowan has quit (Ping timeout: 264 seconds)
294 2013-06-10 06:32:19 macboz has joined
295 2013-06-10 06:41:52 agnostic98 has quit (Read error: Connection reset by peer)
296 2013-06-10 06:42:02 agnostic98 has joined
297 2013-06-10 06:42:53 shesek has joined
298 2013-06-10 06:47:16 defunctzombie_zz is now known as defunctzombie
299 2013-06-10 06:50:22 ThomasV has quit (Ping timeout: 255 seconds)
300 2013-06-10 06:53:41 <sipa> TheUni: which branch?
301 2013-06-10 06:55:26 defunctzombie is now known as defunctzombie_zz
302 2013-06-10 06:56:06 <TheUni> sipa: gitian-cleanup. 1 sec though, just spotted a prob
303 2013-06-10 06:57:18 ivan`_ is now known as ivan`
304 2013-06-10 06:58:53 <sipa> TheUni: that does not include the autotools changes?
305 2013-06-10 06:59:21 <TheUni> sipa: no. i thought you could have gitian use discriptors for another branch?
306 2013-06-10 06:59:25 <TheUni> need me to merge em?
307 2013-06-10 07:00:01 <sipa> TheUni: yes, please
308 2013-06-10 07:00:27 <sipa> (i have a wrapper script around all things gitian, and it just uses the descriptors from the branch you're building, wherever it is)
309 2013-06-10 07:00:28 <TheUni> sipa: done. note the deps bumps for boost and -deps
310 2013-06-10 07:01:03 defunctzombie_zz is now known as defunctzombie
311 2013-06-10 07:01:05 <TheUni> i haven't fixed up the main one yet, since it should be trivial. i'm only messing with windows atm
312 2013-06-10 07:01:58 <sipa> yeah those are likely the biggest hurdle
313 2013-06-10 07:02:20 zer0def has joined
314 2013-06-10 07:03:33 defunctzombie is now known as defunctzombie_zz
315 2013-06-10 07:03:58 <sipa> ah, no need to rebuild qt?
316 2013-06-10 07:04:17 <TheUni> no, it was ok
317 2013-06-10 07:04:53 <sipa> ok, rebuilding deps boost and win32
318 2013-06-10 07:05:12 <TheUni> great, thanks for helping out :)
319 2013-06-10 07:05:16 <sipa> i have to leave now, i'll report back this evening, but the result will likely be "machine crashed" :p
320 2013-06-10 07:05:16 Thepok has joined
321 2013-06-10 07:05:59 Nerevar has joined
322 2013-06-10 07:06:19 <TheUni> heh
323 2013-06-10 07:07:01 Diapolis has quit (Read error: Connection reset by peer)
324 2013-06-10 07:07:30 Diapolis has joined
325 2013-06-10 07:09:13 Diapolis_ has joined
326 2013-06-10 07:09:19 Diapolis has quit (Read error: Connection reset by peer)
327 2013-06-10 07:17:17 Insty has quit (Ping timeout: 240 seconds)
328 2013-06-10 07:21:33 Insty has joined
329 2013-06-10 07:25:17 LainZ has quit (Ping timeout: 240 seconds)
330 2013-06-10 07:25:29 toffoo has quit ()
331 2013-06-10 07:28:33 tyn has quit (Ping timeout: 276 seconds)
332 2013-06-10 07:28:33 <shesek> Can someone please help me a bit with some p2sh/multisig issues?
333 2013-06-10 07:28:47 da2ce7 has joined
334 2013-06-10 07:30:12 oleganza has joined
335 2013-06-10 07:30:16 Eiii has joined
336 2013-06-10 07:30:16 Eiii has quit (Changing host)
337 2013-06-10 07:30:16 Eiii has joined
338 2013-06-10 07:32:32 saulimus has joined
339 2013-06-10 07:32:53 agnostic98 has quit (Remote host closed the connection)
340 2013-06-10 07:33:26 agnostic98 has joined
341 2013-06-10 07:33:52 jaequery has quit (Quit: Computer has gone to sleep.)
342 2013-06-10 07:34:38 jaequery has joined
343 2013-06-10 07:35:42 agnostic98 has quit (Read error: Connection reset by peer)
344 2013-06-10 07:36:13 melvster has joined
345 2013-06-10 07:37:44 grau has quit (Remote host closed the connection)
346 2013-06-10 07:38:51 jaequery has quit (Client Quit)
347 2013-06-10 07:39:33 grau has joined
348 2013-06-10 07:39:45 ericmuys_ has quit (Remote host closed the connection)
349 2013-06-10 07:39:59 oleganza has quit (Quit: oleganza)
350 2013-06-10 07:40:22 Subo1978_ has joined
351 2013-06-10 07:40:38 da2ce7 has quit (Quit: KVIrc 4.2.0 Equilibrium http://www.kvirc.net/)
352 2013-06-10 07:40:42 Subo1978 has quit (Ping timeout: 240 seconds)
353 2013-06-10 07:45:44 ericmuyser has joined
354 2013-06-10 07:47:11 CodesInChaos has joined
355 2013-06-10 07:48:03 Eiii has quit (Ping timeout: 276 seconds)
356 2013-06-10 08:00:32 BlackPrapor has joined
357 2013-06-10 08:04:06 agnostic98 has joined
358 2013-06-10 08:04:14 panzerfaust is now known as panzer
359 2013-06-10 08:10:59 agnostic98 has quit (Read error: Connection reset by peer)
360 2013-06-10 08:12:06 grau has quit (Remote host closed the connection)
361 2013-06-10 08:13:15 nikolaj has joined
362 2013-06-10 08:15:41 grau has joined
363 2013-06-10 08:15:47 viperhr has joined
364 2013-06-10 08:21:05 Nerevar has quit (Remote host closed the connection)
365 2013-06-10 08:24:26 viperhr has quit (Ping timeout: 255 seconds)
366 2013-06-10 08:24:43 t7 has joined
367 2013-06-10 08:29:34 ByronJohnson has quit (Ping timeout: 256 seconds)
368 2013-06-10 08:35:58 macboz_ has joined
369 2013-06-10 08:36:16 FredEE has joined
370 2013-06-10 08:38:32 macboz has quit (Ping timeout: 252 seconds)
371 2013-06-10 08:39:23 agnostic98 has joined
372 2013-06-10 08:42:53 agnostic98 has quit (Read error: Connection reset by peer)
373 2013-06-10 08:45:57 paracyst has quit ()
374 2013-06-10 08:47:49 msvb-lab has joined
375 2013-06-10 08:49:05 ByronJohnson has joined
376 2013-06-10 08:50:41 Neozonz has quit (Ping timeout: 245 seconds)
377 2013-06-10 08:55:22 ovidiusoft has joined
378 2013-06-10 08:59:13 stalled has quit (Ping timeout: 260 seconds)
379 2013-06-10 09:02:31 Neozonz has joined
380 2013-06-10 09:02:31 Neozonz has quit (Changing host)
381 2013-06-10 09:02:32 Neozonz has joined
382 2013-06-10 09:04:12 epscy has joined
383 2013-06-10 09:09:05 fourlights has joined
384 2013-06-10 09:09:57 fourlights_ has quit (Ping timeout: 276 seconds)
385 2013-06-10 09:10:36 debiantoruser has quit (Ping timeout: 276 seconds)
386 2013-06-10 09:11:51 debiantoruser has joined
387 2013-06-10 09:14:30 Jackneill has joined
388 2013-06-10 09:14:30 Jackneill has quit (Changing host)
389 2013-06-10 09:14:30 Jackneill has joined
390 2013-06-10 09:19:15 FredEE has quit (Quit: FredEE)
391 2013-06-10 09:20:02 ericmuys_ has joined
392 2013-06-10 09:20:09 ericmuyser has quit (Read error: Connection reset by peer)
393 2013-06-10 09:27:05 stalled has joined
394 2013-06-10 09:27:21 ericmuys_ has quit (Read error: Connection reset by peer)
395 2013-06-10 09:27:22 B0g4r7_ has quit (Ping timeout: 248 seconds)
396 2013-06-10 09:27:31 ericmuyser has joined
397 2013-06-10 09:32:45 B0g4r7_ has joined
398 2013-06-10 09:33:18 GordonG3kko has quit (Remote host closed the connection)
399 2013-06-10 09:34:10 GordonG3kko has joined
400 2013-06-10 09:39:02 debiantoruser has quit (Ping timeout: 252 seconds)
401 2013-06-10 09:40:58 debiantoruser has joined
402 2013-06-10 09:46:24 ericmuys_ has joined
403 2013-06-10 09:46:32 Arnavion has quit (Ping timeout: 245 seconds)
404 2013-06-10 09:47:01 ericmuyser has quit (Read error: Connection reset by peer)
405 2013-06-10 09:47:39 Arnavion has joined
406 2013-06-10 09:47:42 <warren> TheUni: got anyone to test your gitian build yet? let me know, I'll be back soon.
407 2013-06-10 09:49:56 <sipa> i tried, but it failed building boost apparently
408 2013-06-10 09:50:10 wamatt_ has quit (Quit: wamatt_)
409 2013-06-10 09:50:55 wamatt has joined
410 2013-06-10 09:51:06 ericmuys_ has quit (Ping timeout: 245 seconds)
411 2013-06-10 09:51:45 <melvster> sipa: you disagree with "Bitcoin's fundamental philosophy was one CPU one vote"
412 2013-06-10 09:55:40 <owowo> now it's more like one asic one vote and the one with the most asic has most of the votes
413 2013-06-10 09:56:11 <melvster> lol
414 2013-06-10 09:57:22 <sipa> melvster: i agree with it, but only for ordering transactions
415 2013-06-10 09:57:31 <sipa> melvster: absolutely not for choosing the rules
416 2013-06-10 09:57:32 <BlueMatt> melvster: ummm...that was never the case
417 2013-06-10 09:57:46 BlackPrapor has quit (Ping timeout: 248 seconds)
418 2013-06-10 09:57:56 <melvster> sipa: what mechanism would you suggest for choosing the rules?
419 2013-06-10 09:58:05 <BlueMatt> the one we have now?
420 2013-06-10 09:58:12 <sipa> melvster: there is only one: consensus
421 2013-06-10 09:58:20 <sipa> how to achieve that is something else
422 2013-06-10 09:58:24 <BlueMatt> developer consensus followed by request for comments and eventual consensus among most users
423 2013-06-10 09:58:35 <melvster> consensus is extremely easy to game, especially in a young community
424 2013-06-10 09:58:36 <sipa> but without very widespread consensus among all affected parties, doing a hardfork is suicide
425 2013-06-10 09:59:02 <TheUni> sipa: warren's built boost without issie afaik
426 2013-06-10 09:59:05 ericmuyser has joined
427 2013-06-10 09:59:22 <sipa> TheUni: yes, most likely a problem on my side
428 2013-06-10 09:59:42 <melvster> this is quite a good document about consensus:
429 2013-06-10 09:59:42 <melvster> http://datatracker.ietf.org/doc/draft-resnick-on-consensus/?include_text=1
430 2013-06-10 10:00:08 <sipa> melvster: and yes, consensus is a fragile and hard to define thing, when the boundaries of the community aren't well established in the first place
431 2013-06-10 10:00:15 <melvster> "We reject: kings, presidents and voting. We believe in: rough consensus and running code."
432 2013-06-10 10:01:04 <sipa> melvster: i'm actually worried about the fact that so many people just accept some changes without even stopping and saying "wowowow how was this decided?", even if they'd agree with it
433 2013-06-10 10:01:52 <melvster> sipa: I met someone from a very large company last year, at a big internet meeting, he told me that there were only 7 people in the standards world that they could not buy, and that they knew every single one of them, he also said there was no standard they could not torpedo ... consensus is very easy to game
434 2013-06-10 10:01:54 <BlueMatt> sipa: Im not sure that problem has a solution, sadly
435 2013-06-10 10:02:11 <sipa> BlueMatt: indeed
436 2013-06-10 10:03:12 <sipa> melvster: but consensus about changing the network rules in bitcoin means you need to convince _everyone_ to upgrade their nodes
437 2013-06-10 10:03:16 <melvster> sipa: it's a well known phenomenon that people will accept defaults most of the time
438 2013-06-10 10:04:32 <melvster> sipa: I agree with what you're saying, but one wrong rule change and bitcoin could be toast ... ok we still have litecoin, but im saying there's a risk ... so caution should be the default stance
439 2013-06-10 10:04:36 mE\Ta has quit (Ping timeout: 264 seconds)
440 2013-06-10 10:04:43 <sipa> melvster: it is
441 2013-06-10 10:04:51 <sipa> melvster: without consensus, rules do not change :)
442 2013-06-10 10:05:46 <melvster> sipa: what's your personal view on the block size limit?
443 2013-06-10 10:06:15 <sipa> melvster: i'm in favor of increasing it using a schedule that tries to match technological evolution
444 2013-06-10 10:06:26 <melvster> re the document above, the IETF rejects voting, because you can stuff votes, they "hum" instead to get a feeling for opinion on an issue ...
445 2013-06-10 10:06:52 <sipa> that's imho the only thing that matters; "we want to do more transactions" is a bad reason imho
446 2013-06-10 10:06:52 <melvster> not that the IETF is perfect!
447 2013-06-10 10:06:55 <melvster> far from it
448 2013-06-10 10:07:39 <petertodd> melvster: bitcoin has serious debate over even which technological evolutions are relevant - I seriously doubt that IETF style consensus is really adequate
449 2013-06-10 10:08:18 <melvster> petertodd: im not saying it is, but the take away is that voting is a BAD way to reach consensus
450 2013-06-10 10:08:29 <BlueMatt> melvster: no one ever said we vote
451 2013-06-10 10:08:30 <melvster> votes can easily be manipulated in a small group
452 2013-06-10 10:08:44 <sipa> i'd be absolutely against voting
453 2013-06-10 10:08:49 <BlueMatt> there has never been a technical feature on which developers voted, nor will there be
454 2013-06-10 10:09:00 <petertodd> melvster: voting is a perfectly good way of achieving consensus *if* the consensus is that you accept the results of the vote
455 2013-06-10 10:09:07 <sipa> except when it's required to have a (near) unanymous outcome
456 2013-06-10 10:09:38 <melvster> petertodd: sure but in technical standards and protocols, it's poor, the w3c vote on all rec, and luckily very few are contentious
457 2013-06-10 10:10:14 <petertodd> melvster: right, and john's not talking about voting for a technical standard or protocol, john's talking about voting about what we want Bitcoin to be
458 2013-06-10 10:10:26 <BlueMatt> melvster: it appears you are suggesting alternates to a system we dont use because you think we vote on everything...we dont, and Im not sure what you're complaining about...
459 2013-06-10 10:11:27 <melvster> BlueMatt: context was on the voting proposal presented on the mail list, I'm just looking at sipa's comments
460 2013-06-10 10:12:06 <BlueMatt> yes, $SOMEONE suggested we vote, and I dont really see any agreement on it...so.....
461 2013-06-10 10:13:15 <sipa> melvster: note that i didn't actually comment about the proof-of-stake-voting proposal itself; i'd have to think more about that
462 2013-06-10 10:13:26 <sipa> melvster: just trying to comment on some misconceptions
463 2013-06-10 10:13:35 <melvster> sipa: got it, thanks :)
464 2013-06-10 10:14:11 <melvster> just wanted to check with you before replying incorrectly, i wanted to understand your answer, thanks for explaining
465 2013-06-10 10:15:07 <BlueMatt> having some high-stake users provide statements on what they want bitcoin to be would be informative, but making that a binding vote.....
466 2013-06-10 10:15:30 <melvster> i think one guy owns 10% plus of all btc
467 2013-06-10 10:15:42 <petertodd> BlueMatt: well, this is in the context of Gavin saying miners should decide the issue completely
468 2013-06-10 10:16:18 <BlueMatt> petertodd: I was under the impression it would be a limit on how much miners are allowed to swing the size per block
469 2013-06-10 10:16:36 <sipa> i'm not sure i'd like that
470 2013-06-10 10:16:56 <sipa> but the most concrete thing being said by now, is that we first need data about what block sizes the network can tolerate
471 2013-06-10 10:16:57 <petertodd> BlueMatt: The way I read it is the vote is to determine the upper limit, because we know miners can set the limit below the upper limit to whatever they want it to be.
472 2013-06-10 10:17:01 <BlueMatt> plus theres a propensity for people to jump up and down and say "over 1MB kills all anonymity" when, in fact, that is false
473 2013-06-10 10:17:30 <petertodd> sipa: but, what is "the network", over 1MB quickly kills anonymity, yet has little effect for many other operators
474 2013-06-10 10:17:39 <sipa> it's true that _any_ increase hurts anonimity to some extent
475 2013-06-10 10:17:44 <petertodd> sipa: (anonymity for miners of course)
476 2013-06-10 10:17:51 <sipa> the degree to which this is acceptable is something else
477 2013-06-10 10:17:57 <melvster> so there seem to be 3 conversations 1) is it reasonable to increase block size with moore's law (note moore's law is already factored in to some pieces of btc) 2. Do we want more tx 3. Was the 1MB limit well thought out in the first place ...
478 2013-06-10 10:18:09 <petertodd> sipa: yes, potentially we'd see stakeholders actually vote for a decrease in jdillon's proposal
479 2013-06-10 10:18:34 <sipa> and before we can even start talking about that, we need to know the technical impact of block sizes
480 2013-06-10 10:18:38 Maxvalor has quit ()
481 2013-06-10 10:18:39 <sipa> which we don't even know now
482 2013-06-10 10:18:59 <petertodd> sipa: it's strongly dominated by bandwidth, that's dead easy to determine
483 2013-06-10 10:19:04 * BlueMatt votes for a ban on discussion of block size increase unless new information is being proposed
484 2013-06-10 10:19:15 chorao has quit (Read error: Connection reset by peer)
485 2013-06-10 10:19:16 <BlueMatt> s/proposed/presented/
486 2013-06-10 10:19:37 chorao has joined
487 2013-06-10 10:19:37 chorao has quit (Changing host)
488 2013-06-10 10:19:37 chorao has joined
489 2013-06-10 10:19:46 <petertodd> BlueMatt: I dunno, a clever and thought-provoking voting mechanism seems new to me
490 2013-06-10 10:19:49 <melvster> 4) anonymity concerns ... I guess
491 2013-06-10 10:19:52 <sipa> petertodd: in theory
492 2013-06-10 10:20:55 <melvster> btw this was ben laurie's criticism of bitcoin
493 2013-06-10 10:21:15 <BlueMatt> petertodd: and yet we still have a dearth of real technical details on exactly what effect it would have
494 2013-06-10 10:21:17 <petertodd> sipa: I've done the Tor measurements, and it's easy to get info on non-tor methods like radio - it's not rocket science
495 2013-06-10 10:21:21 ovidiusoft has quit (Quit: leaving)
496 2013-06-10 10:21:22 <BlueMatt> so...lets get everyone to vote blind :)
497 2013-06-10 10:21:46 <petertodd> BlueMatt: if you see anonymity as important you don't need any more technical details than bandwidth, it's that limiting
498 2013-06-10 10:22:00 <BlueMatt> umm....no
499 2013-06-10 10:22:03 <sipa> petertodd: please
500 2013-06-10 10:23:06 <petertodd> Look, it's a 1MB block, you need to fill that up with transactions to stay competitive, and those transactions have no guarantee of being in anyone elses mempools. So push 1MB of data through your link fast enough to keep your orphan rate low.
501 2013-06-10 10:23:37 <BlueMatt> instead of arguing, why not write code to solve that problem?
502 2013-06-10 10:23:43 <BlueMatt> (because its easily doable...)
503 2013-06-10 10:23:58 <petertodd> Because that's impossible. The issue of the transactions not being in others mempools isn't solvable.
504 2013-06-10 10:24:01 <BlueMatt> eg...p2pool which does some of it
505 2013-06-10 10:24:03 <sipa> there are tons of technical improvements possible
506 2013-06-10 10:24:12 <sipa> but it's not solvable completely in theory
507 2013-06-10 10:24:16 <petertodd> It's not solvable in a zero-trust situation.
508 2013-06-10 10:24:21 <petertodd> p2pool is *not* zero trust
509 2013-06-10 10:24:27 rdymac has joined
510 2013-06-10 10:24:28 rdymac has quit (Changing host)
511 2013-06-10 10:24:28 rdymac has joined
512 2013-06-10 10:25:05 <BlueMatt> you can improve it a lot (in +/- zero trust)
513 2013-06-10 10:25:08 <petertodd> I know damn well how p2pool works, and applied to Bitcoin in general that mechanism is dangerous and gives ways to change the best case and worst case block propagation.
514 2013-06-10 10:25:32 <BlueMatt> until I see a cubesat broadcasting blocks, I dont really care about being able to mine over your mom's dial-up connection
515 2013-06-10 10:25:32 <petertodd> BlueMatt: Yes, you can improve it by adding trust. That's unacceptable to me.
516 2013-06-10 10:25:47 <sipa> petertodd: I guess my point is that there are still tons of technical improvements possible
517 2013-06-10 10:25:49 <petertodd> BlueMatt: cubecat's are not decentralized infrastructure
518 2013-06-10 10:25:59 <petertodd> sipa: But that's it, there aren't.
519 2013-06-10 10:26:06 <melvster> petertodd: why is adding trust unacceptable?
520 2013-06-10 10:26:12 <BlueMatt> then you are being deliberately ignorant
521 2013-06-10 10:26:30 <BlueMatt> also: tor is neither anonymous of decentralized in your view
522 2013-06-10 10:26:32 <petertodd> melvster: Adding trust means someone else can start attacking those Tor connected miners - it's bad enough we have the block-withholding attack with pools.
523 2013-06-10 10:27:16 <petertodd> BlueMatt: It's the best we've got, and it's representative of alternatives that are better. In particular Bitcoin has a serious problem with traffic analysis, so you want to be able to have your bandwidth usage == the usage you upload when you build a new block.
524 2013-06-10 10:27:55 * BlueMatt -> more productive things
525 2013-06-10 10:28:05 <petertodd> But, again, this discussion shows how the issues are not technical, they are what we think is valuable, which is why I reluctantly support John's proposal.
526 2013-06-10 10:28:44 <sipa> petertodd: if your view is "increasing the block size hurts anonymity, so it shouldn't be done", you're not being realistic; it's a compromise in any case, and it's all about which point is acceptible there
527 2013-06-10 10:29:11 <sipa> petertodd: and before we can talk about what is acceptable, we need to know what becomes the limiting factor given the current technology
528 2013-06-10 10:29:17 <BlueMatt> ;;slap petertodd "the above technical discussion"
529 2013-06-10 10:29:18 * gribble slaps petertodd the above technical discussion with a whipping up banjorine
530 2013-06-10 10:29:21 <petertodd> sipa: No, my view is more subtle than that. It's that by increasing the blocksize rather than pursuing other solutions, we're not even valuing anonymity and aren't really trying.
531 2013-06-10 10:29:48 <BlueMatt> then go try
532 2013-06-10 10:29:56 <BlueMatt> btw, bitcoin is oss...you can work on things too
533 2013-06-10 10:30:02 <petertodd> sipa: It means that the next time volume is an issue, we'll just punt and further damage anonymity. John's proposal just means that when that happens at least we're clear that it's the will of the people./
534 2013-06-10 10:30:23 gfinn has quit (Ping timeout: 240 seconds)
535 2013-06-10 10:30:57 <petertodd> BlueMatt: I thought about that really early on, and quickly figured out that actually making those solutions just makes me a target. Much better to convince others that the solutions really are needed.
536 2013-06-10 10:31:11 <sipa> petertodd: that is why imho the only meaningful block increase schedule is one that follows technical evolution
537 2013-06-10 10:31:23 <sipa> petertodd: but before that, we need to find what is acceptable, and we don't know
538 2013-06-10 10:31:37 <sipa> i have no clue right now in what way larger blocks affect propagation delays
539 2013-06-10 10:31:46 <petertodd> sipa: indeed we don't, because we can't even decide who "we" are
540 2013-06-10 10:32:08 <sipa> and i don't know at which point bandwidth becomes the limiting factor there, rather than verification speed/database updating
541 2013-06-10 10:32:27 <petertodd> sipa: and frankly, who cares? transaction volume is an issue of orders of magnitude, micro-optimizations like tx hashes vs. whole tx's don't actually make very big differences
542 2013-06-10 10:32:29 <melvster> petertodd: could you say in a sentence or two the exact issue with anonymity ... does tor prevent you sending blocks over a certain size?
543 2013-06-10 10:33:17 filleokus has quit (Excess Flood)
544 2013-06-10 10:33:31 <petertodd> melvster: ways that provide anonymous and/or decentralized bandwidth, and tor *is just one example*, do not follow moores law for a lot of reasons. you need to upload a block fast enough and propagate it fast enough to keep your orphan rate down to stay profitable as a miner
545 2013-06-10 10:33:51 <petertodd> melvster: without anonymous mining, Bitcoin is easy to control by tracking down miners and forcing them to implement changes to bitcoin that you want
546 2013-06-10 10:34:15 <melvster> petertodd: thanks
547 2013-06-10 10:34:24 filleokus has joined
548 2013-06-10 10:34:26 <sipa> meh, that's the sort of argument that makes me want to ignore you
549 2013-06-10 10:34:35 <melvster> O_o
550 2013-06-10 10:35:02 <melvster> sipa: you dont think it has merit?
551 2013-06-10 10:35:02 <petertodd> sipa: it's why people don't use PGP :P
552 2013-06-10 10:35:23 <sipa> (and it's not because i disagree with it, just the tone of presenting it)
553 2013-06-10 10:35:33 <BlueMatt> melvster: essentially, the argument here boils down to lots of handwaving and almost no technical reasoning to back it up
554 2013-06-10 10:36:35 <sipa> well, i agree completely up to the point that larger block sizes make anonymous mining harder
555 2013-06-10 10:36:47 Guest81724 has quit (Ping timeout: 248 seconds)
556 2013-06-10 10:36:51 <BlueMatt> I thought that was the whole argument?
557 2013-06-10 10:36:53 <petertodd> sipa: that's part of what appeals to me about john's proposal to be honest, it at least lets us directly vote on what we value
558 2013-06-10 10:37:29 <BlueMatt> it doesnt, because no one actually knows what effect a given value will have...we have very little data
559 2013-06-10 10:37:31 <melvster> petertodd: but you assume that the only stake holders in bitcoin are the bitcoin holders
560 2013-06-10 10:37:42 <petertodd> Of course, at the other end of scaling, at some point running a bitcoin node costs tens of thousands of dollars, and legality aside control winds up in a few hands. By restricting the blocksize we make it clear that there are limits, and there need to be betters solutions than "just make it bigger"
561 2013-06-10 10:37:45 <BlueMatt> so voting at this point would just be a bunch of handwaving and guessing
562 2013-06-10 10:37:51 has joined
563 2013-06-10 10:37:51 has quit (Clown|!Clown@static-87-79-93-140.netcologne.de|Changing host)
564 2013-06-10 10:37:51 has joined
565 2013-06-10 10:38:08 ovidiusoft has joined
566 2013-06-10 10:38:09 <petertodd> BlueMatt: well, voting at this point would do nothing, because we aren't yet at 1MB anyway
567 2013-06-10 10:38:38 <petertodd> BlueMatt: but it does give a clear path, with unclear results
568 2013-06-10 10:38:56 <jouke_> melvster: good point.
569 2013-06-10 10:39:33 <petertodd> melvster: If you don't own any Bitcoins, your involvement in Bitcoin is to provide a service to those who do. Tell them how you want them to vote.
570 2013-06-10 10:39:35 <sipa> petertodd: ok fine, making the acceptable size a decision made by the users of the system, in some way, is nice
571 2013-06-10 10:39:54 <sipa> petertodd: but right now, *I* don't even know what size would be acceptable to *me*
572 2013-06-10 10:40:05 <melvster> petertodd: my whole argument is that im wary of any voting at all
573 2013-06-10 10:40:06 <sipa> how would i expect it to be chosen meaningfully by others?
574 2013-06-10 10:40:15 <petertodd> sipa: yes! it removes a lot of the uglier politics, and also means the decision isn't a one-time thing, you can always change your vote
575 2013-06-10 10:40:28 <melvster> but id like to understand the exact issue and details before commenting further
576 2013-06-10 10:40:44 <melvster> the anonymity case seems quite compelling
577 2013-06-10 10:40:53 <petertodd> sipa: you can always try convincing them, and anyway, it's more reliable than ugly and contentious hard-forks
578 2013-06-10 10:41:14 <melvster> i do agree a contentious hard fork is potential suicide
579 2013-06-10 10:41:17 <sipa> petertodd: so ok, you give a way to make it politically less dangerous - i like that
580 2013-06-10 10:41:29 <BlueMatt> melvster: no one disagrees, but considering no one knows what level you get at any given block size...
581 2013-06-10 10:41:37 <sipa> petertodd: but it's all moot before we have a clear idea about how much it hurts
582 2013-06-10 10:41:46 <sipa> petertodd: yes, bandwidth has an impact, absolutely
583 2013-06-10 10:41:57 <petertodd> sipa: yeah, and jdillon's proposal quite cleverly means that the default, if everyone stops voting, is that the blocksize limit does reduce a bit over time, not back to the original setting, but halfway
584 2013-06-10 10:42:17 Skav has joined
585 2013-06-10 10:42:19 <melvster> in fact probably the easiest way to destroy bitcoin would be to create an issue just contentious enough that it maximizes the probability of a hard fork ... you see this in elections all the time, for example, divide and conquer public opinion ...
586 2013-06-10 10:42:21 <sipa> but if you go to the point of saying "all constant factors don't matter, only look at completely", you disregard reality
587 2013-06-10 10:42:25 <petertodd> sipa: the mechanism makes for a very convincing consensus, while at the same time meaning that just a small group of users still increase the limit a bit
588 2013-06-10 10:42:31 ralphtheninja has joined
589 2013-06-10 10:43:36 MobPhone has quit (Ping timeout: 245 seconds)
590 2013-06-10 10:43:41 <warren> TheUni: I'm back
591 2013-06-10 10:43:47 mrkent has quit (Ping timeout: 240 seconds)
592 2013-06-10 10:43:48 <warren> TheUni: you still need it tested?
593 2013-06-10 10:43:53 <melvster> sipa: is there any place I can read about "zero trust" systems?
594 2013-06-10 10:44:03 <TheUni> warren: would be great if you could
595 2013-06-10 10:44:06 <tumak> irc is zero trust
596 2013-06-10 10:44:07 <warren> TheUni: url again?
597 2013-06-10 10:44:11 <tumak> everything said on irc is true
598 2013-06-10 10:44:19 <petertodd> melvster: we're inventing them as we go along really
599 2013-06-10 10:44:19 <sipa> melvster: zero trust = verify all data you receive
600 2013-06-10 10:44:22 <TheUni> warren: i merged into one branch for easier testing
601 2013-06-10 10:44:30 <warren> lmk
602 2013-06-10 10:44:36 <melvster> got it
603 2013-06-10 10:44:54 MobPhone has joined
604 2013-06-10 10:44:54 <sipa> melvster: i.e., the system does not rely on a trusted party to make some decisions
605 2013-06-10 10:44:57 <TheUni> https://github.com/theuni/bitcoin/tree/gitian-cleanup
606 2013-06-10 10:45:05 <sipa> melvster: which is true for bitcoin, except for the order of otherwise valid transactions
607 2013-06-10 10:45:20 <sipa> (which is decided by miners, and not 'verified')
608 2013-06-10 10:45:29 <tumak> i just wish gitian had a fiat trusted .exe for average shmoes to download :(
609 2013-06-10 10:45:37 <sipa> tumak: lol
610 2013-06-10 10:45:44 <tumak> gitian downloader that is
611 2013-06-10 10:45:54 anarchy5 has joined
612 2013-06-10 10:46:21 * sipa afk
613 2013-06-10 10:46:22 sipa has left ()
614 2013-06-10 10:46:30 macboz_ has quit (Quit: This computer has gone to sleep)
615 2013-06-10 10:46:31 Skav has quit (Ping timeout: 245 seconds)
616 2013-06-10 10:46:36 <petertodd> melvster: what's interesting is there doesn't appear to be any way of achieving distributed zero-trust consensus without proof-of-work
617 2013-06-10 10:47:00 <petertodd> melvster: even proof-of-stake needs proof-of-work under the hood to actually work
618 2013-06-10 10:47:01 <tumak> (or derivations from pow)
619 2013-06-10 10:47:30 <warren> TheUni: interested in gitian for macos?
620 2013-06-10 10:47:32 <melvster> petertodd: yes that is interesting ... although it's not zero, but very close to zero ... there are other mechanisms of trust that work quite well ... ELO rankings in chess for example
621 2013-06-10 10:47:55 <petertodd> melvster: elo rankings? how?
622 2013-06-10 10:48:05 <TheUni> warren: it should just work (tm) after my next round of hacking
623 2013-06-10 10:48:06 <melvster> petertodd: there is a relationship between reputation and trust
624 2013-06-10 10:48:16 owowo has quit (Quit: sayonara)
625 2013-06-10 10:48:22 <melvster> reputation is from observable events, e.g. kasparov beats karpov
626 2013-06-10 10:48:27 <petertodd> melvster: consensus isn't about reputation or trust
627 2013-06-10 10:48:35 <melvster> and it becomes exponentially more difficult to to gain reputation
628 2013-06-10 10:48:39 <melvster> the trust is now a function
629 2013-06-10 10:48:44 <melvster> probability of X beating Y
630 2013-06-10 10:48:47 <melvster> based on reputation
631 2013-06-10 10:48:47 <tumak> consensus is distributed nothing-up-my-sleeve number generator
632 2013-06-10 10:48:51 <petertodd> melvster: your observable event example is a proof-of-work, kasparov did work to bear karpov
633 2013-06-10 10:48:52 <melvster> in ELO it's quite accurate
634 2013-06-10 10:48:57 Namworld has joined
635 2013-06-10 10:49:00 <tumak> basically generate a random number and prove nobody influenced it, independently
636 2013-06-10 10:49:07 <tumak> you can do that only with pow ledger :(
637 2013-06-10 10:49:23 <petertodd> tumak: heh, yes, for proof-of-stake that's the minimum required to make it actually work; I've got a half-finished paper on that exact topic
638 2013-06-10 10:49:38 <melvster> i like PoW .. im just saying it's not the only system ... ripple uses another system entirely ... but that's another matter
639 2013-06-10 10:49:42 <tumak> petertodd: oh, you got the distributed rand gen w/o pow?
640 2013-06-10 10:49:49 <tumak> petertodd: that would be so awesome, not just for POS
641 2013-06-10 10:50:00 <petertodd> melvster: ripple isn't a consensus system, at least ripple that actually works...
642 2013-06-10 10:50:25 <melvster> petertodd: opencoin's redux of ripple uses consensus
643 2013-06-10 10:50:46 <petertodd> tumak: nope, I've got the opposite: PoS that uses PoW as a random beacon/nothing up my sleeve RNG
644 2013-06-10 10:51:06 <coingenuity> question for yall..... how is the bitcoin leveldb setup built?
645 2013-06-10 10:51:15 <coingenuity> normally leveldb needs a LOCK directory
646 2013-06-10 10:51:19 <petertodd> melvster: nope, ripple is just trusting central authority distributed among a bunch of people
647 2013-06-10 10:51:21 <coingenuity> bitcoin...doesn't do that
648 2013-06-10 10:51:25 <tumak> petertodd: bummers, at least good to know my intuition about this relation was valid :)
649 2013-06-10 10:51:45 <melvster> also no one has said how coloured coins fit in to proof of stake ...
650 2013-06-10 10:52:01 <tumak> colored coins is just a hack
651 2013-06-10 10:52:04 <tumak> to have something viable
652 2013-06-10 10:52:07 <TheUni> warren: i'm headed out for a bit, will check backlog
653 2013-06-10 10:52:12 <petertodd> tumak: Yeah, I've far from proven the relationship formally, but that's my intuition too. The best I could do was use timestamping to make it difficult to use PoS to change history after the fact.
654 2013-06-10 10:52:14 <coingenuity> nobody?
655 2013-06-10 10:52:16 <warren> TheUni: k
656 2013-06-10 10:52:18 <tumak> eventually we will just do https://github.com/bitcoinx/colored-coin-tools/wiki/crosschain-p2ptrade
657 2013-06-10 10:52:33 <petertodd> melvster: colored coins are proof of stake
658 2013-06-10 10:53:54 <coingenuity> oh shit, i got it working now
659 2013-06-10 10:53:56 <coingenuity> nvm, thanks
660 2013-06-10 10:54:46 <melvster> i am working on a system where you can put any currency into a bitcoin wallet
661 2013-06-10 10:55:05 <melvster> so a wallet might look like it's worth 0.1 btc, but actually can hold millions of dollars
662 2013-06-10 10:55:12 <tumak> yet another colouring scheme?
663 2013-06-10 10:55:18 <tumak> i think we have like 4 or 5 now
664 2013-06-10 10:55:32 <tumak> excluding petertodd's, because its just too academic for my taste :)
665 2013-06-10 10:55:42 <melvster> tumak: im unsure, but i would be very surprised if my system overlapped with any of the colored coin 'hacks'
666 2013-06-10 10:55:54 PiZZaMaN2K has quit (away!~PiZZaMaN2@host-72-2-137-170.csinet.net|Ping timeout: 276 seconds)
667 2013-06-10 10:55:58 <petertodd> melvster: what you are doing isn't a technical problem and can't be solved with crypo
668 2013-06-10 10:56:01 <petertodd> *crypto
669 2013-06-10 10:56:09 <tumak> melvster: anything making distinctions between tx inputs is colored coin
670 2013-06-10 10:56:09 <melvster> it's simply another distributed ledger of bitcoin:addres -> balance ^^ currency
671 2013-06-10 10:56:17 <tumak> that is formal description, actual algorithm is irrelevant
672 2013-06-10 10:56:43 <petertodd> melvster: solve the problem of making the coins *actually* represent currency first
673 2013-06-10 10:57:37 <melvster> petertodd: it's more oriented at newly minted currencies than legacy ones .. so instead of have 10 alt coins, on 10 networks, one btc address can hold balances in 10 coins
674 2013-06-10 10:57:53 <melvster> orthogonal currencies help to solve double spend
675 2013-06-10 10:58:25 <petertodd> melvster: what do you think makes those newly minted currencies actually mean anything?
676 2013-06-10 10:59:05 <melvster> petertodd: that's another topic :) ...
677 2013-06-10 10:59:28 <petertodd> melvster: well, solve that problem, because the colored coins stuff is the easy part
678 2013-06-10 10:59:34 <coingenuity> so, now that i'm happily reading bitcoin's block leveldb i see that it's just metadata in the form of ����\n�#'���O�n�7h%�.}g�F]�~;X�x C�
679 2013-06-10 10:59:46 <coingenuity> what the hell do i do with this to correlate it to a blkxxxx.dat
680 2013-06-10 10:59:51 <melvster> but if i was to answer, you look at precident of coins that have value ... there seem to be 3 aspects to optimize, 1) liquidity 2) utility 3) trust
681 2013-06-10 11:00:21 <tumak> petertodd: occams razor
682 2013-06-10 11:00:34 <tumak> people want to innovate, but get-rich-quick as well
683 2013-06-10 11:00:41 <tumak> they should just follow satoshi
684 2013-06-10 11:00:54 <tumak> start with altruistic intentions, and hope for the best
685 2013-06-10 11:00:56 <tumak> not vice versa
686 2013-06-10 11:01:12 <melvster> petertodd: im not trying to create an alt coin, im trying to make a framework such that anyone can make an alt coin using a plain old commodity web server, and link it to the block chain, the ripple ledger, GPG, or any public/private key system
687 2013-06-10 11:01:33 <tumak> melvster: dont forget to add lots of buzzwords
688 2013-06-10 11:01:39 <melvster> in this way you can send money from a bitcoin address to GPG key
689 2013-06-10 11:01:39 <tumak> that seems to be certainly the trend
690 2013-06-10 11:01:47 <petertodd> tumak: heh, indeed, I think people get too carried in big ideas and buzzwords rather than the nitty gritty of how the incentives and consensus works
691 2013-06-10 11:01:52 <melvster> tumaK: will try .. but im not a marketing person :)
692 2013-06-10 11:01:59 <tumak> why not use bitcoin oracle in open transactions?
693 2013-06-10 11:02:25 <melvster> im very interested in opentransactions too
694 2013-06-10 11:02:44 <petertodd> melvster: opentransactions is way over hyped
695 2013-06-10 11:04:01 <melvster> I'm a big fan of OT been following it a lot time ... my main motivation is a web integration ... bitcoin is great but it doesnt scale ... the web scales but it doesnt have a currency ... I want to try and solve these two problems
696 2013-06-10 11:04:31 <melvster> I like OT for off chain tx
697 2013-06-10 11:04:31 <tumak> petertodd: well, OT is actually implemented
698 2013-06-10 11:04:35 <tumak> which is bonus
699 2013-06-10 11:04:41 <melvster> i like btc as the reserve currency of the internet
700 2013-06-10 11:04:52 <melvster> i like the web for mass distribution
701 2013-06-10 11:04:56 <tumak> i filter the hype automagically, but otherwise i'm overall satisfied with what OT can do
702 2013-06-10 11:05:01 <tumak> i'd just replace the scripting with lua
703 2013-06-10 11:05:08 <tumak> and use json instead of xml
704 2013-06-10 11:05:13 <petertodd> melvster: go read up on my fidelity bonded banking stuff - if you can't understand how that's an alternative to global consensus and central authority, you don't understand the problem well enough yet
705 2013-06-10 11:05:22 <tumak> but other that, its very well researched
706 2013-06-10 11:05:24 <petertodd> tumak: true, but OT is just a library
707 2013-06-10 11:05:27 <melvster> petertodd: pointer?
708 2013-06-10 11:05:59 <melvster> OT has a few front ends now
709 2013-06-10 11:06:04 <tumak> isnt fidelity bonds just unnecesarily complicated cc tracking? :)
710 2013-06-10 11:06:06 <tumak> </troll>
711 2013-06-10 11:06:28 <petertodd> melvster: https://bitcointalk.org/index.php?topic=146307.0 http://sourceforge.net/mailarchive/message.php?msg_id=30531383 https://bitcointalk.org/index.php?topic=175156.msg2392095#msg2392095
712 2013-06-10 11:06:34 <melvster> thx!
713 2013-06-10 11:06:49 Anduckk has joined
714 2013-06-10 11:06:49 <petertodd> melvster: not to say that fidelity bonded banking actually works in practice, but understand what the intended security model is
715 2013-06-10 11:06:58 <petertodd> tumak: cc tracking?
716 2013-06-10 11:07:45 <tumak> https://github.com/petertodd/trustbits/blob/master/fidelitybond.md
717 2013-06-10 11:07:56 <tumak> it's just cc with ricardian on top of it
718 2013-06-10 11:08:04 <tumak> (it's a good write up, but a bit ahead its time)
719 2013-06-10 11:08:14 <tumak> cc=colored coin
720 2013-06-10 11:08:30 <petertodd> tumak: ah, well, that's just a low-level representation, fidelity bonds are only *really* useful with fraud proofs, which is a whole 'nother issue
721 2013-06-10 11:09:27 <melvster> petertodd: chaumian blinding ... nice :)
722 2013-06-10 11:09:47 <tumak> yeah, chaum is super easy offchain
723 2013-06-10 11:10:02 <tumak> once again, i prefer OT because one can actually read the code
724 2013-06-10 11:10:15 <melvster> opentransactions is just an implementation of the the original true ledger blinding system
725 2013-06-10 11:10:20 <tumak> indeed
726 2013-06-10 11:10:40 <petertodd> melvster: don't get distracted by the chuam stuff... the key thing to understand is under what conditions could a fraud proof work, and it's subtle
727 2013-06-10 11:10:57 <petertodd> melvster: subtle enough that it is easy to have it not work :)
728 2013-06-10 11:11:10 * melvster reading
729 2013-06-10 11:12:08 agnostic98 has joined
730 2013-06-10 11:12:19 <tumak> melvster: it's actually market theory
731 2013-06-10 11:12:35 daybyter has joined
732 2013-06-10 11:12:53 <melvster> tumak: you mean the efficient market theory?
733 2013-06-10 11:12:54 <tumak> the low-level trick is the network will simply burn your born if you're caught scamming
734 2013-06-10 11:13:06 <tumak> nah, fidelity bonds to prevent fraud
735 2013-06-10 11:13:41 <petertodd> tumak: yeah, and one way of "burning a bond" is to simply consider it useless... which has a lot of subtle issues
736 2013-06-10 11:13:55 <petertodd> tumak: with scripting support in Bitcoin it's much easier, because the network itself can evaluate a fraud proof
737 2013-06-10 11:13:58 <tumak> it's an odd cross between MAD and escrow, but far superior than latter two assuming the market emerges :)
738 2013-06-10 11:14:28 <tumak> s/born/bond/
739 2013-06-10 11:14:35 <tumak> your first born perhaps too
740 2013-06-10 11:14:43 <petertodd> tumak: lol
741 2013-06-10 11:15:42 agnostic98 has quit (Read error: Connection reset by peer)
742 2013-06-10 11:18:45 <melvster> my idea is slightly different
743 2013-06-10 11:19:06 <melvster> but fidelity bonds looks interesting ... it's going to take me a bit more time to digest (no puns intended :))
744 2013-06-10 11:19:26 <tumak> melvster: dont worry i had to learn my lesson at first too
745 2013-06-10 11:19:33 <tumak> the first rule of bitcoin-dev should be
746 2013-06-10 11:19:35 <tumak> LURK MOAR
747 2013-06-10 11:19:40 <melvster> lol
748 2013-06-10 11:19:53 <melvster> my idea is REVERSE looking of hashes
749 2013-06-10 11:19:57 <tumak> if you have a great idea, chances are there are 5 similiar, much better developed, and some even implemented
750 2013-06-10 11:20:06 <melvster> so normally you get a data structure and a hash together
751 2013-06-10 11:20:08 <petertodd> Yeah, think about it some more. I may be wrong, but I'm pretty sure it's essentially one of the only design options available for distributed consensus.
752 2013-06-10 11:20:09 <tumak> people are focusing too much on the pow :(
753 2013-06-10 11:20:26 <melvster> but what if you just got the hash, and could lookup the data structure on demand?
754 2013-06-10 11:20:31 <petertodd> Basically fidelity bonds are your "consensus, but punish fraud" option, vs. all consensus, and all trust.
755 2013-06-10 11:20:33 <melvster> you can do this quite easily on the web
756 2013-06-10 11:20:48 <tumak> what data structure?
757 2013-06-10 11:20:53 <melvster> in that way you can verify that the hash and data were consistent (zero trust) kind of idea
758 2013-06-10 11:20:54 <petertodd> melvster: think more about *why* you can do that on the web... you don't understand the problem yet
759 2013-06-10 11:21:11 <melvster> tumak: the data structure can be anything, but ill model blocks, tx, and chain states first
760 2013-06-10 11:21:20 <melvster> and have a linked block chain
761 2013-06-10 11:21:36 <melvster> which could be a single hash per block for lite nodes
762 2013-06-10 11:21:42 <tumak> melvster: if you want to replace proof-of-work with proof-of-website, google timekoin
763 2013-06-10 11:21:53 <petertodd> melvster: yeah, I'll be mean and tell you to go away and think about this stuff more. Here's your homework problem: what's the issue with merge mining and validation?
764 2013-06-10 11:22:53 <tumak> (small modification to first rule of webcoin-dev - even ridiculous ideas are sometimes well implemented)
765 2013-06-10 11:23:06 <petertodd> tumak: lol
766 2013-06-10 11:23:20 <melvster> petertodd: i may not be up to date on all the details that you are, but I've been thinking about this problem for 15 years, and am happy to still be learning, thanks for the homework assignment :)
767 2013-06-10 11:23:26 anarchy5 has quit (Quit: Computer has gone to sleep.)
768 2013-06-10 11:23:37 <petertodd> tumak: and good ideas are usually only on paper because their creators are ivory tower theorists... and there are bad ideas from those theorists too!
769 2013-06-10 11:23:57 <tumak> petertodd: there are opportunist coders like me, killerstorm or grau!
770 2013-06-10 11:24:02 <melvster> I very much doubt that anyone else has a standards based approach of translating crypto coins to the web ... there's a tiny intersection of web and p2p developers
771 2013-06-10 11:24:04 <tumak> or justmoon, who defected to ripple, sadly :/
772 2013-06-10 11:24:05 <petertodd> melvster: no problem, I was thinking about this idea 10 years ago, and got absolutely no-where until I understood how bitcoin worked :)
773 2013-06-10 11:24:22 <melvster> yes well i think bitcoin opened all our eyes, thanks to the genius of satoshi
774 2013-06-10 11:24:37 grau has quit (Remote host closed the connection)
775 2013-06-10 11:24:40 <petertodd> tumak: heh, well if you're feeling really opportunistic, go implement my zookeyv thing for me :P
776 2013-06-10 11:25:32 <melvster> the trick in making the web into a distributed database is to make the data structure machine readable and extensible using reverse lookup and the di: uri scheme and also storing in a consistent place /.well-known/di/foo
777 2013-06-10 11:25:35 one_zero has quit ()
778 2013-06-10 11:25:48 <tumak> petertodd: in the sense we will strain the elaborate ivory towers of yours or mike hearn's into something with remote possibility of being implemented :)
779 2013-06-10 11:25:52 <tumak> at least thats the vision anyway
780 2013-06-10 11:25:59 <petertodd> melvster: go away for a few days, trust me :)
781 2013-06-10 11:26:10 <melvster> i do :)
782 2013-06-10 11:26:18 <melvster> I dream about this stuff
783 2013-06-10 11:26:24 <melvster> sometimes wake up with a new idea :)
784 2013-06-10 11:26:31 abrkn has joined
785 2013-06-10 11:27:00 <tumak> petertodd: keyv ... sounds like namecoin, but done right and generalized for other stuff?
786 2013-06-10 11:27:06 <petertodd> tumak: heh.... mike does that himself, he keeps on going on about crummy fidelity bonds with just high-fee txs that any miner can create for free...
787 2013-06-10 11:27:20 <melvster> one final point is that you dont need scripts, just balances
788 2013-06-10 11:27:30 <petertodd> tumak: yeah, just a pure proof-of-sacrifice blockchain, well, actually block-directed-acyclic-graph
789 2013-06-10 11:27:52 <petertodd> melvster: that's homework problem #2 for you, figure out why you do in fact need transactions
790 2013-06-10 11:28:14 <tumak> petertodd: writeup?
791 2013-06-10 11:28:35 <melvster> petertodd: transactions are key ... they are what creates the ledger, and allows you to verify
792 2013-06-10 11:29:02 <tumak> not necessarily :) there are blockchain-less, account-based ledgers too
793 2013-06-10 11:29:02 <melvster> and also grow the total proof of work
794 2013-06-10 11:29:04 da2ce7 has joined
795 2013-06-10 11:29:13 <tumak> and can be quite efficient if it werent for that darn distributed random number generator
796 2013-06-10 11:29:29 <nsh> the universe is a distributed random number generator
797 2013-06-10 11:29:32 <nsh> hth
798 2013-06-10 11:29:38 <petertodd> tumak: http://pastebin.com/Rj4bshY3 it's about halfway down
799 2013-06-10 11:29:57 <petertodd> tumak: think about how you verify those account-based ledgers...
800 2013-06-10 11:30:08 <Plinker> So its not even an inverted triangle, more like an imaginary snake
801 2013-06-10 11:30:29 <Plinker> Why do you say the universe is random, can you prove this?
802 2013-06-10 11:31:15 <melvster> does god play dice lol?
803 2013-06-10 11:31:27 abrkn\ has joined
804 2013-06-10 11:31:46 <Plinker> Exactly
805 2013-06-10 11:31:53 <tumak> petertodd: as i said, it's impossible without at least PoS :(
806 2013-06-10 11:31:53 anarchy5 has joined
807 2013-06-10 11:32:27 <petertodd> tumak: I mean the mechanics of verification, you have to verify why the accounts changed, and why they changed was due to transactions...
808 2013-06-10 11:32:27 <tumak> ie the lottery numbers seed generator or some shim like that unless some major math/algo breakthrough comes up
809 2013-06-10 11:33:10 <tumak> petertodd: the RNG decides which double-spend wins and which loses
810 2013-06-10 11:33:22 <petertodd> tumak: I think we're talking about something else
811 2013-06-10 11:33:33 <Plinker> I like the lottery concept, like a largeponzi scheme..
812 2013-06-10 11:33:33 <petertodd> tumak: s/else/different/
813 2013-06-10 11:33:50 _sgstair has joined
814 2013-06-10 11:33:50 sgstair has quit (Disconnected by services)
815 2013-06-10 11:33:50 _sgstair is now known as sgstair
816 2013-06-10 11:34:25 <tumak> petertodd: everyone has table of accounts, right. transaction are broadcasted, transfer from account a to account b, for each such transaction (including double-spend) you create dirty version of such tables
817 2013-06-10 11:34:29 <tumak> let there be God
818 2013-06-10 11:34:45 <tumak> god says there is random number, same across the world, every 10 minutes
819 2013-06-10 11:34:54 abrkn has quit (Ping timeout: 276 seconds)
820 2013-06-10 11:35:16 <petertodd> tumak: right, I guess my point is that you *can't* get away from having transactions, and a transaction history, because that's what you audit to achieve meaningful consensus, the actual proof-of-* mechanism is irrelevant
821 2013-06-10 11:35:48 CodesInChaos has quit (Read error: Connection reset by peer)
822 2013-06-10 11:35:54 <petertodd> tumak: some people seem to think you can someone have consensus with just accounts, bizzarely
823 2013-06-10 11:35:54 <tumak> the history is just the dirty copies of table
824 2013-06-10 11:35:54 <tumak> once the Number is known, history is set in stone
825 2013-06-10 11:35:54 <tumak> problem is
826 2013-06-10 11:35:57 <tumak> there is no such number generator
827 2013-06-10 11:36:05 CodesInChaos has joined
828 2013-06-10 11:36:13 <tumak> i think satoshi called it "distributed timestamping" problem
829 2013-06-10 11:36:42 <melvster> it's odd mining is essentially gambling, yet I thought gambling was illegal in the US
830 2013-06-10 11:37:01 <nsh> the hash of the last block is a globally-consensuated pseudorandom number generated approximately every ten minutes
831 2013-06-10 11:37:05 <tumak> melvster: there is no house edge
832 2013-06-10 11:37:12 <tumak> nsh: yup
833 2013-06-10 11:37:23 <tumak> nsh: but even that random number is not definitive
834 2013-06-10 11:37:27 <petertodd> Yup. An interesting problem would be to figure out a decent way to pay for a PoW timestamping system. Possibly jus thaving "timestamp coin" with fast UDP block headers and a shortist blocktime is good enough, and have participating currencies only use a subset of blocks.
835 2013-06-10 11:37:31 <nsh> definitive of what?
836 2013-06-10 11:37:32 <tumak> so you still need to keep full history
837 2013-06-10 11:37:37 <tumak> because reorg can happen
838 2013-06-10 11:37:43 <nsh> mm
839 2013-06-10 11:37:43 <tumak> and suddenly, the number is invalid
840 2013-06-10 11:37:44 <Plinker> And the concept of gambling is that some win and some have to lose
841 2013-06-10 11:37:50 * nsh nods
842 2013-06-10 11:38:17 <Plinker> Input = output
843 2013-06-10 11:38:36 <melvster> tumak: gambling does not imply a house edge
844 2013-06-10 11:38:40 ralphtheninja has quit (Ping timeout: 246 seconds)
845 2013-06-10 11:38:53 <Plinker> Makes a wonderful laundering method though you have to admit!
846 2013-06-10 11:39:05 <tumak> petertodd: there can be made interesting hacks even in bitcoin, though
847 2013-06-10 11:39:06 <petertodd> Basically the timestamping thing is just your master PoW function for your global set of PoS blockchains, and add rules where successful PoW's gain you something in the PoS... somehow.
848 2013-06-10 11:39:24 <tumak> petertodd: say, why hardcode only checkpoints in official client, why not hardcode utxo set at that date, too?
849 2013-06-10 11:39:33 <petertodd> tumak: that will be done
850 2013-06-10 11:39:45 <nsh> petertodd, PoS here being Proof of Stake or Sacrifice?
851 2013-06-10 11:39:54 <tumak> petertodd: thats basically "half-assed accounts" by then :)
852 2013-06-10 11:40:00 <petertodd> nsh: heh, actually both works
853 2013-06-10 11:40:03 <tumak> tx history only for the sake of random number generator
854 2013-06-10 11:40:24 <tumak> petertodd: might be interesting to develop this idea into account-based altcoin abusing bitcoin for timestamps though :)
855 2013-06-10 11:40:27 <petertodd> tumak: oh for sure
856 2013-06-10 11:40:31 <tumak> though the rng would be useful in span of days
857 2013-06-10 11:40:44 <Plinker> I am still waiting on the DOJ to see what further steps they take
858 2013-06-10 11:41:03 saulimus has quit (Quit: saulimus)
859 2013-06-10 11:41:21 <petertodd> tumak: heh, yeah one neat things about a PoW/PoS hybrid is it *may* be possible to get enough activity around it that attacking the underlying PoW is infeasible... good luck though
860 2013-06-10 11:42:28 <tumak> petertodd: i'll wait for your formal proof that PoW == distributed rng first :)
861 2013-06-10 11:42:35 <Plinker> The main problem is tracking where money comes from and where it goes..... Not the ponzi aspect of bitcoin
862 2013-06-10 11:42:51 <petertodd> tumak: lol, that won't be easy.
863 2013-06-10 11:43:10 agnostic98 has joined
864 2013-06-10 11:43:23 metabyte_ has joined
865 2013-06-10 11:43:39 <tumak> i wonder if bitcoin foundation could sponsor math bounty like there is for other interesting math conjenctures :)
866 2013-06-10 11:43:46 <petertodd> tumak: one trick in my yet unpublished semi-paper was to do the evaluation of which PoS wins by computing H(P|R) with P=PoS and R=RNG
867 2013-06-10 11:43:50 <tumak> frankly, your testnet seed is crap, this could be much more useful
868 2013-06-10 11:44:09 <petertodd> tumak: that means provided P's are unknown to the miner, manipulating R doesn't help
869 2013-06-10 11:44:26 Sealy has joined
870 2013-06-10 11:44:32 <petertodd> tumak: for sure, but it was a decent first project to test things out
871 2013-06-10 11:44:53 Sealy has quit (Client Quit)
872 2013-06-10 11:45:12 <nsh> heh
873 2013-06-10 11:45:14 Sealy has joined
874 2013-06-10 11:45:49 agnostic98 has quit (Read error: Connection reset by peer)
875 2013-06-10 11:46:01 metabyte has quit (Ping timeout: 246 seconds)
876 2013-06-10 11:46:06 * nsh has a strange idea for an altcoin where the PoW is pseudorandom proof-generation and the blockchain represents a reasoning system of constantly growing complexity
877 2013-06-10 11:46:28 <tumak> nsh: yeah, i tried to ponder something like that
878 2013-06-10 11:46:43 <tumak> nsh: like nodes betting head-tails with each other
879 2013-06-10 11:46:48 <nsh> it makes my brain hurt in a nice way
880 2013-06-10 11:46:48 <nsh> :)
881 2013-06-10 11:46:49 <tumak> but ultimately, you always end up with PoS
882 2013-06-10 11:46:56 <nsh> mm
883 2013-06-10 11:47:11 ThomasV has joined
884 2013-06-10 11:47:23 zer0def has quit (Ping timeout: 255 seconds)
885 2013-06-10 11:47:54 <tumak> (PoS alone still needs the rng)
886 2013-06-10 11:48:59 comboy has joined
887 2013-06-10 11:49:32 <petertodd> tumak: random beacon is the terminology btw
888 2013-06-10 11:49:43 <petertodd> tumak: they come up in zero-knowledge protocols too
889 2013-06-10 11:50:24 <tumak> ah, ok
890 2013-06-10 11:50:41 <tumak> btw, wrt the hardcoded utxo set hack, second level would be actually including it in coinbase
891 2013-06-10 11:50:43 <petertodd> tumak: look up non-interactive zero-knowledge protocols, interesting stuff
892 2013-06-10 11:50:48 <nsh> ;;title http://link.springer.com/chapter/10.1007%2F978-3-642-33536-5_25
893 2013-06-10 11:50:48 <gribble> Error: This url is not on the whitelist.
894 2013-06-10 11:50:49 <tumak> say, every 14 days
895 2013-06-10 11:50:54 <petertodd> tumak: yeah, you need to read bitcointalk more...
896 2013-06-10 11:50:59 <tumak> :(
897 2013-06-10 11:51:00 ralphtheninja has joined
898 2013-06-10 11:51:02 <tumak> that forum is horrible
899 2013-06-10 11:51:08 <nsh> gribble, eat a dick. (Scalable Byzantine Agreement with a Random Beacon)
900 2013-06-10 11:51:11 * tumak still needs first rule
901 2013-06-10 11:51:26 <petertodd> tumak: there's a project to implement UTXO in coinbase
902 2013-06-10 11:52:10 <petertodd> anyway, later
903 2013-06-10 11:52:11 <tumak> cant find in greg's altcoin wishlist
904 2013-06-10 11:52:27 <nsh> https://en.bitcoin.it/wiki/User:Gmaxwell/alt_ideas ?
905 2013-06-10 11:52:55 <petertodd> tumak: it's a part of the first idea, see the link to https://bitcointalk.org/index.php?topic=137933.0 which depends on it
906 2013-06-10 11:53:11 <tumak> found https://bitcointalk.org/index.php?topic=127169.msg1350177
907 2013-06-10 11:53:16 grau has joined
908 2013-06-10 11:53:52 grau has quit (Remote host closed the connection)
909 2013-06-10 11:53:55 <tumak> ah, grau caught red-handed in plagiarism!
910 2013-06-10 11:54:44 grau has joined
911 2013-06-10 11:54:51 <nsh> (that paper is pretty useless, it assumes a random beacon and a partially synchronous communication model)
912 2013-06-10 11:54:52 <tumak> yeah, we'll need to deal with utxo sets in cc soon
913 2013-06-10 11:55:25 grau has quit (Remote host closed the connection)
914 2013-06-10 11:55:33 <tumak> to keep it simple, issuers will just publish em, but its a horrible ugly hack :/
915 2013-06-10 11:55:59 <tumak> (fidelity bonds might come handy in there as well, not just IOUs)
916 2013-06-10 11:56:05 grau has joined
917 2013-06-10 11:56:15 <nsh> tumak, in cc?
918 2013-06-10 11:56:19 <tumak> colored coins :)
919 2013-06-10 11:56:23 <nsh> ah
920 2013-06-10 11:56:39 <tumak> basically the resolution algo is horribly inneficient
921 2013-06-10 11:56:49 <tumak> so we hit scalability issues much sooner than bitcoin itself
922 2013-06-10 11:56:54 <nsh> like marriage then
923 2013-06-10 11:57:14 <nsh> (resolution algo being inefficient. nm...)
924 2013-06-10 11:58:02 Tantadruj has joined
925 2013-06-10 12:01:01 grau has quit (Ping timeout: 264 seconds)
926 2013-06-10 12:01:46 da2ce7 has quit (Ping timeout: 246 seconds)
927 2013-06-10 12:02:35 patcon has quit (Ping timeout: 248 seconds)
928 2013-06-10 12:02:40 Squidicuz has joined
929 2013-06-10 12:04:53 Squid_ has quit (Ping timeout: 252 seconds)
930 2013-06-10 12:06:13 tyn has joined
931 2013-06-10 12:07:05 Squidicuz has quit (Ping timeout: 252 seconds)
932 2013-06-10 12:12:03 metabyte_ is now known as metabyte
933 2013-06-10 12:14:15 agnostic98 has joined
934 2013-06-10 12:14:52 da2ce7 has joined
935 2013-06-10 12:16:13 fourlights_ has joined
936 2013-06-10 12:16:47 fourlights has quit (Ping timeout: 240 seconds)
937 2013-06-10 12:16:51 agnostic98 has quit (Read error: Connection reset by peer)
938 2013-06-10 12:19:55 klmist has quit (Ping timeout: 256 seconds)
939 2013-06-10 12:28:19 daybyter has quit (Quit: Konversation terminated!)
940 2013-06-10 12:31:28 grau has joined
941 2013-06-10 12:32:17 nikolaj has quit (Ping timeout: 240 seconds)
942 2013-06-10 12:33:04 robocoin has joined
943 2013-06-10 12:33:04 robocoin has quit (Changing host)
944 2013-06-10 12:33:04 robocoin has joined
945 2013-06-10 12:35:19 BlackPrapor has joined
946 2013-06-10 12:37:10 cads has quit (Quit: Leaving)
947 2013-06-10 12:38:31 debiantoruser has quit (Ping timeout: 246 seconds)
948 2013-06-10 12:38:47 crank has quit (Ping timeout: 240 seconds)
949 2013-06-10 12:40:11 tyn has quit (Ping timeout: 264 seconds)
950 2013-06-10 12:40:32 debiantoruser has joined
951 2013-06-10 12:42:31 Sealy has quit (Quit: Sealy)
952 2013-06-10 12:43:07 daz_ has joined
953 2013-06-10 12:43:30 daz_ is now known as Guest47017
954 2013-06-10 12:44:31 agricocb has quit (Quit: Leaving.)
955 2013-06-10 12:45:53 chorao has quit (Ping timeout: 255 seconds)
956 2013-06-10 12:53:25 copumpkin has quit (Quit: Computer has gone to sleep.)
957 2013-06-10 12:58:10 <shesek> when preparing a transaction for hashing/signing, what script should I use for p2sh inputs - the original output hashed script (HASH_160 ... OP_EQUAL), or the "real" pre-hashed script?
958 2013-06-10 12:58:56 Tantadruj has quit (Ping timeout: 255 seconds)
959 2013-06-10 12:59:47 caedes has joined
960 2013-06-10 12:59:47 caedes has quit (Changing host)
961 2013-06-10 12:59:47 caedes has joined
962 2013-06-10 13:00:06 Tantadruj has joined
963 2013-06-10 13:02:19 tyn has joined
964 2013-06-10 13:05:22 RazielZ has quit (Quit: Leaving)
965 2013-06-10 13:07:10 abracadabra has quit (Ping timeout: 245 seconds)
966 2013-06-10 13:08:43 agricocb has joined
967 2013-06-10 13:09:00 abracadabra has joined
968 2013-06-10 13:09:14 nikolaj has joined
969 2013-06-10 13:11:00 malexmedia has joined
970 2013-06-10 13:12:35 <warren> TheUni: unzip: cannot find or open ../qt-win32-4.8.3-gitian-r1.zip, ../qt-win32-4.8.3-gitian-r1.zip.zip or ../qt-win32-4.8.3-gitian-r1.zip.ZIP.
971 2013-06-10 13:12:46 <warren> TheUni: using your branch
972 2013-06-10 13:14:15 <warren> qt-win32-4.8.3-gitian-r1.zip was in inputs/
973 2013-06-10 13:14:25 <warren> so not sure what's going on
974 2013-06-10 13:27:31 ovidiusoft has quit (Ping timeout: 246 seconds)
975 2013-06-10 13:29:29 ovidiusoft has joined
976 2013-06-10 13:29:34 grau has quit (Remote host closed the connection)
977 2013-06-10 13:31:14 <judah> I'm just a noob trying to get started but it seems that bitcoind is not quite up to json-rpc level 2.0, no?
978 2013-06-10 13:33:14 tgs3 has quit (Quit: leaving)
979 2013-06-10 13:34:51 tgs3 has joined
980 2013-06-10 13:34:56 <nsh> judah, can you elaborate?
981 2013-06-10 13:35:11 <judah> bitcoind is returning json result AND error (error is null, but still..) and not returning the request ID
982 2013-06-10 13:36:22 <judah> my parser is very strict, took me a bit of fiddling get a result from it. O.o
983 2013-06-10 13:36:37 jgarzik has joined
984 2013-06-10 13:36:54 <jgarzik> mornin'
985 2013-06-10 13:37:13 altamic has joined
986 2013-06-10 13:37:16 <TD> hello
987 2013-06-10 13:37:30 altamic has left ()
988 2013-06-10 13:39:26 nikolaj has quit (Ping timeout: 255 seconds)
989 2013-06-10 13:39:53 <BlueMatt> morning? Im already on my like 5th coffee today...it cant be that early
990 2013-06-10 13:40:01 <jgarzik> real life is forcing me to become a 9-to-5'er
991 2013-06-10 13:40:24 <nsh> sounds awful
992 2013-06-10 13:40:24 * jgarzik was in Raleigh all weekend, moving his office to Atlanta (we told the movers "don't you dare touch Jeff's office")
993 2013-06-10 13:40:34 <BlueMatt> lol
994 2013-06-10 13:40:43 debiantoruser has quit (Ping timeout: 248 seconds)
995 2013-06-10 13:42:07 cc_8 has joined
996 2013-06-10 13:42:07 cc_8 is now known as alphaguru
997 2013-06-10 13:42:51 debiantoruser has joined
998 2013-06-10 13:44:13 <jgarzik> petertodd, what issues did you find with OP_RETURN pullreq?
999 2013-06-10 13:46:01 jaequery has joined
1000 2013-06-10 13:46:29 jaequery has quit (Client Quit)
1001 2013-06-10 13:46:55 zer0def has joined
1002 2013-06-10 13:47:28 agnostic98 has joined
1003 2013-06-10 13:48:35 abrkn has joined
1004 2013-06-10 13:48:57 abrkn has quit (Client Quit)
1005 2013-06-10 13:48:59 <kjj_> moving sucks
1006 2013-06-10 13:49:02 kjj_ is now known as kjj
1007 2013-06-10 13:50:11 <jgarzik> moving an entire family across states sucks++
1008 2013-06-10 13:50:26 Guest83631 has joined
1009 2013-06-10 13:50:41 <BlueMatt> moving an entire family across the atlantic sucks+++
1010 2013-06-10 13:50:52 JDuke128_ has quit (Quit: Computer has gone to sleep.)
1011 2013-06-10 13:50:57 agnostic98 has quit (Read error: Connection reset by peer)
1012 2013-06-10 13:51:07 abrkn\ has quit (Ping timeout: 245 seconds)
1013 2013-06-10 13:51:15 <jgarzik> kjj, was good to meet you at the conf, BTW. You always have enjoyable forum posts, saying the stuff I would like to say to trolls [but never have the energy to]. I applaud your energy ;p
1014 2013-06-10 13:51:24 <kjj> just moving myself an hour sucked. but I think I have more "stuff" than many families
1015 2013-06-10 13:51:50 <kjj> it was nice meeting you too. wish you'd been able to make it to supper
1016 2013-06-10 13:51:58 <nsh> i applaud anyone trying to hose 100 bitcointalk trolls, but take it from this old flamerat, such a program will do more damage than good....
1017 2013-06-10 13:52:12 <nsh> ;)
1018 2013-06-10 13:52:45 <kjj> nsh: I mostly try to confine myself, these days, to people that imagine mythical powers in the foundation or the dev team
1019 2013-06-10 13:53:09 * nsh smiles, was just quoting a silly reddit meme
1020 2013-06-10 13:53:24 <nsh> ( http://www.reddit.com/r/science/comments/6nz1k/got_six_weeks_try_the_hundred_push_ups_training/ )
1021 2013-06-10 13:53:34 MC1984 has quit (Quit: Leaving)
1022 2013-06-10 13:53:46 <nsh> s/quoting/invoking/
1023 2013-06-10 13:54:00 tyn has quit (Ping timeout: 264 seconds)
1024 2013-06-10 13:54:37 <nsh> but we all know the commit bit bestows both prescience and a supernatural ability to bend the laws of mathematics
1025 2013-06-10 13:57:24 yubrew has joined
1026 2013-06-10 14:01:43 nikolaj has joined
1027 2013-06-10 14:05:20 tyn has joined
1028 2013-06-10 14:05:40 <r0sc0e> how can i connect to a note i add manually @ bitcoin qt 0.8.2
1029 2013-06-10 14:07:06 denom has joined
1030 2013-06-10 14:07:21 <nsh> r0sc0e, come again? connect to a node?
1031 2013-06-10 14:07:59 robocoin has quit (Ping timeout: 252 seconds)
1032 2013-06-10 14:08:17 <r0sc0e> i knew how to add
1033 2013-06-10 14:08:23 <r0sc0e> but not to "jump" to ;)
1034 2013-06-10 14:10:16 nikolaj has quit (Ping timeout: 276 seconds)
1035 2013-06-10 14:10:39 rdymac has quit (Quit: This computer has gone to sleep)
1036 2013-06-10 14:10:42 GordonG3kko has quit (Remote host closed the connection)
1037 2013-06-10 14:11:56 saulimus has joined
1038 2013-06-10 14:14:33 macboz has joined
1039 2013-06-10 14:18:33 agnostic98 has joined
1040 2013-06-10 14:18:41 Ashaman has joined
1041 2013-06-10 14:20:36 kwikness has joined
1042 2013-06-10 14:21:01 agnostic98 has quit (Read error: Connection reset by peer)
1043 2013-06-10 14:21:21 bitRipperX has joined
1044 2013-06-10 14:22:20 <bitRipperX> I piad for something using coin base and my transaction went through in a few seconds. From what I read it may take a couple hours for a tx to be confirmed. How do they do it so fast? Are they doing anything special?
1045 2013-06-10 14:22:39 <runeks> r0sc0e: You mean you've added a node in bitcoin.conf via "addnode=" but you want to connect specifically to that node?
1046 2013-06-10 14:24:55 malexmedia has quit (Ping timeout: 240 seconds)
1047 2013-06-10 14:26:17 <judah> bitRipper: I think they notified back with zero confirmations, then the next blocks will have more confirmations in them
1048 2013-06-10 14:27:18 <r0sc0e> runeks: yes
1049 2013-06-10 14:27:24 <r0sc0e> because the one i am connected to
1050 2013-06-10 14:27:30 <r0sc0e> is too slow for synchronisation
1051 2013-06-10 14:27:43 <bitRipperX> judah: doesn't that leave them open for someone double spending?
1052 2013-06-10 14:28:38 <judah> bitRipperX: yah, but who does that? they'll take that risk for "customer satisfaction"
1053 2013-06-10 14:28:47 <runeks> r0sc0e: I think you might need to connect to your desired peer using "connect=" instead of "addnode=". I'm not sure, but I don't think you can choose which nodes are connected to, when you add them via "addnode=".
1054 2013-06-10 14:32:11 macboz has quit (Quit: This computer has gone to sleep)
1055 2013-06-10 14:32:56 malexmedia has joined
1056 2013-06-10 14:36:35 gfinn has joined
1057 2013-06-10 14:36:49 patcon has joined
1058 2013-06-10 14:38:22 patcon has quit (Remote host closed the connection)
1059 2013-06-10 14:40:37 patcon has joined
1060 2013-06-10 14:46:53 roconnor has joined
1061 2013-06-10 14:48:43 whiterabbit has joined
1062 2013-06-10 14:49:10 crescend1 has quit (Remote host closed the connection)
1063 2013-06-10 14:49:35 agnostic98 has joined
1064 2013-06-10 14:50:03 wrabbit has quit (Ping timeout: 240 seconds)
1065 2013-06-10 14:50:10 whiterabbit is now known as wrabbit
1066 2013-06-10 14:51:25 tyn has quit (Ping timeout: 264 seconds)
1067 2013-06-10 14:52:02 lolcookie__ has joined
1068 2013-06-10 14:52:07 whiterabbit has joined
1069 2013-06-10 14:52:13 agnostic98 has quit (Read error: Connection reset by peer)
1070 2013-06-10 14:53:37 i2pRelay has quit (Remote host closed the connection)
1071 2013-06-10 14:55:23 wrabbit has quit (Ping timeout: 240 seconds)
1072 2013-06-10 14:55:23 whiterabbit is now known as wrabbit
1073 2013-06-10 14:56:14 tyn has joined
1074 2013-06-10 14:58:53 i2pRelay has joined
1075 2013-06-10 15:01:40 denom has quit (Ping timeout: 246 seconds)
1076 2013-06-10 15:02:08 fanquake has left ()
1077 2013-06-10 15:04:36 altamic has joined
1078 2013-06-10 15:04:39 altamic has quit (Client Quit)
1079 2013-06-10 15:05:24 denom has joined
1080 2013-06-10 15:10:23 MiningBuddy has joined
1081 2013-06-10 15:10:23 MiningBuddy has quit (Changing host)
1082 2013-06-10 15:10:23 MiningBuddy has joined
1083 2013-06-10 15:11:44 defunctzombie_zz is now known as defunctzombie
1084 2013-06-10 15:11:57 * jgarzik does the make-sipa-appear dance
1085 2013-06-10 15:12:05 <jgarzik> just merged #2154
1086 2013-06-10 15:12:12 <jgarzik> pull pull pull, test test test
1087 2013-06-10 15:17:03 rdymac has joined
1088 2013-06-10 15:17:03 rdymac has quit (Changing host)
1089 2013-06-10 15:17:03 rdymac has joined
1090 2013-06-10 15:17:54 <runeks> So what would happen if more than 1 MB of transactions were generated per "time-between-blocks" (~10 mins)? Would the memory pool of bitcoind just keep filling up until it crashes from excessive memory usage?
1091 2013-06-10 15:18:10 BCBot has quit (Remote host closed the connection)
1092 2013-06-10 15:20:05 <jgarzik> runeks, Answer to first question: transactions build up in the memory pool
1093 2013-06-10 15:20:38 agnostic98 has joined
1094 2013-06-10 15:20:41 <jgarzik> runeks, Answer to second question: yes (but that takes many thousands of transactions, possibly hundreds of thousands)
1095 2013-06-10 15:20:53 <runeks> jgarzik: So once the memory pool is 1MB, no more transactions are added to it?
1096 2013-06-10 15:21:11 <jgarzik> 144MB of transactions (1 day's worth at max block size) is not much
1097 2013-06-10 15:21:23 <jgarzik> runeks, the memory pool does not have a size limit
1098 2013-06-10 15:21:25 <runeks> Yeah that's true.
1099 2013-06-10 15:21:31 <jgarzik> runeks, blocks have a size limit, not memory pool
1100 2013-06-10 15:21:48 BCBot has joined
1101 2013-06-10 15:22:24 <runeks> jgarzik: Ok. So if there are more than 1 MB of txs in the memory pool, bitcoind just selects 1MB of them if I issue the "getblocktemplate" command?
1102 2013-06-10 15:24:40 agnostic98 has quit (Ping timeout: 240 seconds)
1103 2013-06-10 15:27:28 patcon_ has joined
1104 2013-06-10 15:28:23 <jgarzik> runeks, correct
1105 2013-06-10 15:28:46 cjsw3 has joined
1106 2013-06-10 15:28:56 <runeks> Ok. Makes sense. So this really isn't a problem. Good to know.
1107 2013-06-10 15:29:00 <jgarzik> runeks, block creation pulls from the memory pool, but that does not therefore imply any limitations on the mempool
1108 2013-06-10 15:29:33 <petertodd> runeks: basically it takes a certain amount of BTC for an attacker to create enough transactions to crash nodes via memory exhaustion
1109 2013-06-10 15:29:34 Thepok has quit (Ping timeout: 276 seconds)
1110 2013-06-10 15:29:48 <runeks> jgarzik: Right. And thinking about it, that's the only thing that makes sense. Otherwise transaction fees wouldn't matter, and it would just be a first come first serve thing, which we don't want.
1111 2013-06-10 15:29:53 <jgarzik> runeks, mempool is the set of transactions /valid and available/ for block creation, whether the mempool is 1k or 100MB in size doesn't matter
1112 2013-06-10 15:30:00 <jgarzik> yep
1113 2013-06-10 15:30:00 <petertodd> runeks: 1mBTC/KB, so that's 100BTC for 1GiB of mempool size
1114 2013-06-10 15:30:25 patcon has quit (Ping timeout: 264 seconds)
1115 2013-06-10 15:30:35 <runeks> petertodd: Is that the soft rule of bitcoind?
1116 2013-06-10 15:30:46 <petertodd> runeks: it's not that much money... although I'm working on a mempool rewrite which will make it easy to limit the mempool
1117 2013-06-10 15:31:13 <petertodd> runeks: 1mBTC/KB is a lower limit on fees yes
1118 2013-06-10 15:31:20 <dugo> 100BTC buys you quite some ram
1119 2013-06-10 15:31:35 <petertodd> brb
1120 2013-06-10 15:31:38 <runeks> Hmm. 1 mBTC/KiB is 1BTC/MiB, which is 1000BTC/GiB
1121 2013-06-10 15:31:56 lolcookie__ has quit (Read error: Connection reset by peer)
1122 2013-06-10 15:32:15 lolcookie__ has joined
1123 2013-06-10 15:32:17 <runeks> petertodd: And yeah, it's not a serious issue in any way. Just one of those nice-to-have things that aren't necessary yet.
1124 2013-06-10 15:32:34 jaequery has joined
1125 2013-06-10 15:33:21 <runeks> dugo: Yeah I'm sure most miners would be happy to offer 1 GiB of memory for 1000 BTC :)
1126 2013-06-10 15:33:25 Belkaar has quit (Ping timeout: 264 seconds)
1127 2013-06-10 15:35:49 Belkaar has joined
1128 2013-06-10 15:40:11 hpprinter100 has quit (Read error: Connection reset by peer)
1129 2013-06-10 15:40:50 vragnaroda has joined
1130 2013-06-10 15:41:02 agnostic98 has joined
1131 2013-06-10 15:49:14 xnptzk` has joined
1132 2013-06-10 15:50:42 Nothing4You has quit (K-Lined)
1133 2013-06-10 15:51:06 <tgs3> dugo: I guess the problem is that everyone needs to
1134 2013-06-10 15:51:40 phebus has quit (Quit: POKE 1,0)
1135 2013-06-10 15:52:07 phebus has joined
1136 2013-06-10 15:52:26 helo is now known as aTrailOfYogurt
1137 2013-06-10 15:52:47 aTrailOfYogurt is now known as helo
1138 2013-06-10 15:52:53 vragnaroda has quit (Quit: bye)
1139 2013-06-10 15:54:39 t7 has quit (Quit: ChatZilla 0.9.90 [Firefox 21.0/20130511120803])
1140 2013-06-10 15:56:41 chorao has joined
1141 2013-06-10 15:56:41 chorao has quit (Changing host)
1142 2013-06-10 15:56:41 chorao has joined
1143 2013-06-10 15:58:16 Subo1978 has joined
1144 2013-06-10 16:00:23 Subo1978_ has quit (Ping timeout: 240 seconds)
1145 2013-06-10 16:04:18 agath_pd has joined
1146 2013-06-10 16:04:27 <petertodd> jgarzik: first issue, you put the OP_RETURN on the wrong end, it should be at the front of the script
1147 2013-06-10 16:05:28 agath has quit (Ping timeout: 256 seconds)
1148 2013-06-10 16:05:34 patcon_ has quit (Remote host closed the connection)
1149 2013-06-10 16:07:21 Ashaman has quit (Remote host closed the connection)
1150 2013-06-10 16:13:03 <jgarzik> petertodd, any other issues?
1151 2013-06-10 16:13:25 ovidiusoft has quit (Ping timeout: 246 seconds)
1152 2013-06-10 16:14:45 Nothing4You has joined
1153 2013-06-10 16:15:14 <petertodd> jgarzik: ScriptSigArgsExpected() and similar was done a bit odd, in many cases you are saying that OP_RETURN expects signatures, which implies it's spendable
1154 2013-06-10 16:15:51 <petertodd> jgarzik: Also I'd suggest s/null/unspendable/
1155 2013-06-10 16:16:16 <jgarzik> petertodd, the latter was largely based on the context -- how its used, which muddies the waters a bit
1156 2013-06-10 16:16:47 <petertodd> jgarzik: I yeah, I'm thinking more what users will think null means.
1157 2013-06-10 16:17:05 <jgarzik> petertodd, the former was a trap -- seeing if anybody was paying attention
1158 2013-06-10 16:17:15 <jgarzik> (ok, it was an error, but one spotted very early on :))
1159 2013-06-10 16:17:21 <petertodd> jgarzik: lol
1160 2013-06-10 16:18:10 <petertodd> jgarzik: Something to consider is we may want to say that OP_RETURN is supposed to have digests as the data, because of gmaxwell's P2SH^2 pre-image proposal.
1161 2013-06-10 16:19:00 <jgarzik> petertodd, "say" == update the commit to add code comments, or just thinking about messaging and discussion outside the code?
1162 2013-06-10 16:19:23 <petertodd> jgarzik: I also worked it out, and if we limit OP_RETURN to 20 bytes, but allow an unlimited number, it's 3x cheaper to not bloat the UTXO set than it is to bloat it, while still 2x more expensive than the usual per-kb fee because of the overhead.
1163 2013-06-10 16:19:35 malexmedia has quit (Quit: Leaving.)
1164 2013-06-10 16:19:36 <warren> trying to send this https://github.com/wtogami/bitcoin/commit/0e222a3c480b77e863ba5b68099d05dc264fdff1 as a pull request to theuni/bitcoin but github doesn't show him as a base option
1165 2013-06-10 16:19:46 <petertodd> jgarzik: No, I mean technical stuff like making OP_RETURN only allow 20 and/or 32byte digests
1166 2013-06-10 16:20:25 <petertodd> jgarzik: er, well, combined with the comment saying "in the future we may require a pre-image to relay your tx"
1167 2013-06-10 16:21:00 denom has quit (Ping timeout: 264 seconds)
1168 2013-06-10 16:22:21 <jgarzik> bbiab, lunch
1169 2013-06-10 16:24:08 * helo watches as VISA hires half of the bitcoin developers to make a centralized bitcoin-like payment system (ECDSA, free 'private' accounts, etc)
1170 2013-06-10 16:25:57 <warren> grrr... github pull requests are finickey
1171 2013-06-10 16:26:13 <warren> you can't choose certain users as a pull target
1172 2013-06-10 16:27:13 <warren> anyway, need sleep
1173 2013-06-10 16:27:18 <warren> TheUni: https://github.com/wtogami/bitcoin/commit/0e222a3c480b77e863ba5b68099d05dc264fdff1
1174 2013-06-10 16:27:25 <warren> TheUni: that fixes your win32 gitian build
1175 2013-06-10 16:27:32 <warren> TheUni: read the inline comments
1176 2013-06-10 16:27:50 <warren> TheUni: something is different about your acccount. unable to send you pull requestes.
1177 2013-06-10 16:28:43 jaequery has quit (Quit: Computer has gone to sleep.)
1178 2013-06-10 16:29:09 richcollins has joined
1179 2013-06-10 16:31:37 fourlights_ has quit (Ping timeout: 264 seconds)
1180 2013-06-10 16:33:17 Squidicuz has joined
1181 2013-06-10 16:33:36 Belkaar has quit (Ping timeout: 264 seconds)
1182 2013-06-10 16:33:50 Thepok has joined
1183 2013-06-10 16:36:20 Belkaar has joined
1184 2013-06-10 16:36:44 richcollins has quit (Quit: richcollins)
1185 2013-06-10 16:37:05 Muis has quit (Read error: Connection reset by peer)
1186 2013-06-10 16:37:43 beethoven2 has quit (Ping timeout: 240 seconds)
1187 2013-06-10 16:39:23 Muis has joined
1188 2013-06-10 16:41:02 RazielZ has joined
1189 2013-06-10 16:41:11 FredEE has joined
1190 2013-06-10 16:42:12 jaequery has joined
1191 2013-06-10 16:42:26 jaequery has quit (Client Quit)
1192 2013-06-10 16:43:20 cjsw3 has quit (Quit: Textual IRC Client: www.textualapp.com)
1193 2013-06-10 16:43:29 fourlights_ has joined
1194 2013-06-10 16:43:40 tyn has quit (Ping timeout: 276 seconds)
1195 2013-06-10 16:44:13 alphaguru has quit (Ping timeout: 264 seconds)
1196 2013-06-10 16:46:50 cc_8 has joined
1197 2013-06-10 16:46:50 cc_8 is now known as alphaguru
1198 2013-06-10 16:47:14 crank has joined
1199 2013-06-10 16:48:03 BlackPrapor has quit (Ping timeout: 252 seconds)
1200 2013-06-10 16:51:14 yubrew has quit (Remote host closed the connection)
1201 2013-06-10 16:51:39 random_cat has quit (Remote host closed the connection)
1202 2013-06-10 16:52:00 Heimdall has joined
1203 2013-06-10 16:53:21 tyn has joined
1204 2013-06-10 16:55:27 cc_8 has joined
1205 2013-06-10 16:55:37 ralphtheninja has quit (Ping timeout: 264 seconds)
1206 2013-06-10 16:57:31 cc_8 has quit (Client Quit)
1207 2013-06-10 16:57:40 richcollins has joined
1208 2013-06-10 17:00:53 FredEE has quit (Quit: FredEE)
1209 2013-06-10 17:07:53 jaequery has joined
1210 2013-06-10 17:10:44 Thepok has quit (Ping timeout: 240 seconds)
1211 2013-06-10 17:12:15 tyn has quit (Ping timeout: 252 seconds)
1212 2013-06-10 17:13:09 daybyter has joined
1213 2013-06-10 17:13:45 qLooper_ has joined
1214 2013-06-10 17:13:57 ralphtheninja has joined
1215 2013-06-10 17:15:20 patcon has joined
1216 2013-06-10 17:16:07 daybyter has quit (Excess Flood)
1217 2013-06-10 17:16:13 nsh has quit (Excess Flood)
1218 2013-06-10 17:16:20 nsh_ has joined
1219 2013-06-10 17:16:27 daybyter has joined
1220 2013-06-10 17:17:07 Thepok has joined
1221 2013-06-10 17:17:43 nsh_ is now known as nsh
1222 2013-06-10 17:17:54 anarchy5 has quit (Quit: Computer has gone to sleep.)
1223 2013-06-10 17:20:16 qLooper_ has quit (Quit: Page closed)
1224 2013-06-10 17:23:30 <melvster> question : why does difficulty rebalance every 2 weeks rather than a shorter period?
1225 2013-06-10 17:23:37 Thepok has quit (Ping timeout: 245 seconds)
1226 2013-06-10 17:23:59 <melvster> like 1 week
1227 2013-06-10 17:24:21 ThomasV has quit (Quit: Leaving)
1228 2013-06-10 17:29:38 <CodeShark> the variance is pretty large - if the period were too short it would be too subject to statistical flukes
1229 2013-06-10 17:30:57 <CodeShark> dunno how the 2 week period was decided upon - but I suppose one could figure out how long you need to have a higher than p probability that the error is less than s
1230 2013-06-10 17:32:25 BlackPrapor has joined
1231 2013-06-10 17:36:15 <gmaxwell> Longer times also increase the work factor for an attacker to mine down a fork to a low difficulty to feed to network isolated nodes.
1232 2013-06-10 17:38:36 brson has joined
1233 2013-06-10 17:40:37 debiantoruser has quit (Ping timeout: 264 seconds)
1234 2013-06-10 17:41:10 robocoin has joined
1235 2013-06-10 17:41:10 robocoin has quit (Changing host)
1236 2013-06-10 17:41:10 robocoin has joined
1237 2013-06-10 17:41:59 debiantoruser has joined
1238 2013-06-10 17:42:05 peetaur2 has joined
1239 2013-06-10 17:44:42 <helo> bitcoin could survive forever in its current incarnation, and i would really like to see how it would evolve if the block size was left alone. i'm starting to think i won't get to though :(
1240 2013-06-10 17:46:43 GeorgeJ has joined
1241 2013-06-10 17:47:23 chorao has quit (Ping timeout: 255 seconds)
1242 2013-06-10 17:47:43 peetaur2 has quit (Ping timeout: 248 seconds)
1243 2013-06-10 17:48:33 michagogo has joined
1244 2013-06-10 17:49:08 alphaguru has quit ()
1245 2013-06-10 17:49:48 debiantoruser has quit (Ping timeout: 264 seconds)
1246 2013-06-10 17:51:23 peetaur2 has joined
1247 2013-06-10 17:51:26 debiantoruser has joined
1248 2013-06-10 17:53:51 GeorgeJ has quit (Remote host closed the connection)
1249 2013-06-10 17:56:28 <helo> at 7000 tps in a cash replacement global-scale scenario, how many non-miners will run validating nodes given that they may require 112MBit/s (if each tx is 250 bytes) of outbound traffic?
1250 2013-06-10 17:56:48 SchmalzTech has joined
1251 2013-06-10 17:57:00 debiantoruser has quit (Ping timeout: 264 seconds)
1252 2013-06-10 17:57:34 <helo> that's assuming 8 peers that they share their transactions with... not sure if that's an accurate reflection of outbound traffic relative to transaction rate...
1253 2013-06-10 17:58:44 aspect__ has joined
1254 2013-06-10 17:59:00 <helo> i guess on average half of your peers will have already seen the transactions you see? so half of 112MBit/s?
1255 2013-06-10 17:59:35 debiantoruser has joined
1256 2013-06-10 17:59:41 akl has joined
1257 2013-06-10 17:59:41 akl has quit (Changing host)
1258 2013-06-10 17:59:41 akl has joined
1259 2013-06-10 18:00:09 a_meteor has joined
1260 2013-06-10 18:00:15 chorao has joined
1261 2013-06-10 18:00:16 chorao has quit (Changing host)
1262 2013-06-10 18:00:16 chorao has joined
1263 2013-06-10 18:00:31 weex_ has joined
1264 2013-06-10 18:00:32 digitalmagus2 has joined
1265 2013-06-10 18:00:32 digitalmagus2 has quit (Changing host)
1266 2013-06-10 18:00:32 digitalmagus2 has joined
1267 2013-06-10 18:00:34 avarab has joined
1268 2013-06-10 18:00:35 avarab has quit (Changing host)
1269 2013-06-10 18:00:35 avarab has joined
1270 2013-06-10 18:00:37 dugo_ has joined
1271 2013-06-10 18:00:39 o3u_ has joined
1272 2013-06-10 18:00:39 sanders_ has joined
1273 2013-06-10 18:00:40 nowan has joined
1274 2013-06-10 18:01:32 crank_ has joined
1275 2013-06-10 18:02:48 yubrew has joined
1276 2013-06-10 18:02:49 amiller_ has joined
1277 2013-06-10 18:03:23 reizuki_ has joined
1278 2013-06-10 18:03:23 reizuki_ has quit (Changing host)
1279 2013-06-10 18:03:23 reizuki_ has joined
1280 2013-06-10 18:04:14 whiterabbit has joined
1281 2013-06-10 18:05:03 dugo_ has quit (Ping timeout: 252 seconds)
1282 2013-06-10 18:05:45 tumak_ has joined
1283 2013-06-10 18:05:50 [7] has quit (Disconnected by services)
1284 2013-06-10 18:05:59 TheSeven has joined
1285 2013-06-10 18:06:21 crank_ is now known as crank
1286 2013-06-10 18:06:23 whiterabbit is now known as wrabbit
1287 2013-06-10 18:06:29 jurov_ has joined
1288 2013-06-10 18:08:09 Guest83631 has joined
1289 2013-06-10 18:08:09 variousnefarious has joined
1290 2013-06-10 18:08:09 marcop has joined
1291 2013-06-10 18:08:09 hnz has joined
1292 2013-06-10 18:08:34 <runeks> helo: Well the answer to that depends on how far out into the future this scenario takes place. I imagine a 100 Mbit line will be quite common in 10 years. So if it doesn't happen until then we'll be fine.
1293 2013-06-10 18:09:01 jurov_ is now known as jurov
1294 2013-06-10 18:11:22 <TheUni> warren: thanks a bunch
1295 2013-06-10 18:11:43 chorao has quit (Read error: Connection reset by peer)
1296 2013-06-10 18:12:07 chorao has joined
1297 2013-06-10 18:12:08 chorao has quit (Changing host)
1298 2013-06-10 18:12:08 chorao has joined
1299 2013-06-10 18:12:12 <runeks> helo: Bear in mind that we're at roughly 0.7 TPS right now. An increase of 10,000x will surely take some time. And the "cash replacement global-scale scenario" will surely take *at least* 20 years. And by then I imagine we'll all be sitting on 1 Gbit lines (at least) with 100 TB hard drives.
1300 2013-06-10 18:14:24 ivan\ has joined
1301 2013-06-10 18:14:24 dugo has joined
1302 2013-06-10 18:14:33 <helo> yeah, that's sounds reasonable
1303 2013-06-10 18:16:31 fourlights has joined
1304 2013-06-10 18:17:22 fourlights_ has quit (Ping timeout: 245 seconds)
1305 2013-06-10 18:17:34 cypher has joined
1306 2013-06-10 18:17:35 gdbz has joined
1307 2013-06-10 18:21:46 LainZ has joined
1308 2013-06-10 18:25:09 stalled has joined
1309 2013-06-10 18:26:12 xnptzk` has quit (Read error: Connection reset by peer)
1310 2013-06-10 18:26:47 tyn has joined
1311 2013-06-10 18:28:17 nikolaj has joined
1312 2013-06-10 18:32:18 MobiusL has quit (Quit: Ex-Chat)
1313 2013-06-10 18:33:26 CodesInChaos has quit (Ping timeout: 246 seconds)
1314 2013-06-10 18:34:43 Mobius_ has joined
1315 2013-06-10 18:35:21 Mobius_ is now known as MobiusL
1316 2013-06-10 18:35:42 Tantadruj has quit (Ping timeout: 245 seconds)
1317 2013-06-10 18:36:16 denom has joined
1318 2013-06-10 18:37:42 vigilyn has left ("Leaving")
1319 2013-06-10 18:38:38 Tantadruj has joined
1320 2013-06-10 18:38:56 ralphtheninja has quit (Quit: leaving)
1321 2013-06-10 18:39:31 [Author] has quit (Ping timeout: 256 seconds)
1322 2013-06-10 18:39:36 vigilyn has joined
1323 2013-06-10 18:40:02 a_meteor is now known as a_meteorite
1324 2013-06-10 18:41:39 malexmedia has joined
1325 2013-06-10 18:42:08 [Author] has joined
1326 2013-06-10 18:43:44 saulimus has quit (Read error: Connection reset by peer)
1327 2013-06-10 18:44:05 saulimus has joined
1328 2013-06-10 18:44:23 digitalmagus has joined
1329 2013-06-10 18:44:43 Grouver has joined
1330 2013-06-10 18:44:52 Neozonz has quit (Ping timeout: 245 seconds)
1331 2013-06-10 18:44:52 digitalmagus has joined
1332 2013-06-10 18:45:38 nimdAHK_ has quit (Read error: Connection reset by peer)
1333 2013-06-10 18:46:03 nimdAHK_ has joined
1334 2013-06-10 18:47:05 wamatt has quit (Ping timeout: 246 seconds)
1335 2013-06-10 18:47:23 owowo has joined
1336 2013-06-10 18:48:29 danieldaniel has joined
1337 2013-06-10 18:48:44 <danieldaniel> Some chars can't be in btc addresses because they're not in base32, right/
1338 2013-06-10 18:48:44 <danieldaniel> ?
1339 2013-06-10 18:49:01 <danieldaniel> (sorry for two lines)
1340 2013-06-10 18:49:51 wamatt has joined
1341 2013-06-10 18:51:40 BeeDeePee has quit (Remote host closed the connection)
1342 2013-06-10 18:52:15 lolcookie__ has quit (Read error: Connection reset by peer)
1343 2013-06-10 18:52:26 BeeDeePee has joined
1344 2013-06-10 18:52:32 lolcookie__ has joined
1345 2013-06-10 18:52:57 Chat2092 has joined
1346 2013-06-10 18:53:30 Chat2092 has quit (Client Quit)
1347 2013-06-10 18:53:51 danieldaniel has quit (Quit: Page closed)
1348 2013-06-10 18:54:02 crank has joined
1349 2013-06-10 18:54:12 Guest47017 has quit (Quit: Page closed)
1350 2013-06-10 18:54:58 brson has quit (Ping timeout: 276 seconds)
1351 2013-06-10 18:56:30 Thepok has joined
1352 2013-06-10 18:56:45 Neozonz has joined
1353 2013-06-10 18:57:25 caedes has quit (Ping timeout: 264 seconds)
1354 2013-06-10 19:01:21 <lianj> digitalmagus: yes, only base58 chars
1355 2013-06-10 19:01:34 PiZZaMaN2K has joined
1356 2013-06-10 19:01:55 PiZZaMaN2K has quit (Changing host)
1357 2013-06-10 19:01:55 PiZZaMaN2K has joined
1358 2013-06-10 19:02:00 BCBot has quit (Remote host closed the connection)
1359 2013-06-10 19:02:22 BCBot has joined
1360 2013-06-10 19:02:29 mhanne_ has joined
1361 2013-06-10 19:04:25 oleganza has joined
1362 2013-06-10 19:05:36 mhanne_ is now known as mhanne
1363 2013-06-10 19:07:09 cdecker_ has joined
1364 2013-06-10 19:08:18 brson has joined
1365 2013-06-10 19:10:50 Toresh has joined
1366 2013-06-10 19:11:38 GordonG3kko has joined
1367 2013-06-10 19:12:25 shesek has joined
1368 2013-06-10 19:12:46 MKCoin has joined
1369 2013-06-10 19:13:29 agnostic98 has joined
1370 2013-06-10 19:13:29 Diapolis has joined
1371 2013-06-10 19:13:29 gjj has joined
1372 2013-06-10 19:17:14 qeb has joined
1373 2013-06-10 19:17:15 qeb has quit (Remote host closed the connection)
1374 2013-06-10 19:17:46 qeb has joined
1375 2013-06-10 19:18:15 77CAACAGL has joined
1376 2013-06-10 19:18:15 ppcko has joined
1377 2013-06-10 19:18:15 jordandotdev has joined
1378 2013-06-10 19:18:15 datagutt has joined
1379 2013-06-10 19:18:25 datagutt has quit (Max SendQ exceeded)
1380 2013-06-10 19:20:30 ovidiusoft has joined
1381 2013-06-10 19:20:34 daybyter has quit (Read error: Connection reset by peer)
1382 2013-06-10 19:22:00 tonikt has joined
1383 2013-06-10 19:23:17 tonikt has quit (Client Quit)
1384 2013-06-10 19:24:02 wiretapped has joined
1385 2013-06-10 19:24:35 guruvan has joined
1386 2013-06-10 19:25:35 sensorii has joined
1387 2013-06-10 19:26:22 random_cat has joined
1388 2013-06-10 19:28:58 beethoven2 has joined
1389 2013-06-10 19:31:16 tonikt has joined
1390 2013-06-10 19:31:18 mologie has quit (Quit: ZNC - http://znc.in)
1391 2013-06-10 19:31:50 mologie has joined
1392 2013-06-10 19:32:28 lolcookie__ has quit (Read error: Connection reset by peer)
1393 2013-06-10 19:33:27 richcollins has quit (Quit: richcollins)
1394 2013-06-10 19:33:36 nikolaj has quit (Ping timeout: 264 seconds)
1395 2013-06-10 19:34:17 Maxvalor has joined
1396 2013-06-10 19:34:58 oleganza has quit (Quit: oleganza)
1397 2013-06-10 19:35:16 Thepok has quit (Ping timeout: 276 seconds)
1398 2013-06-10 19:40:16 phma has quit (Ping timeout: 240 seconds)
1399 2013-06-10 19:41:41 chorao has quit (Ping timeout: 255 seconds)
1400 2013-06-10 19:46:04 chorao has joined
1401 2013-06-10 19:46:04 chorao has quit (Changing host)
1402 2013-06-10 19:46:05 chorao has joined
1403 2013-06-10 19:46:25 lmacken has joined
1404 2013-06-10 19:47:41 paracyst has joined
1405 2013-06-10 19:49:20 denom has quit (Quit: Leaving)
1406 2013-06-10 19:49:51 RoboTedd_ has quit (Remote host closed the connection)
1407 2013-06-10 19:50:26 agnostic98 has quit (Remote host closed the connection)
1408 2013-06-10 19:51:13 Thepok has joined
1409 2013-06-10 19:51:50 debiantoruser has quit (Ping timeout: 246 seconds)
1410 2013-06-10 19:54:51 sipa has joined
1411 2013-06-10 19:57:37 tonikt has quit (Quit: Leaving)
1412 2013-06-10 19:57:48 patcon has quit (Remote host closed the connection)
1413 2013-06-10 19:58:15 patcon has joined
1414 2013-06-10 20:00:44 Adrastia has joined
1415 2013-06-10 20:00:56 gjj has quit (Remote host closed the connection)
1416 2013-06-10 20:01:10 <warren> TheUni: was linux gitian tested?
1417 2013-06-10 20:01:31 <TheUni> warren: not yet. win32 is the hard one, i was working on it first
1418 2013-06-10 20:01:47 <TheUni> linux gitian should basically just be a c/p of the ./configure procedure without all of the extra crap
1419 2013-06-10 20:02:05 <sipa> yeah, i expect the linux gitian descriptors to become trivial
1420 2013-06-10 20:02:17 <TheUni> it should basically be ./configure && make
1421 2013-06-10 20:02:39 <TheUni> warren: would you mind trying to move the faketime stuff to after configure, and see if there are any problems with it
1422 2013-06-10 20:02:42 <TheUni> ?
1423 2013-06-10 20:02:49 patcon has quit (Ping timeout: 240 seconds)
1424 2013-06-10 20:03:22 <warren> TheUni: I didn't check if the current libfaketime hack results in a deterministic build or not
1425 2013-06-10 20:03:30 <warren> I had barely 3 hours of sleep.
1426 2013-06-10 20:03:37 <warren> I'll check later.
1427 2013-06-10 20:03:40 <TheUni> ok
1428 2013-06-10 20:03:45 <TheUni> linux is deterministic without it
1429 2013-06-10 20:03:51 <warren> TheUni: any idea why I couldn't send you a pull request?
1430 2013-06-10 20:03:59 <TheUni> win32 only isn't because the toolchain has an ancient binutils
1431 2013-06-10 20:04:34 <sipa> imho we should upgrade (at least) the win32 env to a more recent ubuntu
1432 2013-06-10 20:04:52 <sipa> there is absolutely no reason to stick to a 3 year old OS for that
1433 2013-06-10 20:04:58 <TheUni> well depending on ubuntu is kinda silly to begin with. should be using a linaro toolchain
1434 2013-06-10 20:05:27 <TheUni> or any readily-available pre-packaged toolchain
1435 2013-06-10 20:05:29 <sipa> never heard of that
1436 2013-06-10 20:05:33 <sipa> seems for ARM?
1437 2013-06-10 20:06:08 <TheUni> er, right, heh. i forgot which world i was in for a min :)
1438 2013-06-10 20:06:18 <TheUni> s/linaro/linaro-ish/
1439 2013-06-10 20:06:43 <TheUni> point is, it should be a static toolchain rather than some OS package that could change on a whim
1440 2013-06-10 20:07:21 <michagogo> sipa: I thought the choice of lucid was a gitian thing, not a bitcoin thing?
1441 2013-06-10 20:07:31 <sipa> michagogo: eh, no
1442 2013-06-10 20:08:14 debiantoruser has joined
1443 2013-06-10 20:08:19 <michagogo> sipa: Could have sworn that when I cloned gitian-builder and ran make-base-vm it gave me lucid
1444 2013-06-10 20:08:27 <michagogo> Or was that a flag? I don't remember at this point
1445 2013-06-10 20:08:32 <Luke-Jr> TheUni: static vs shared libraries is not a toolchain thing
1446 2013-06-10 20:08:57 o-p has joined
1447 2013-06-10 20:08:59 <TheUni> Luke-Jr: hmm? where did that come from?
1448 2013-06-10 20:09:12 <Luke-Jr> or you mean in case Ubuntu changes the lucid compiler packages? O.o
1449 2013-06-10 20:09:22 <TheUni> yes, that
1450 2013-06-10 20:09:28 <Luke-Jr> ok
1451 2013-06-10 20:09:29 <TheUni> static package, not static results
1452 2013-06-10 20:09:42 <Luke-Jr> really, gitian should be like Gentoo and build everything from source ;p
1453 2013-06-10 20:09:51 <jgarzik> guffaw
1454 2013-06-10 20:09:54 <Luke-Jr> but things take time
1455 2013-06-10 20:10:14 <TheUni> Luke-Jr: working on it
1456 2013-06-10 20:10:45 <sipa> michagogo: lucid is the default, yes
1457 2013-06-10 20:10:46 <gmaxwell> TheUni: "some OS package that could change on a whim" huh?
1458 2013-06-10 20:11:24 <michagogo> (note that I know nearly nothing about gitian -- I just run the scripts)
1459 2013-06-10 20:11:32 <TheUni> gmaxwell: http://changelogs.ubuntu.com/changelogs/pool/main/b/binutils/binutils_2.20.1-3ubuntu7.1/changelog
1460 2013-06-10 20:11:39 <TheUni> that's binutils alone
1461 2013-06-10 20:11:42 <gmaxwell> TheUni: the whole idea of gitian is to create a static encapsulated build enviroment which is adequate to create a determinstic binary.
1462 2013-06-10 20:11:50 <gmaxwell> TheUni: yes, and? Whats the relevance?
1463 2013-06-10 20:12:02 <TheUni> gmaxwell: i understand. i'm arguing that it operates on flawed assumptions
1464 2013-06-10 20:12:13 <gmaxwell> TheUni: I think your understanding is in error.
1465 2013-06-10 20:12:36 atweiden has joined
1466 2013-06-10 20:12:38 <sipa> i don't think he's wrong
1467 2013-06-10 20:12:52 <Luke-Jr> gmaxwell: I think his point is that someone could bump the binutils package in lucid with a trojan-injector, and all our builds would match with a trojan included
1468 2013-06-10 20:12:54 patcon has joined
1469 2013-06-10 20:12:55 <gmaxwell> TheUni: The fact that someone someplace else changes things doesn't make your builder VMs change.
1470 2013-06-10 20:13:05 ThomasV has joined
1471 2013-06-10 20:13:06 <Luke-Jr> gmaxwell: gitian does a package update every build
1472 2013-06-10 20:13:14 <gmaxwell> Ah, well thats moronic.
1473 2013-06-10 20:13:26 <TheUni> gmaxwell: please explain. you build gitian from up-to-date ubuntu. ubuntu updates their gcc, and i do a gitian build with the current fresh ubuntu. our binaries don't match.
1474 2013-06-10 20:13:28 <gmaxwell> I didn't realize it did that, my apologies.
1475 2013-06-10 20:13:55 <Luke-Jr> I'm not sure it updates *everything*, but at least some things do
1476 2013-06-10 20:14:08 <TheUni> as i keep saying, gitian goes out of its way to do extra work in order to produce binaries that are only slightly more likely to be bit-exact
1477 2013-06-10 20:14:27 <Luke-Jr> I'm almost certain there'd be a way to infect the binaries if someone at Ubuntu wanted to
1478 2013-06-10 20:14:39 <TheUni> Luke-Jr: that's by far not the weakest link
1479 2013-06-10 20:14:59 enquirer has quit (Quit: Nettalk6 - www.ntalk.de)
1480 2013-06-10 20:15:00 <sipa> TheUni: what is the weakest link in your opinion?
1481 2013-06-10 20:15:05 <gmaxwell> TheUni: I think thats an orthorgonal issue. The auto updating is a silly weakness, but its not necessary to keep it.
1482 2013-06-10 20:15:09 <TheUni> there are several huge targets that would be much more trivial to take advantage of
1483 2013-06-10 20:15:17 <michagogo> Luke-Jr: Someone at Ubuntu?
1484 2013-06-10 20:15:23 enquirer has joined
1485 2013-06-10 20:15:24 <michagogo> Do you mean Canonical?
1486 2013-06-10 20:15:30 <Luke-Jr> michagogo: whoever has access to the signing keys
1487 2013-06-10 20:15:49 qeb has quit (Quit: ["Textual IRC Client: www.textualapp.com"])
1488 2013-06-10 20:16:04 <TheUni> gmaxwell: it's not orthogonal at all. gitian has a single purpose. it does a really bad job of making assumptions to achieve that purpose
1489 2013-06-10 20:16:12 stochasm has quit (Ping timeout: 256 seconds)
1490 2013-06-10 20:16:28 <TheUni> rather than addressing those goals, i'm taking the approach of assuming that gitian is unneccessary, and working from there
1491 2013-06-10 20:16:43 <TheUni> sipa: i'd prefer not to say
1492 2013-06-10 20:16:51 <gmaxwell> TheUni: ... If the build script doesn't foolishly apply updates and just builds from a static enviroment then its no less determinstic than any other static build enviroment.
1493 2013-06-10 20:17:04 <Luke-Jr> gmaxwell: the static environment is missing packages we need
1494 2013-06-10 20:17:10 * nsh is now convinced gitian is a silly thing
1495 2013-06-10 20:17:19 <nsh> and should be replaced by something that does what it attempt to do
1496 2013-06-10 20:17:22 <TheUni> gmaxwell: correct. you've just made my point :)
1497 2013-06-10 20:17:38 <gmaxwell> TheUni: Your point was pointless.
1498 2013-06-10 20:17:55 enquirer has quit (Client Quit)
1499 2013-06-10 20:18:02 <sipa> so, if not gitian, how do you propose to achieve a static build environment?
1500 2013-06-10 20:18:06 <gmaxwell> So comment out the stupid command and thereby remove the problem. You haven't offered an alternative which is actually successful at getting a repeatable determintsic build.
1501 2013-06-10 20:18:08 * Luke-Jr thinks we're all on the same page now, regardless of who pointed what
1502 2013-06-10 20:18:12 <nsh> (a point that contained a point would be dimensionally perverse)
1503 2013-06-10 20:18:19 <TheUni> gmaxwell: i've said repeatedly, i'm working on one
1504 2013-06-10 20:18:36 <sipa> wait, what?
1505 2013-06-10 20:18:43 <TheUni> gmaxwell: i've submitted several pull requests which are prerequisites
1506 2013-06-10 20:18:50 <gmaxwell> ...
1507 2013-06-10 20:19:06 <gmaxwell> autotools != a determinstic build enviroment.
1508 2013-06-10 20:19:30 agnostic98 has joined
1509 2013-06-10 20:19:34 <TheUni> gmaxwell: autotools = an abstracted build environment which allows for a single build procedure
1510 2013-06-10 20:19:38 <TheUni> and yes, the results are deterministic
1511 2013-06-10 20:19:49 <sipa> given a deterministic host, the result will be, sure
1512 2013-06-10 20:19:57 <sipa> but how do you achieve the former?
1513 2013-06-10 20:20:35 <nsh> VM
1514 2013-06-10 20:20:37 <TheUni> no need for a deterministic host, only a deterministic self-hosting toolchain
1515 2013-06-10 20:20:38 <TheUni> NO.
1516 2013-06-10 20:21:10 stochasm has joined
1517 2013-06-10 20:21:13 RoboTeddy has joined
1518 2013-06-10 20:21:30 RoboTeddy has quit (Read error: Connection reset by peer)
1519 2013-06-10 20:21:31 yubrew has quit (Remote host closed the connection)
1520 2013-06-10 20:21:39 <gmaxwell> TheUni: and yes, have you actually ever installed a complete cross compilation enviroment from scratch? it's an enormous burden. A static OS install in a VM is a much more straightforward way to get there, as far as I can tell.
1521 2013-06-10 20:22:00 RoboTeddy has joined
1522 2013-06-10 20:22:02 <Luke-Jr> gmaxwell: it is?
1523 2013-06-10 20:22:07 <jgarzik> yes
1524 2013-06-10 20:22:08 <gmaxwell> Of courseâ I'm open to being shown otherwise. But the way to do that is to ... do that, not to sit around on IRC idly attacking stuff that actually works.
1525 2013-06-10 20:22:11 <Luke-Jr> I just type crossdev -t <triplet>
1526 2013-06-10 20:22:22 <sipa> gmaxwell: well, he is working :)
1527 2013-06-10 20:22:29 <TheUni> gmaxwell: i'm intimately familiar with such environments, it's what i've done for $dayjob for years
1528 2013-06-10 20:22:48 <nsh> great, everything's settled then.
1529 2013-06-10 20:22:51 <Luke-Jr> works for everything except Mac :P
1530 2013-06-10 20:22:55 enquirer has joined
1531 2013-06-10 20:22:58 <Luke-Jr> and Android
1532 2013-06-10 20:22:59 chorao has quit (Ping timeout: 256 seconds)
1533 2013-06-10 20:23:04 <TheUni> Luke-Jr: works on both
1534 2013-06-10 20:23:23 Jackneill has quit (Ping timeout: 248 seconds)
1535 2013-06-10 20:23:28 <Luke-Jr> TheUni: crossdev doesn't, I mean
1536 2013-06-10 20:23:29 <gmaxwell> TheUni: This needs to be accessible to a larger audience than the small set of cross compiling development ninjas.
1537 2013-06-10 20:23:50 <gmaxwell> Luke-Jr: that isn't from scratch, and so it won't get you the same enviroment that I have. (unless I run the same OS as you)
1538 2013-06-10 20:23:52 <TheUni> mac builds fine, android obviously fails to build due to toolchain issues. i've asked about providing android patches, but there seemed to be little interest
1539 2013-06-10 20:24:09 <Luke-Jr> gmaxwell: what do you mean from scratch then?
1540 2013-06-10 20:24:14 B0g4r7 has quit (Remote host closed the connection)
1541 2013-06-10 20:24:14 B0g4r7_ is now known as B0g4r7
1542 2013-06-10 20:24:16 <Luke-Jr> gmaxwell: crossdev builds everything from source
1543 2013-06-10 20:24:27 <TheUni> gmaxwell: ok, i agree with you. it's senseless to discuss. please understand that i'm working to provide a POC, and assume that i'm capable of doing so
1544 2013-06-10 20:24:36 <sipa> TheUni: i think you've been very mysterious about what your ideal end solution would look like
1545 2013-06-10 20:24:38 <TheUni> if it's not interesting, that's fine, i'll understand
1546 2013-06-10 20:24:41 <sipa> TheUni: i still have no clue
1547 2013-06-10 20:25:03 <TheUni> sipa: i even gave you a git repo to clone and try :)
1548 2013-06-10 20:25:12 <gmaxwell> TheUni: it's interesting, of courseâ but dear lord, I'd hate to see you waste a bunch of time on something non-viable because you missed some of the requirements.
1549 2013-06-10 20:25:21 <sipa> TheUni: hmm?
1550 2013-06-10 20:25:41 <nsh> why aren't these things hashed out in some shared environment, like a section of the wiki?
1551 2013-06-10 20:25:57 <nsh> to prevent wasted/replicated labour
1552 2013-06-10 20:26:04 <TheUni> sipa: i've done this before for a project with ~60 dependencies. requirements were the same, except that it also had to support ios/android/bsd/etc
1553 2013-06-10 20:26:11 <gmaxwell> nsh: because then we wouldn't have the pleasure of you constantly interrupting the discussion?
1554 2013-06-10 20:26:13 <warren> Luke-Jr: I've been told to ask you about your previous efforts on a linux-based cross compiler to macos target.
1555 2013-06-10 20:26:20 <warren> Luke-Jr: how far did you get?
1556 2013-06-10 20:26:33 <Luke-Jr> warren: worked until bitcoind 0.7 I think
1557 2013-06-10 20:26:33 <nsh> gmaxwell, trust me, the pleasure is all mine.
1558 2013-06-10 20:26:33 metabyte has quit (Read error: Connection reset by peer)
1559 2013-06-10 20:26:47 <Luke-Jr> warren: https://gitorious.org/cross-osx/cross-osx
1560 2013-06-10 20:26:51 <gmaxwell> TheUni: "This" ? You mean built a public enviroment where independant parties were able to reliably produce identical binaries?
1561 2013-06-10 20:27:03 metabyte has joined
1562 2013-06-10 20:27:27 <gmaxwell> TheUni: if so, please link to it. ... if it wasn't public, then the requirements really were not the same.
1563 2013-06-10 20:27:50 smiddi has quit (Remote host closed the connection)
1564 2013-06-10 20:27:50 <warren> Luke-Jr: we're willing to contribute to a bounty to fix this and make it robust, do you think Bitcoin people can contribute as well?
1565 2013-06-10 20:28:24 <sipa> TheUni: i know you worked on xbmc (iirc?), but i'm not sure we're talking about the same thing still
1566 2013-06-10 20:28:30 <Luke-Jr> warren: nothing from that previous effort is useful to get it working today
1567 2013-06-10 20:28:45 <warren> Luke-Jr: oh, why is that?
1568 2013-06-10 20:28:46 <Luke-Jr> warren: it's on my todo list to get something working now, but pending on some other efforts
1569 2013-06-10 20:28:51 <TheUni> gmaxwell: ok, fair enough, the requirements are different
1570 2013-06-10 20:28:54 saulimus has quit (Quit: saulimus)
1571 2013-06-10 20:28:56 <Luke-Jr> warren: the previous effort was based on GCC, and Apple is using LLVM now
1572 2013-06-10 20:29:25 <TheUni> ok, i'll not discuss it here anymore. seems it only attracts negative attention
1573 2013-06-10 20:29:29 <warren> ok, so it would be a serious pain to get it to work again
1574 2013-06-10 20:29:35 <sipa> TheUni: then stop being mysterious!
1575 2013-06-10 20:29:43 <gmaxwell> TheUni: in any case, I think whatever you to do make the build enviroment better also helps something like gitian even if we don't change how that works... so I'm certantly happy to see improvements there.
1576 2013-06-10 20:29:58 <sipa> TheUni: i want to know what you envision
1577 2013-06-10 20:29:59 <TheUni> sipa: i'm not. i've said over and over what my plan is
1578 2013-06-10 20:30:06 <sipa> i must have missed it
1579 2013-06-10 20:30:19 <TheUni> sipa: git clone git@github.com:xbmc/xbmc.git
1580 2013-06-10 20:30:25 <TheUni> cd tools/depends && ./configure --help
1581 2013-06-10 20:30:35 <warren> TheUni: does your account block pull requests or something? I can't for reasons I don't understand.
1582 2013-06-10 20:30:49 <sipa> TheUni: if we both run that, it would certainly surprise me very much if we end up with identical binaries
1583 2013-06-10 20:30:54 <TheUni> warren: unsure
1584 2013-06-10 20:31:08 <Luke-Jr> TheUni: if you can help me get an OS X cross-toolchain going, I'd appreciate it :D
1585 2013-06-10 20:31:35 <jgarzik> hah
1586 2013-06-10 20:31:39 <jgarzik> would be nice
1587 2013-06-10 20:31:40 <TheUni> sipa: yes, the deterministic bit is new, but the procedure is the same. i've verified that it can be made to be deterministic. that's what i'm working on
1588 2013-06-10 20:31:50 <TheUni> Luke-Jr: to avoid the need for a native builder?
1589 2013-06-10 20:32:00 <sipa> TheUni: that's not an answer (though certainly a massive improvement on its own)
1590 2013-06-10 20:32:15 <sipa> TheUni: it doesn't explain how we'll end up with an identical toolchain in a way that can be verified
1591 2013-06-10 20:32:28 <TheUni> sipa: wget $toolchain
1592 2013-06-10 20:32:33 <Luke-Jr> TheUni: because we need to build it from Ubuntu
1593 2013-06-10 20:32:47 <Ry4an> [dev.drama9.com] out: ValueError: Unable to configure handler 'log_file': [Errno 13] Permission denied: '/persistent/dramafever/branch/run/django.log'
1594 2013-06-10 20:32:48 <sipa> TheUni: who decides $toolchain?
1595 2013-06-10 20:32:59 <TheUni> sipa: that's why i was commenting on grabbing a static toolchain rather than depending on an ubuntu package
1596 2013-06-10 20:33:17 <sipa> well that is no solution, unless there is a reason to trust that toolchain
1597 2013-06-10 20:33:34 <TheUni> sipa: that's not a fair question. the answer right now is "canonical". the answer should be "we do".
1598 2013-06-10 20:33:52 <Luke-Jr> â¦
1599 2013-06-10 20:34:02 <gmaxwell> TheUni: "we do" works only in a space of sufficiently obvious answers.
1600 2013-06-10 20:34:11 <sipa> agree, we now trust canonical to an extent
1601 2013-06-10 20:34:18 <TheUni> if you want to grab the same canonical toolchain as before to avoid having to answer the question, then that's a possibility as well
1602 2013-06-10 20:34:24 <gmaxwell> ::nods::
1603 2013-06-10 20:34:24 o-p is now known as akrmn
1604 2013-06-10 20:34:27 <Luke-Jr> "trust canonical for compiler" is better than "trust <random bitcoin developer> for compiler"
1605 2013-06-10 20:34:44 <gmaxwell> TheUni: okay, I'm happy now that we all basically understand each other!
1606 2013-06-10 20:35:08 <TheUni> who said that? i certainly wouldn't trust some random developer. i'm simply suggesting that it's a static toolchain that comes from a trusted source rather than one that's implied to be safe because it was never considered
1607 2013-06-10 20:35:22 <akrmn> the real way is to use something like this: http://bellard.org/tcc/tccboot.html
1608 2013-06-10 20:35:32 <sipa> ideally, the reason a toolchain is trusted, is because everyone is already using it
1609 2013-06-10 20:35:47 <TheUni> sipa: ok, then wget $current-toolchain. done.
1610 2013-06-10 20:35:52 <akrmn> but probably not possible yet
1611 2013-06-10 20:35:56 <TheUni> though
1612 2013-06-10 20:36:00 <gmaxwell> TheUni: huh? Thats a question which has been considered.
1613 2013-06-10 20:36:03 * Luke-Jr facepalms
1614 2013-06-10 20:36:18 <warren> "trust unpatched dependency library that we never update"
1615 2013-06-10 20:36:45 <Luke-Jr> no matter what toolchain we use, we need to update it eventually
1616 2013-06-10 20:36:59 <gmaxwell> warren: we already deal with thatâ we make a concious and controlled decision to update dependencies like boost from time to time.
1617 2013-06-10 20:37:09 <warren> ok
1618 2013-06-10 20:37:26 <gmaxwell> And the manner in which it's done (using well known widely used release versions) is a nothing-up-my-sleeve factor.
1619 2013-06-10 20:38:13 <sipa> TheUni: so basically the issue is reduced to "how do we establish a trusted build environment", and our answer to this was "gitian" (but gitian does way more than that, and probably shouldn't), and your answer is "it doesn't matter"
1620 2013-06-10 20:38:18 <warren> Is the old lucid environment updating things itself?
1621 2013-06-10 20:38:35 <TheUni> sipa: no. my answer is "it's not as trusted as you think"
1622 2013-06-10 20:38:55 <Luke-Jr> warren: every build fetches the latest packages for <dependencies named in gitian yml>
1623 2013-06-10 20:39:02 <TheUni> and, "the same level of trust could be achieved with infinitely less hassle"
1624 2013-06-10 20:39:11 <Luke-Jr> warren: those dependencies pull in their own dependencies, potentially including updated toolchain packages
1625 2013-06-10 20:39:17 <sipa> TheUni: imho, you've not given a better alternative yet
1626 2013-06-10 20:39:57 <sipa> by which i don't mean to say you're wrong
1627 2013-06-10 20:40:12 * Luke-Jr tempted to make a fully deterministic OS with Gentoo, but resists the urge to waste time <.<
1628 2013-06-10 20:40:25 <warren> TheUni: with moderate effort I got gitian to work on fedora, and Luke-Jr got it work on gentoo. with a little more effort it can boot on a macos host.
1629 2013-06-10 20:40:49 <TheUni> sipa: this is getting frustrating. i'm actively working on it. i'm making the changes required. i'm pointing out why i believe the work is necessary. and i'm doing my work openly and with an open mind. i'm not sure what else i can do...
1630 2013-06-10 20:40:56 malexmedia has quit (Ping timeout: 240 seconds)
1631 2013-06-10 20:41:03 normanrichards has joined
1632 2013-06-10 20:41:09 <sipa> TheUni: and you're working on making the build process awesome
1633 2013-06-10 20:41:18 <devrandom> TheUni: how will you get the result to be deterministic without a VM? the environment does affect the build
1634 2013-06-10 20:41:21 normanrichards has quit (Max SendQ exceeded)
1635 2013-06-10 20:41:26 * jgarzik has been reading and scrolling back, and is still not sure what is being /usefully/ argued
1636 2013-06-10 20:41:27 <jgarzik> :)
1637 2013-06-10 20:41:27 <sipa> TheUni: but i haven't seen a single suggestion how to establish a trusted host environment
1638 2013-06-10 20:41:37 <TheUni> jgarzik: heh, very little :)
1639 2013-06-10 20:41:38 <jgarzik> autoconf'ing is nice
1640 2013-06-10 20:41:49 <jgarzik> let's focus on close, near term steps
1641 2013-06-10 20:41:51 <TheUni> sipa: i'm arguing that you don't have to
1642 2013-06-10 20:42:03 <sipa> TheUni: if your suggestion is "wget $toolchain", sorry... that's handwaving
1643 2013-06-10 20:42:12 <warren> jgarzik: autoconf needs a --allow-insane-environment option =)
1644 2013-06-10 20:42:31 <sipa> TheUni: you're minimizing the thing we're caring about
1645 2013-06-10 20:42:49 <sipa> and solving everything else (in from what i can see, a very good way)
1646 2013-06-10 20:42:56 <TheUni> sipa: you're caring about something that i consider minimal for my work :)
1647 2013-06-10 20:43:01 <warren> jgarzik: https://github.com/wtogami/bitcoin/commit/0e222a3c480b77e863ba5b68099d05dc264fdff1
1648 2013-06-10 20:43:02 <sipa> TheUni: *bingo*
1649 2013-06-10 20:43:34 <jgarzik> heh
1650 2013-06-10 20:43:58 <sipa> TheUni: i love how much work you've already invested in this, and i really want these things merged
1651 2013-06-10 20:44:13 <TheUni> sipa: cross compilation assumes that the dev environment has nothing to do with the runtime environment
1652 2013-06-10 20:44:13 <sipa> TheUni: but you haven't solved "how to get a trusted build environment", except saying it doesn't matter
1653 2013-06-10 20:44:32 <TheUni> i'm applying that idea here to assume that everything is a cross build
1654 2013-06-10 20:44:52 <sipa> TheUni: of course, that's a great improvemtn
1655 2013-06-10 20:44:52 <TheUni> if there's a tool needed to generate a target binary, it should itself be generated from the build machine first.
1656 2013-06-10 20:45:10 <sipa> sure
1657 2013-06-10 20:45:20 <TheUni> for the same reason that arm binaries can be built from x86 machines, it is possible to build deterministic target binaries without needing a deterministic builder
1658 2013-06-10 20:45:29 <Luke-Jr> â¦â¦â¦â¦â¦
1659 2013-06-10 20:45:34 <TheUni> that's what i mean by self-hosting
1660 2013-06-10 20:45:47 <TheUni> and that's what i'm working to build.
1661 2013-06-10 20:45:52 <Luke-Jr> if it builds deterministic target binaries, it *is* a deterministic builder <.<
1662 2013-06-10 20:45:53 <sipa> well, if you're using gcc 4.6 and i use 4.7, there's no way we end up with the same binary
1663 2013-06-10 20:46:16 <TheUni> sipa: as i've said, wget $toolchain
1664 2013-06-10 20:46:29 <sipa> which completely sidesteps the question how that toolchain is trusted
1665 2013-06-10 20:46:30 <TheUni> and you can't say "that's hand-waving". because that's what gitian does now
1666 2013-06-10 20:46:34 <TheUni> it just does it in the form of an apt-get
1667 2013-06-10 20:46:49 <sipa> gitian relies on using a very popular operating system distribution to establish trust
1668 2013-06-10 20:46:52 cdecker_ has quit (Remote host closed the connection)
1669 2013-06-10 20:46:53 <sipa> it's certainly not perfect
1670 2013-06-10 20:46:53 <jgarzik> you -can- cross-compile, but really you want a native build
1671 2013-06-10 20:46:55 agnostic98 has quit (Remote host closed the connection)
1672 2013-06-10 20:47:04 <sipa> but it's far better than saying "download X"
1673 2013-06-10 20:47:08 <jgarzik> gcc is forced to emulate certain things, sometimes imperfectly, when host != target
1674 2013-06-10 20:47:25 cdecker_ has joined
1675 2013-06-10 20:47:33 <TheUni> jgarzik: it emulates nothing, it makes guesses
1676 2013-06-10 20:47:49 cdecker_ has quit (Remote host closed the connection)
1677 2013-06-10 20:48:13 caedes has joined
1678 2013-06-10 20:48:13 caedes has quit (Changing host)
1679 2013-06-10 20:48:13 caedes has joined
1680 2013-06-10 20:48:22 <jgarzik> TheUni, Not true. When gcc performs algebraic simplifications and the like, you must follow the CPU rules precisely
1681 2013-06-10 20:48:24 <dugo> someone who want super optimised build roles their own anyway
1682 2013-06-10 20:48:31 <dugo> +s
1683 2013-06-10 20:48:32 sanders_ has quit (Ping timeout: 248 seconds)
1684 2013-06-10 20:48:41 <jgarzik> to the point where you wind up emulating an FPU sometimes, depending on platform
1685 2013-06-10 20:48:48 <TheUni> jgarzik: yep, which is what -mcpu, -march, and -mtune are for
1686 2013-06-10 20:49:00 <jgarzik> or messy integer maths when bitness mismatch occurs
1687 2013-06-10 20:49:21 <jgarzik> TheUni, no, this is much more fundamental
1688 2013-06-10 20:49:22 <TheUni> jgarzik: again, that's set ahead of time. softabi/hardabi isn't guessed, it's set during toolchain production
1689 2013-06-10 20:49:26 <TheUni> and overridden if necessary
1690 2013-06-10 20:49:51 <sipa> TheUni: anyway, please keep doing what you're doing :)
1691 2013-06-10 20:50:53 <TheUni> jgarzik: same for optims. sure, if you use march=native, you're going to get builder=target binaries. but if you provide safe defaults (which any sane toolchain, including ubuntu's does), then you get pre-determinied abi compatibility
1692 2013-06-10 20:51:23 <jgarzik> TheUni, you and I are talking about two different things
1693 2013-06-10 20:51:37 <jgarzik> anyway, not relevant to #bitcoin-dev, getting back on topic
1694 2013-06-10 20:51:43 <jgarzik> continue autoconf'ing
1695 2013-06-10 20:51:47 yubrew has joined
1696 2013-06-10 20:52:00 <jgarzik> configure outside root is decidedly non-standard
1697 2013-06-10 20:52:03 <jgarzik> but not unheard of
1698 2013-06-10 20:52:12 <jgarzik> e.g. BDB sticks theirs in build/unix or somesuch.
1699 2013-06-10 20:52:23 <jgarzik> so slight preference for root
1700 2013-06-10 20:52:35 <jgarzik> that enables capturing of all files via "make dist"
1701 2013-06-10 20:52:36 <TheUni> yea, theirs is wonky. though i believe mysql/samba/others stick theirs in src/
1702 2013-06-10 20:52:54 <TheUni> jgarzik: yes, that was one of my questions...
1703 2013-06-10 20:53:01 <jgarzik> sticking in src/ makes capturing and installing productions outside that dir highly difficult
1704 2013-06-10 20:53:04 patcon has quit (Remote host closed the connection)
1705 2013-06-10 20:53:04 <jgarzik> thus,
1706 2013-06-10 20:53:06 <TheUni> 'make dist' currently works, but excludes all of that top-level stuff
1707 2013-06-10 20:53:07 <jgarzik> so slight preference for root
1708 2013-06-10 20:53:17 <TheUni> ok, fair enough
1709 2013-06-10 20:53:20 RazielZ has quit (Ping timeout: 248 seconds)
1710 2013-06-10 20:53:37 <TheUni> as i mentioned, my reasoning for doing a single makefile in src/ was for qt-creator
1711 2013-06-10 20:53:38 <Luke-Jr> TheUni: once upon a time, I had a need to test for roundl() support in my autoconf code
1712 2013-06-10 20:53:41 Diablo-D3 has quit (Quit: do coders dream of sheep()?)
1713 2013-06-10 20:53:44 <jgarzik> top level makes it trivial to include or exclude any dirs
1714 2013-06-10 20:53:53 <TheUni> so the question is: which of those is more important
1715 2013-06-10 20:54:01 <Luke-Jr> TheUni: but I found it was detecting roundl() even on cross-compilers where roundl() was not defined
1716 2013-06-10 20:54:03 <jgarzik> it's painful to reverse course later, if you start out in src/
1717 2013-06-10 20:54:10 <jgarzik> and then move autoconf files to root
1718 2013-06-10 20:54:22 <TheUni> (i don't care either way, i'm happy to change it to whatever is more desired)
1719 2013-06-10 20:54:31 <TheUni> jgarzik: ok, will do then
1720 2013-06-10 20:54:48 <Luke-Jr> TheUni: it turned out that GCC was completely removing the roundl(constant), and just figuring out the answer itself at compile-time. Thus, when it linked, roundl wasn't actually being needed/checked for
1721 2013-06-10 20:54:51 <TheUni> Luke-Jr: that's not a compiler issue, that's a libc/libm issue
1722 2013-06-10 20:54:59 Diablo-D3 has joined
1723 2013-06-10 20:55:13 <Luke-Jr> but it *was* a compiler issue âº
1724 2013-06-10 20:55:40 <sipa> well you were assuming compiling code that called a function resulted in a test for that function existing :p
1725 2013-06-10 20:55:41 sanders has joined
1726 2013-06-10 20:55:43 <TheUni> Luke-Jr: still, that's a compiler optim, and one that's predictable and deterministic
1727 2013-06-10 20:55:46 <Luke-Jr> this is what jgarzik meant above
1728 2013-06-10 20:55:51 Elmf has quit (Ping timeout: 256 seconds)
1729 2013-06-10 20:56:00 <Luke-Jr> by the compiler emulating the host platform
1730 2013-06-10 20:56:16 <TheUni> Luke-Jr: no, you've misunderstood what happened there
1731 2013-06-10 20:56:32 Maxvalor has quit ()
1732 2013-06-10 20:56:52 <warren> Luke-Jr: you intend on working on the mac cross-compile problem soon? or it's fine to pay my guy to do it?
1733 2013-06-10 20:57:19 <TheUni> warren: what is the desired goal?
1734 2013-06-10 20:57:46 <warren> TheUni: gitian environment to cross-compile mac builds in the same way we cross-compile win32
1735 2013-06-10 20:57:46 <TheUni> just to eliminate the need for the darwin builder?
1736 2013-06-10 20:58:14 <warren> TheUni: also port gitian to work on macos
1737 2013-06-10 20:58:26 <TheUni> warren: heh, seems those are direct opposite goals
1738 2013-06-10 20:58:43 <warren> how so?
1739 2013-06-10 20:59:11 <TheUni> nevermind, will try to stay on topic
1740 2013-06-10 20:59:12 <dugo> ..and trust macos?
1741 2013-06-10 20:59:21 <Luke-Jr> warren: gitian already works on Mac
1742 2013-06-10 20:59:30 <warren> Luke-Jr: ok good, less work
1743 2013-06-10 20:59:35 <TheUni> jgarzik: did you see my question about qt-creator above?
1744 2013-06-10 20:59:47 <warren> Luke-Jr: virtualbox based?
1745 2013-06-10 20:59:49 <sipa> TheUni: jgarzik has never built bitcoin-qt :p
1746 2013-06-10 20:59:53 <Luke-Jr> warren: I forget
1747 2013-06-10 20:59:56 <jgarzik> ;p
1748 2013-06-10 21:00:04 <sipa> TheUni: i doubt he feels strongly about the qt project in any way
1749 2013-06-10 21:00:08 <warren> jgarzik: fixed your bootloader issue yet? =)
1750 2013-06-10 21:00:11 <jgarzik> TheUni, saw the question, what is the issue?
1751 2013-06-10 21:00:24 <TheUni> sipa: heh, well i don't either. but i'm doing my best to keep everyone happy :)
1752 2013-06-10 21:00:37 <Luke-Jr> warren: feel free to have someone else to do it, the reason I'm waiting is because others are doing stuff to make it easier
1753 2013-06-10 21:00:38 <jgarzik> warren, heh, not with chaotic home life. Currently on MacBook Pro w/ VirtualBox for Linux dev.
1754 2013-06-10 21:00:51 * jgarzik just got Internet access at home? two hours ago
1755 2013-06-10 21:00:59 <jgarzik> after a couple weeks living in Atlanta
1756 2013-06-10 21:01:02 <sipa> TheUni: i think nobody but wumpus and Diapolo care about it (which doesn't mean that perhaps, if building becomes easier, more people start caring about it)
1757 2013-06-10 21:01:07 <TheUni> jgarzik: reason for using a single makefile was keeping autotools regression-free for qt-creator users
1758 2013-06-10 21:01:11 <warren> Luke-Jr: "doing stuff to make it easier" like?
1759 2013-06-10 21:01:20 <TheUni> jgarzik: ah, interesting. i live in atlanta :)
1760 2013-06-10 21:01:23 <jgarzik> TheUni, totally broken, screw that
1761 2013-06-10 21:01:26 <jgarzik> :)
1762 2013-06-10 21:01:37 <Luke-Jr> warren: like getting a working cross-toolchain that can actually build Mac binariesâ¦
1763 2013-06-10 21:01:49 <warren> Luke-Jr: oh, who?
1764 2013-06-10 21:01:53 maaku_ is now known as maaku
1765 2013-06-10 21:02:11 <Luke-Jr> warren: Ray Donnelly
1766 2013-06-10 21:02:36 <TheUni> jgarzik: so my question was: which is more of a concern: moving to split makefiles and having the configure at the top, or keeping qt-creator compatibility
1767 2013-06-10 21:02:40 <Luke-Jr> warren: there's a small team of people interested in this, including myself and Mike Perry (from the tor project)
1768 2013-06-10 21:02:42 <TheUni> seems you've just answered that though
1769 2013-06-10 21:03:11 Elmf has joined
1770 2013-06-10 21:03:12 <Luke-Jr> jgarzik: wumpus and Diapolo will be mad if you break Qt Creator :P
1771 2013-06-10 21:03:35 <jgarzik> I disagree that it will break Qt Creator
1772 2013-06-10 21:03:48 <sipa> TheUni: you mean breaking qt-creator, or breaking bitcoin-qt.pro
1773 2013-06-10 21:03:49 <jgarzik> if Qt Creator demands a makefile, one only, it is already broken
1774 2013-06-10 21:04:02 <TheUni> jgarzik: you understand that the qt.pro file is being removed, right?
1775 2013-06-10 21:04:06 <jgarzik> indeed -- sipa asks the relevant question :)
1776 2013-06-10 21:04:12 <jgarzik> bitcoin-qt.pro may be generated
1777 2013-06-10 21:04:29 <jgarzik> bitcoin-qt.pro.in
1778 2013-06-10 21:04:49 Prattler has quit (Quit: ZNC - http://znc.in)
1779 2013-06-10 21:04:54 tyn has quit (Ping timeout: 252 seconds)
1780 2013-06-10 21:04:58 <TheUni> jgarzik: qt-creator can use a Makefile.am, and it works quite nicely
1781 2013-06-10 21:05:12 <TheUni> i was hoping to remove the need for any duplication or manual generation
1782 2013-06-10 21:05:13 <sipa> TheUni: i saw no way to access forms/resources
1783 2013-06-10 21:05:34 <sipa> (but that was the first time i used to program at all, so that may not be very relevant)
1784 2013-06-10 21:05:58 <jgarzik> anyway, time to poof
1785 2013-06-10 21:06:01 jgarzik has quit (Quit: poof)
1786 2013-06-10 21:06:08 <TheUni> sipa: file->open works fine, and seems like a reasonable workaround to me
1787 2013-06-10 21:06:14 <sipa> ok
1788 2013-06-10 21:06:19 sanders has quit (Read error: Operation timed out)
1789 2013-06-10 21:06:23 <sipa> well, check with wumpus or Diapolo
1790 2013-06-10 21:08:26 c_k has quit (Ping timeout: 245 seconds)
1791 2013-06-10 21:08:37 Gnaf has joined
1792 2013-06-10 21:08:41 patcon has joined
1793 2013-06-10 21:08:41 caedes has quit (Remote host closed the connection)
1794 2013-06-10 21:10:03 <TheUni> ok
1795 2013-06-10 21:10:22 <TheUni> whew. you guys sure can be demotivating, you realize that? :p
1796 2013-06-10 21:12:36 Thepok has quit (Ping timeout: 245 seconds)
1797 2013-06-10 21:13:30 <sipa> TheUni: i'm doing nothing but encourage you! (at least, that's my intention...)
1798 2013-06-10 21:14:51 Thepok has joined
1799 2013-06-10 21:14:57 <TheUni> sipa: heh, that was mainly a joke. most discussion here has been pretty constructive, really
1800 2013-06-10 21:16:23 c_k has joined
1801 2013-06-10 21:17:07 kwikness has quit (Remote host closed the connection)
1802 2013-06-10 21:17:29 yubrew has quit (Remote host closed the connection)
1803 2013-06-10 21:18:19 saintcajetan has joined
1804 2013-06-10 21:18:46 OneFixt has quit ()
1805 2013-06-10 21:22:46 Keefe has joined
1806 2013-06-10 21:22:47 optimator has joined
1807 2013-06-10 21:23:01 jarpiain is now known as Guest11387
1808 2013-06-10 21:24:05 vigilyn has left ("Leaving")
1809 2013-06-10 21:25:56 peetaur2 has quit (Quit: Konversation terminated!)
1810 2013-06-10 21:26:03 sanders has joined
1811 2013-06-10 21:26:39 agnostic98 has joined
1812 2013-06-10 21:28:20 testnode9 has joined
1813 2013-06-10 21:30:29 owowo has quit (Read error: Connection reset by peer)
1814 2013-06-10 21:31:25 owowo has joined
1815 2013-06-10 21:32:09 nikolaj has joined
1816 2013-06-10 21:37:30 SchmalzTech has quit ()
1817 2013-06-10 21:37:32 root2 has quit (Quit: Leaving)
1818 2013-06-10 21:39:26 PiZZaMaN2K is now known as PiZZaMaN2K|away
1819 2013-06-10 21:40:34 bitanarchy has joined
1820 2013-06-10 21:40:57 agnostic98 has quit (Ping timeout: 255 seconds)
1821 2013-06-10 21:41:47 owowo has quit (Quit: sayonara)
1822 2013-06-10 21:42:11 agnostic98 has joined
1823 2013-06-10 21:42:58 <bitanarchy> hi, are there any projects that use https://github.com/sipa/secp256k1 ?
1824 2013-06-10 21:43:28 agnostic_ has joined
1825 2013-06-10 21:43:59 defunctzombie is now known as defunctzombie_zz
1826 2013-06-10 21:44:58 richcollins has joined
1827 2013-06-10 21:46:36 agnostic98 has quit (Ping timeout: 256 seconds)
1828 2013-06-10 21:48:14 <Ry4an> bitanarchy: sipa is here, so he probably knows :)
1829 2013-06-10 21:48:59 <bitanarchy> sipa: are there any projects that use https://github.com/sipa/secp256k1 ?
1830 2013-06-10 21:49:14 nelisky has joined
1831 2013-06-10 21:49:35 <sipa> bitanarchy: warren is doing experimental litecoin builds using it
1832 2013-06-10 21:49:54 <warren> I'd like to ship with it default, but I'd probably be shot.
1833 2013-06-10 21:50:55 <bitanarchy> sipa: i would like to have some code that can show me that this code can do the same as openssl... like turning a private key in DER format into a public key in DER format
1834 2013-06-10 21:51:06 PhantomSpark has quit (2!~kvirc@pool-71-251-16-105.nycmny.fios.verizon.net|Ping timeout: 260 seconds)
1835 2013-06-10 21:51:12 <warren> bitanarchy: I personally don't care about bitmessage, but it might be interesting to encourage them to adopt secp256k1 as an option. I think they use the same EC as bitcoin...
1836 2013-06-10 21:51:21 <sipa> bitanarchy: public keys are never encoded in DER
1837 2013-06-10 21:51:33 <nelisky> hope it's not too 'off boundaries' to ask for nmc help; in the current effort to get namecoind up to date I'm seeing a bug where a transaction with output and input sharing an address (think return change to sender) the wallet doesn't get updated with the receiving part, so address gets the balance removed correctly but doesn't get the change part added. Any pointers on where to look?
1838 2013-06-10 21:51:49 <nelisky> just to save me 30 mins of mucking aimlessly around the sources :)
1839 2013-06-10 21:51:58 <warren> nelisky: wow, people are working on rebasing namecoin?
1840 2013-06-10 21:52:14 <nelisky> warren: yes, for a while, I'm one of those people
1841 2013-06-10 21:52:15 PhantomSpark has joined
1842 2013-06-10 21:52:48 <warren> nelisky: you folks plan on making a GUI for domain registration built-in?
1843 2013-06-10 21:52:50 <bitanarchy> sipa: i am not so much into the details, but i thought that PEM is base64 of DER... isnt that the format the openssl uses to write a public key?
1844 2013-06-10 21:52:55 <nelisky> see the last few pages of https://bitcointalk.org/index.php?topic=6017.0
1845 2013-06-10 21:53:02 <sipa> bitanarchy: oh, sure
1846 2013-06-10 21:53:09 <sipa> bitanarchy: i mean, that's not used anywhere in bitcoin
1847 2013-06-10 21:53:22 <nelisky> warren: yes, already being worked on, though I don't really have a use for the UI atm
1848 2013-06-10 21:54:12 <warren> nelisky: you don't, but it'd probably get more people to actually use it
1849 2013-06-10 21:55:09 <nelisky> warren: I know, and others are dealing with that part. I need the server side to be flexible enough, and that's what I'm pushing for, but join in the fun and you'll see many others pushing for the UI
1850 2013-06-10 21:55:49 <bitanarchy> sipa: what format is used in bitcoin for private and public keys? like in the wallet file and in the blockchain.
1851 2013-06-10 21:56:12 <sipa> bitanarchy: DER for private keys
1852 2013-06-10 21:56:20 <sipa> but that's client-side only
1853 2013-06-10 21:56:43 michagogo has quit (Quit: goodnight)
1854 2013-06-10 21:56:47 <sipa> public keys are encoded as defined by the SEC standard, which also defines the curve bitcoin uses
1855 2013-06-10 21:56:53 <sipa> signatures are DER
1856 2013-06-10 21:58:59 <melvster> isnt PEM just DER with the ascii armor?
1857 2013-06-10 22:00:01 oleganza has joined
1858 2013-06-10 22:00:15 <sipa> i think so
1859 2013-06-10 22:00:46 <bitanarchy> sipa: can you tell me the difference between a serialized and non-serialized public key in your optimized code?
1860 2013-06-10 22:01:03 <sipa> a serialized public key is a series of bytes
1861 2013-06-10 22:01:12 <sipa> a public key itself is a black box
1862 2013-06-10 22:01:13 RoboTeddy has quit (Ping timeout: 240 seconds)
1863 2013-06-10 22:01:22 <sipa> just some internal representation
1864 2013-06-10 22:03:39 <sipa> bitanarchy: anyway, libsecp256k1 can import/export/convert between all serialized formats used
1865 2013-06-10 22:04:41 <bitanarchy> sipa: ok, but i see that there are two formats for private keys... serialized and non serialized, they are called privkey and seckey
1866 2013-06-10 22:04:59 <sipa> oh no, they're both serialized formats
1867 2013-06-10 22:05:21 <sipa> "non-serialized format" really makes no sense
1868 2013-06-10 22:05:35 <sipa> that's just a data type
1869 2013-06-10 22:06:05 <sipa> privkey = OpenSSL's DER encoded private key (~250 bytes)
1870 2013-06-10 22:06:19 <bitanarchy> sipa: ok, then what is this function doing: int secp256k1_ecdsa_privkey_import(unsigned char *seckey, const unsigned char *privkey, int privkeylen) ?
1871 2013-06-10 22:06:24 <phantomcircuit> sipa, one of the few things i feel could seriously be improved with a hardfork is the specification for signature encodings
1872 2013-06-10 22:06:33 <sipa> phantomcircuit: no need for a hardfork
1873 2013-06-10 22:06:35 <phantomcircuit> openssl doesn't actually implement DER it's pretty annoying
1874 2013-06-10 22:06:37 <sipa> phantomcircuit: and already done
1875 2013-06-10 22:06:48 <sipa> 0.8.2 enforces a strict DER encoding
1876 2013-06-10 22:06:51 <phantomcircuit> sipa, oh really
1877 2013-06-10 22:07:00 <phantomcircuit> sipa, for relaying or for inclusion?
1878 2013-06-10 22:07:05 <sipa> both
1879 2013-06-10 22:07:09 <sipa> but not for block validity
1880 2013-06-10 22:07:09 <phantomcircuit> hmm
1881 2013-06-10 22:07:17 <sipa> as that requires a soft fork
1882 2013-06-10 22:07:29 <phantomcircuit> when you say strict do you mean it follows the rules about smallest possible encoding?
1883 2013-06-10 22:07:34 caedes has joined
1884 2013-06-10 22:07:51 <sipa> DER allows for exactly one byte sequence only
1885 2013-06-10 22:08:04 paracyst has quit (Read error: Connection reset by peer)
1886 2013-06-10 22:08:05 <sipa> for a given data
1887 2013-06-10 22:08:27 <phantomcircuit> sipa, right
1888 2013-06-10 22:08:53 <phantomcircuit> the specific termonology for bignums is that it has to be the smallest encoding possible
1889 2013-06-10 22:08:59 <phantomcircuit> ie dont pad with null bytes
1890 2013-06-10 22:09:06 <sipa> phantomcircuit: that is incorrect
1891 2013-06-10 22:09:08 <phantomcircuit> openssl does not follow the spec there
1892 2013-06-10 22:09:21 <sipa> it does
1893 2013-06-10 22:09:42 <phantomcircuit> sipa, it encodes following those rules
1894 2013-06-10 22:09:47 <sipa> for example the byte sequence 0xFF 0xFF means -1
1895 2013-06-10 22:09:49 <phantomcircuit> but will accept null padded bignums
1896 2013-06-10 22:10:03 <sipa> so if you mean 65535, you need to use 0x00 0xFF 0xFF
1897 2013-06-10 22:10:12 <sipa> and that is what OpenSSL does, and it is correct
1898 2013-06-10 22:10:24 triciam has quit (Ping timeout: 256 seconds)
1899 2013-06-10 22:10:28 <phantomcircuit> sipa, it will also accept 0x00 0x00 0xFF 0xFF
1900 2013-06-10 22:10:30 <sipa> yes
1901 2013-06-10 22:10:35 <sipa> and that is non-DER
1902 2013-06-10 22:10:37 <phantomcircuit> which is not correct
1903 2013-06-10 22:10:41 <sipa> but 0xFF 0xFF is also non-DER
1904 2013-06-10 22:10:48 <sipa> only 0x00 0xFF 0xFF is correct
1905 2013-06-10 22:11:12 <phantomcircuit> sipa, btw im like 99% sure the DER spec actually says you have to use the shortest possible length
1906 2013-06-10 22:11:19 <sipa> it does
1907 2013-06-10 22:11:28 <sipa> but 0xFF 0xFF means something else
1908 2013-06-10 22:11:39 <sipa> and all numbers in DER are signed
1909 2013-06-10 22:11:48 <sipa> even though OpenSSL knows it expects an unsigned number
1910 2013-06-10 22:11:59 <phantomcircuit> oh
1911 2013-06-10 22:12:00 nikolaj has quit (Ping timeout: 252 seconds)
1912 2013-06-10 22:12:02 <phantomcircuit> sipa, right
1913 2013-06-10 22:12:25 <phantomcircuit> sipa, i was just saying that openssl will accept encodings which are not the minimum number of octets as being valid DER
1914 2013-06-10 22:12:30 <phantomcircuit> which is strictly speaking wrong
1915 2013-06-10 22:12:33 <sipa> phantomcircuit: you're preaching to the choir
1916 2013-06-10 22:12:41 <sipa> i've been fighting that for more than a year now
1917 2013-06-10 22:12:42 paracyst has joined
1918 2013-06-10 22:13:03 <phantomcircuit> sipa, otoh it's how sd keeps getting screwed on unconfirmed bets
1919 2013-06-10 22:13:03 <phantomcircuit> lol
1920 2013-06-10 22:13:04 patcon has quit (Remote host closed the connection)
1921 2013-06-10 22:13:16 OneFixt has joined
1922 2013-06-10 22:13:18 <sipa> the signature malleability problem is wider than that
1923 2013-06-10 22:13:19 triciam has joined
1924 2013-06-10 22:13:34 <phantomcircuit> sipa, how so?
1925 2013-06-10 22:13:52 <phantomcircuit> the signature should just be two bignums in a container
1926 2013-06-10 22:13:55 <sipa> for example if (r,s) is a valid ECDSA signature for message m and pubkey p
1927 2013-06-10 22:14:07 <sipa> then (r,-s) is also a valid ECDSA signature for message m and pubkey p
1928 2013-06-10 22:14:14 <phantomcircuit> oh right
1929 2013-06-10 22:14:24 <phantomcircuit> i noticed the negative form is now non standard
1930 2013-06-10 22:14:28 <sipa> it is not
1931 2013-06-10 22:14:48 <phantomcircuit> there's at least a message about it in the debug log
1932 2013-06-10 22:14:58 <sipa> that can't be
1933 2013-06-10 22:15:24 <sipa> ah, you mean the "non-canonical signature: negative S" ?
1934 2013-06-10 22:15:28 <phantomcircuit> yeah
1935 2013-06-10 22:15:36 ThomasV has quit (Ping timeout: 255 seconds)
1936 2013-06-10 22:15:38 <sipa> that's the 0xFF 0xFF instead of 0x00 0xFF 0xFF
1937 2013-06-10 22:15:56 <sipa> the - in (r,-s) is still modulo n
1938 2013-06-10 22:16:02 <sipa> so it's really (r,n-s)
1939 2013-06-10 22:16:02 <phantomcircuit> oh
1940 2013-06-10 22:16:04 <phantomcircuit> ok then
1941 2013-06-10 22:16:14 <sipa> you can't distinguish the two
1942 2013-06-10 22:16:22 <sipa> except one is odd and the other is even
1943 2013-06-10 22:16:26 darkee_ has joined
1944 2013-06-10 22:16:30 <sipa> so you could in theory enforce that s must be even
1945 2013-06-10 22:16:55 <sipa> https://github.com/bitcoin/bitcoin/pull/2131
1946 2013-06-10 22:17:32 defunctzombie_zz is now known as defunctzombie
1947 2013-06-10 22:17:51 <bitanarchy> sipa: how does the following format ¨openssl ec -inform PEM -in key.pem -pubout¨ compare to the format that is output by secp256k1_ecdsa_pubkey_create () ?
1948 2013-06-10 22:18:08 <phantomcircuit> sipa, right
1949 2013-06-10 22:18:24 <sipa> bitanarchy: can you paste some example output of that?
1950 2013-06-10 22:18:26 <phantomcircuit> sipa, mostly it's an issue for wallets where someone screws with their transactions and they no longer recognize them
1951 2013-06-10 22:18:48 <phantomcircuit> in general though better double spend handling in the wallet code is probably necessary anyways
1952 2013-06-10 22:18:55 <bitanarchy> sipa: let me try
1953 2013-06-10 22:18:57 <phantomcircuit> sipa, is there some other issue im not seeing
1954 2013-06-10 22:19:04 darkee has quit (Ping timeout: 240 seconds)
1955 2013-06-10 22:19:05 <bitanarchy> sipa: alex@ubuntu:~/test$ openssl ec -inform PEM -in key.pem -pubout
1956 2013-06-10 22:19:05 <bitanarchy> read EC key
1957 2013-06-10 22:19:05 <bitanarchy> writing EC key
1958 2013-06-10 22:19:05 <bitanarchy> -----BEGIN PUBLIC KEY-----
1959 2013-06-10 22:19:05 <bitanarchy> MFYwEAYHKoZIzj0CAQYFK4EEAAoDQgAEq7X4IPt4DsgaVRWYEn0qjB12vXExnYwv
1960 2013-06-10 22:19:06 <bitanarchy> hQuvjQ3iH2JBeTU7WPxt1/dvsdSQea0NdNfnXn/6JOZX0K7/IiCAmA==
1961 2013-06-10 22:19:08 <bitanarchy> -----END PUBLIC KEY-----
1962 2013-06-10 22:19:29 <phantomcircuit> bitanarchy, dont do dat
1963 2013-06-10 22:19:39 <bitanarchy> :)
1964 2013-06-10 22:20:01 <bitanarchy> phantomcircuit: sorry...
1965 2013-06-10 22:20:07 msvb-lab has quit (Quit: msvb-lab)
1966 2013-06-10 22:20:45 <sipa> bitanarchy: that's a DER encoded public key; we don't use that
1967 2013-06-10 22:21:15 <melvster> i just noticed the bitcoin uri scheme doesnt mention the address is a hash of the pub key
1968 2013-06-10 22:21:39 <gmaxwell> melvster: uh. An address is an address.
1969 2013-06-10 22:21:48 <bitanarchy> sipa: i just want to check that your code is doing what openssl is doing... so i need a way to convert the sec format into der...
1970 2013-06-10 22:21:56 <melvster> gmaxwell: it's also a hash :)
1971 2013-06-10 22:22:23 <sipa> melvster: it's the base58 encoding of some data, which is prefixed with a version byte and suffixed with a checksum
1972 2013-06-10 22:22:33 <sipa> that data can be the hash of a public key
1973 2013-06-10 22:22:54 <gmaxwell> It is none of the URI spec's darn business how you went about constructing the address. And not all addresses are constructed the same way.
1974 2013-06-10 22:23:03 <melvster> isnt it the sha256 and ripe160 of it?
1975 2013-06-10 22:23:11 <sipa> melvster: yes, but irrelevant
1976 2013-06-10 22:23:16 agnostic_ has quit (Remote host closed the connection)
1977 2013-06-10 22:23:22 <melvster> why?
1978 2013-06-10 22:23:38 <sipa> because of what gmaxwell said
1979 2013-06-10 22:23:52 RoboTeddy has joined
1980 2013-06-10 22:23:52 <sipa> an address is an address, which is something defined by the bitcoin system
1981 2013-06-10 22:23:57 <melvster> bah
1982 2013-06-10 22:24:03 malexmedia has joined
1983 2013-06-10 22:24:11 <melvster> that's wanting to have your cake and eat it :)
1984 2013-06-10 22:24:12 <sipa> and it's not necessarily the hash of a public key
1985 2013-06-10 22:24:19 <sipa> it can be a P2SH address too
1986 2013-06-10 22:24:37 <sipa> and the URI scheme shouldn't concern itself about which possible future address schemes exist
1987 2013-06-10 22:24:51 <melvster> of course it should!
1988 2013-06-10 22:24:56 <melvster> that's why we have standards
1989 2013-06-10 22:25:02 <sipa> absolutely
1990 2013-06-10 22:25:09 <sipa> but they're orthogonal
1991 2013-06-10 22:25:11 bitanarchy has quit (Remote host closed the connection)
1992 2013-06-10 22:25:11 random_cat has quit (Remote host closed the connection)
1993 2013-06-10 22:25:26 <melvster> we define things to that there can be interop
1994 2013-06-10 22:25:39 <sipa> absolutely
1995 2013-06-10 22:25:41 bitanarchy has joined
1996 2013-06-10 22:26:01 <gmaxwell> melvster: and standards for addresses are described elsewhere.
1997 2013-06-10 22:26:17 ralphtheninja has joined
1998 2013-06-10 22:26:29 <gmaxwell> Without compartmentalization everything becomes an incomprehensible mess.
1999 2013-06-10 22:26:36 random_cat has joined
2000 2013-06-10 22:26:48 agnostic98 has joined
2001 2013-06-10 22:26:49 <gmaxwell> There is no description of how an AND gate works in the URI spec, but surely it is impossible to implement it without one.
2002 2013-06-10 22:27:44 <melvster> ?
2003 2013-06-10 22:27:58 <Luke-Jr> melvster: it's quite plausable that the next version address won't be base58
2004 2013-06-10 22:28:01 jurov has quit (Ping timeout: 245 seconds)
2005 2013-06-10 22:28:23 <gmaxwell> And it's _already_ the case that some addresses are not "the hash of a (ecc) public key".
2006 2013-06-10 22:28:27 <melvster> Luke-Jr: you need to define things the way they are ... then alter the spec if it changes ... otherwise no one will know
2007 2013-06-10 22:28:59 <Luke-Jr> melvster: and addresses aren't part of the URI spec, they're more general
2008 2013-06-10 22:29:03 <gmaxwell> melvster: I look forward to your section describing Quark-gluon interaction for the URI spec.
2009 2013-06-10 22:29:11 <Luke-Jr> someone who cares about addresses, does not necessarily care about URIs
2010 2013-06-10 22:29:14 <sipa> melvster: and addresses are well-defined (BIP13 defines the extension to the address scheme for P2SH)
2011 2013-06-10 22:29:33 <Diablo-D3> http://zeptobars.ru/en/read/avalon-bitcoin-mining-unit-rig?utm_source=feedly
2012 2013-06-10 22:29:46 <sipa> perhaps informationally, the URI spec can link to BIP13
2013 2013-06-10 22:30:43 <melvster> this is a mess
2014 2013-06-10 22:30:54 <melvster> you are overloading too many things
2015 2013-06-10 22:31:04 <melvster> and hoping it will all interoperate
2016 2013-06-10 22:31:08 <sipa> the base64 specification doesn't specify you can only use it to store GIF images either, does it?
2017 2013-06-10 22:31:26 jurov has joined
2018 2013-06-10 22:31:42 <melvster> a spec is about defining something clearly and correctly
2019 2013-06-10 22:31:47 <gmaxwell> melvster: Someone who is implementing the URI spec really usually has no business or care wondering what the content of an address is.
2020 2013-06-10 22:31:49 <sipa> and it does
2021 2013-06-10 22:31:51 <gmaxwell> melvster: And the URI spec does so.
2022 2013-06-10 22:31:53 <melvster> so that other specs can work in tandem
2023 2013-06-10 22:32:20 root2 has joined
2024 2013-06-10 22:32:22 <Luke-Jr> gmaxwell's point is good too: to implement the URI spec, you *don't* need to care what an address represents
2025 2013-06-10 22:32:40 <Luke-Jr> and no implementation of it actually does, afaik
2026 2013-06-10 22:33:12 <Luke-Jr> (I'm mentally segregating Satoshi from Bitcoin-Qt here)
2027 2013-06-10 22:33:18 <melvster> gmaxwell: im unsure you have a clear understanding of relevant modularity and separation of concerns, gloun interactions bare little or no relevance, sorry if that comes across harshly, it's not intended that way :)
2028 2013-06-10 22:33:21 <gmaxwell> melvster: The URI spec does not, however, describe how the weak force acts to govern fermion interaction, or how a digital computer implements an adder.. or many many other things. Including what the internal behavior of an address is...
2029 2013-06-10 22:33:51 <melvster> gmaxwell: im sure the ietf will be very interested in your arguments
2030 2013-06-10 22:34:22 <gmaxwell> Oh, will I be seeing you in berlin?
2031 2013-06-10 22:34:35 <melvster> gmaxwell: berlin? when?
2032 2013-06-10 22:34:51 <melvster> Luke-Jr: at least you should have some informative examples on how the addresses are generated, imho
2033 2013-06-10 22:34:52 <gmaxwell> Oh, you're not coming to the next IETF meeting?
2034 2013-06-10 22:35:01 <Luke-Jr> melvster: why? that is entirely unrelated to URIs.
2035 2013-06-10 22:35:06 <melvster> gmaxwell: i try to avoid them :)
2036 2013-06-10 22:35:08 root2_ has joined
2037 2013-06-10 22:35:25 root2 is now known as Guest17394
2038 2013-06-10 22:35:25 root2_ is now known as root2
2039 2013-06-10 22:35:27 <sipa> melvster: the HTTP RFC should specify a way to generate cat videos?
2040 2013-06-10 22:35:49 <Luke-Jr> sipa: the HTML RFC probably thinks it should..
2041 2013-06-10 22:36:01 <melvster> html is a mess
2042 2013-06-10 22:36:28 <sipa> melvster: imho, the URI scheme is a way to convey a piece of black box data to a bitcoin application
2043 2013-06-10 22:36:49 <sipa> the code that handles that URI has no business in knowing what it represents, or how it is encoded
2044 2013-06-10 22:37:06 <melvster> sipa: bitcoin you are arguing that bitcoin addresses are opaque, they are not
2045 2013-06-10 22:37:16 <melvster> lol i called you bitcoin by mistake :P
2046 2013-06-10 22:37:21 <sipa> in what way are they not?
2047 2013-06-10 22:37:24 <Luke-Jr> melvster: they ARE opaque
2048 2013-06-10 22:37:51 Guest17394 has quit (Ping timeout: 252 seconds)
2049 2013-06-10 22:39:12 <sipa> well "is" or "isn't" opaque is meaningless; at some level, of course they aren't
2050 2013-06-10 22:39:18 agnostic98 has quit (Ping timeout: 256 seconds)
2051 2013-06-10 22:39:23 <melvster> sipa: in the most fundamental sense the scheme is not opaque, furthermore the checksum and encoding are well defined, as can the prefix be, and the generation from hashes of the public key is used in other places (i dont know all the ways a bitcoin address can be generated, but id like to, and I'd like to read that in the RFC)
2052 2013-06-10 22:39:36 Arbition has quit (Ping timeout: 240 seconds)
2053 2013-06-10 22:39:39 <melvster> opacity is the hardest to understand of the axioms
2054 2013-06-10 22:39:48 <sipa> the point is that anything except the final code constructing the transaction shouldn't care about what it is
2055 2013-06-10 22:40:09 <sipa> opacity is important because it allows you to separate concerns between different pieces of code
2056 2013-06-10 22:40:21 <melvster> not just code
2057 2013-06-10 22:40:23 <melvster> the whole internet
2058 2013-06-10 22:40:29 <warren> TheUni: btw, with your autoconf work, can you make it possible to static link only certain libs?
2059 2013-06-10 22:40:59 <TheUni> warren: sure, it's kinda ugly though
2060 2013-06-10 22:41:06 <TheUni> for which OS?
2061 2013-06-10 22:41:19 Maxvalor has joined
2062 2013-06-10 22:41:27 <warren> TheUni: well, I personally need it on linux, specifically for boost.
2063 2013-06-10 22:41:31 <sipa> melvster: anyway, sorry you feel different, but no application dealing with bitcoin URIs (except the final step which deals with the construction of transaction) should even know that it's base58, or has a checksum
2064 2013-06-10 22:41:36 <Luke-Jr> melvster: addresses are opaque because you *don't* know what they could be, when you implement URI support
2065 2013-06-10 22:41:37 <TheUni> static or shared boost?
2066 2013-06-10 22:41:59 <warren> TheUni: static boost.
2067 2013-06-10 22:42:33 <Luke-Jr> melvster: for example, I wrote up a new address format a few weeks ago. it's not a standard, and maybe never will be, but it also could be tomorrow, and anything assuming base58 would then be wrong
2068 2013-06-10 22:42:48 agnostic98 has joined
2069 2013-06-10 22:42:51 <Luke-Jr> at some level, that needs transitioning, but it SHOULD NOT break URI code
2070 2013-06-10 22:43:56 malexmedia has left ("PONG :lindbohm.freenode.net")
2071 2013-06-10 22:44:44 <nelisky> so, can someone give me pointers to the code path used to accept txs so I can properly debug the nmc issue with change addresses matching input addresses not being properly updated?
2072 2013-06-10 22:44:45 <warren> TheUni: two reasons: 1) https://github.com/bitcoin/bitcoin/issues/2690 bitcoin isn't compatible with fedora's boost. 2) RHEL's boost is too old.
2073 2013-06-10 22:45:52 <melvster> Luke-Jr: the whole point of an RFC is to document things in a clean way so that completely different systems can understand what they are and what they are not, as closely as possible to the truth, trying to be all things to all people does not scale
2074 2013-06-10 22:45:54 <TheUni> warren: presumably you'll be building your own static version and staging elsewhere then?
2075 2013-06-10 22:46:39 <warren> TheUni: yes
2076 2013-06-10 22:46:48 <melvster> sipa: if you think bitcoin addresses are opaque, then you think they can be any set of strings, which just is a long way from the truth
2077 2013-06-10 22:47:00 <TheUni> warren: oh, easy then. just don't build the shared versions :)
2078 2013-06-10 22:47:19 <warren> TheUni: I can't uninstall boost shared
2079 2013-06-10 22:47:27 <melvster> and who makes that decision anyway ...
2080 2013-06-10 22:47:41 <TheUni> warren: --with-boost-dir=/path/to/staging/prefix
2081 2013-06-10 22:47:56 <warren> oh. it'll static link that and dynamic everything else?
2082 2013-06-10 22:48:01 <melvster> if there's a base 58 encoding, just say so, and if that changes, change the rfc
2083 2013-06-10 22:48:03 <TheUni> er, --with-boost=, sorry
2084 2013-06-10 22:48:05 <TheUni> yes
2085 2013-06-10 22:48:12 <gmaxwell> melvster: As far as I can tell you're just trolling here.
2086 2013-06-10 22:48:12 <warren> TheUni: awesome!
2087 2013-06-10 22:48:26 <TheUni> warren: well, not exactly. it'll just link to whatever boost libs you tell it to, and in this case only static ones exist
2088 2013-06-10 22:48:28 <gmaxwell> melvster: Am I wrong?
2089 2013-06-10 22:49:08 Arbition has joined
2090 2013-06-10 22:49:20 <Luke-Jr> melvster: and URI parsers have no business knowing what addresses are inside
2091 2013-06-10 22:49:40 <Luke-Jr> nelisky: change addresses should never match "input addresses"
2092 2013-06-10 22:50:15 agnostic98 has quit (Ping timeout: 255 seconds)
2093 2013-06-10 22:50:31 milone has quit (Ping timeout: 245 seconds)
2094 2013-06-10 22:50:35 <melvster> Luke-Jr: I think the RFC would benefit from some examples on how some bitcoin addresses are made and/or used ... if they are going to be arbitrary, of course that's less work, but less informative too
2095 2013-06-10 22:50:39 oleganza has quit (Quit: oleganza)
2096 2013-06-10 22:51:05 <nelisky> Luke-Jr: maybe I'm not using the proper semantics... input1[addr1] -> output1[addr2] + output2[addr1] where output2 is not included in wallet unless I run a -rescan
2097 2013-06-10 22:51:05 <sipa> melvster: but also less limiting
2098 2013-06-10 22:51:08 <Luke-Jr> melvster: anything related to how bitcoin addresses are made/used are irrelevant to URIs
2099 2013-06-10 22:51:29 milone has joined
2100 2013-06-10 22:51:43 <sipa> nelisky: does bitcoin have the same problem?
2101 2013-06-10 22:51:43 <Luke-Jr> nelisky: addresses should never be used in more than one transaction, and inputs do not have addresses
2102 2013-06-10 22:51:48 <melvster> sipa: yes exactly there's trade offs between flexibility and scalability
2103 2013-06-10 22:51:51 <Luke-Jr> nelisky: you're doing something fundamentally wrong there
2104 2013-06-10 22:51:51 lolcookie__ has joined
2105 2013-06-10 22:52:06 Grouver has quit (Quit: I love my HydraIRC -> http://www.hydrairc.com <-)
2106 2013-06-10 22:52:24 <sipa> Luke-Jr: regardless, what he says sounds like a bug
2107 2013-06-10 22:52:29 agricocb has quit (Quit: Leaving.)
2108 2013-06-10 22:52:38 <sipa> Luke-Jr: even if it's one that only appears in a "you shouldn't do that" case
2109 2013-06-10 22:52:52 <nelisky> Luke-Jr: again, I must be explaining myself wrong. High level, I'm sending coins from one address that has a single unspent tx to two addresses, where one of them is the same address that the sending coins come from
2110 2013-06-10 22:52:57 <Luke-Jr> sipa: possibly one already fixed, considering namecoin is based on 0.3
2111 2013-06-10 22:53:13 <sipa> nelisky: yes, check whether bitcoin has the same problem; if so, file a ticket and we'll fix it
2112 2013-06-10 22:53:16 <Luke-Jr> nelisky: coins are never sent *from* addresses. they are sent from unspent coins
2113 2013-06-10 22:53:21 <nelisky> sipa: it doesn't
2114 2013-06-10 22:53:32 <Luke-Jr> nelisky: and if an address has been used before, you should never use it again
2115 2013-06-10 22:53:46 <sipa> nelisky: then the problem is in namecoin's code
2116 2013-06-10 22:53:48 <nelisky> Luke-Jr: don't nitpick me :) I know that, and why should I never use it again?
2117 2013-06-10 22:54:06 <melvster> Luke-Jr: do you really think it's feasible that the base58 encoding could change?
2118 2013-06-10 22:54:09 <nelisky> sipa: I know, I'm just asking for pointers to the code path that handles that so I can debug
2119 2013-06-10 22:54:10 <bitanarchy> sipa: thx, gotta go
2120 2013-06-10 22:54:15 bitanarchy has quit (Quit: Leaving)
2121 2013-06-10 22:54:21 <Luke-Jr> nelisky: the system is designed with the assumption that addresses are used once per transaction; reusing it harms (global) privacy, and potentially security in some possible futures
2122 2013-06-10 22:54:29 <Luke-Jr> melvster: I think it's LIKELY
2123 2013-06-10 22:54:40 <melvster> Luke-Jr: wouldnt a ton of stuff break?
2124 2013-06-10 22:54:54 <Luke-Jr> melvster: hence gradual adoption
2125 2013-06-10 22:55:01 <sipa> melvster: a lot would break yes; but there is no reason why URI handling code would or should break
2126 2013-06-10 22:55:07 <melvster> Luke-Jr: im interested in why you think it's likely, unless you feel you've had enough of this conversation for a day :)
2127 2013-06-10 22:55:10 <sipa> melvster: hence it doesn't make sense to have it depend on this
2128 2013-06-10 22:55:15 <Luke-Jr> and only things will break that need to actually work with the internals of addresses, or have the bug you want us to add to URIs
2129 2013-06-10 22:55:18 <nelisky> Luke-Jr: sure, but there's no technical reason not to do that, and it is done in a lot of instances without problems.
2130 2013-06-10 22:55:23 <Luke-Jr> (that is, assuming addresses are formatted a specific way)
2131 2013-06-10 22:55:56 <Luke-Jr> melvster: I'd suggest you read my semi-proposed change from a few weeks ago, to see the problems it addresses in base58
2132 2013-06-10 22:56:00 <nelisky> so, other than "don't do that", is there any help you can lend?
2133 2013-06-10 22:56:08 <Luke-Jr> nelisky: privacy is a technical issue
2134 2013-06-10 22:56:21 <nelisky> Luke-Jr: I'll take that as a no then
2135 2013-06-10 22:56:25 <melvster> Luke-Jr: thanks will do, do you have a pointer, or was it on the mail list or wiki?
2136 2013-06-10 22:56:30 agnostic98 has joined
2137 2013-06-10 22:56:38 <Luke-Jr> melvster: http://www.mail-archive.com/bitcoin-development@lists.sourceforge.net/msg02183.html
2138 2013-06-10 22:56:43 <melvster> thanks!
2139 2013-06-10 22:56:49 paracyst has quit ()
2140 2013-06-10 22:57:25 mrkent has joined
2141 2013-06-10 22:57:30 <sipa> nelisky: well if bitcoin doesn't have the problem, i have no idea where it is introduced
2142 2013-06-10 22:57:51 i2pRelay has quit (Quit: kytv)
2143 2013-06-10 22:57:51 <Luke-Jr> IMO, more likely it was fixed in bitcoin post-0.3
2144 2013-06-10 22:58:00 <Adrastia> The whole enforced privacy change thing pisses off more users than anything else. I don't understand the neurotically strong insistence on not making it a user-configurable option.
2145 2013-06-10 22:58:02 <sipa> ah, that bitcoin originally had it
2146 2013-06-10 22:58:17 digitalmagus has quit (Ping timeout: 256 seconds)
2147 2013-06-10 22:58:24 <Luke-Jr> Adrastia: it's fundamentally flawed on multiple levels
2148 2013-06-10 22:58:25 i2pRelay has joined
2149 2013-06-10 22:58:39 <Luke-Jr> Adrastia: for example, it reinforces the myth that the input's previous destination address is somehow a from/source address
2150 2013-06-10 22:58:51 <Luke-Jr> Adrastia: also, it becomes a security risk when quantum-safe crypto is adopted
2151 2013-06-10 22:58:59 <nelisky> sipa: well, the code on which namecoind is based is quite old, so I'm trying to understand the moving pieces so I can fix it myself. I guess I'll just have to bite the bullet and dig into the code
2152 2013-06-10 22:59:10 vigilyn has joined
2153 2013-06-10 22:59:16 <Luke-Jr> Adrastia: and it doesn't hurt just YOUR OWN privacy, it hurts EVERYONE'S privacy
2154 2013-06-10 23:00:08 <warren> nelisky: how do you fix code without looking at it?
2155 2013-06-10 23:00:31 <sipa> there is one argument right now in favor or reusing addresses that i consider valid, namely that it makes wallet backups easier
2156 2013-06-10 23:00:32 <Adrastia> I am aware. I am also not naive enough to have any expectation of privacy when I use the system as intended, frankly. The option is available in other clients, it IS going to happen. All the insistence on it does is make people less likely to use a client that actually contributes to the net and more likely to go plug into Electrum or something.
2157 2013-06-10 23:00:36 <nelisky> warren: heh, right, I was just trying to limit the scope of entry, sheesh ;)
2158 2013-06-10 23:00:48 <sipa> but that's something that can be fixed independently
2159 2013-06-10 23:01:18 <Adrastia> I spend more time explaining "Sorry, but change addresses are going to piss in your cheerios" to newbies than ANYTHING else about that client.
2160 2013-06-10 23:01:44 <Luke-Jr> Adrastia: so complain to the other clients that they shouldn't enable such bad behaviour
2161 2013-06-10 23:01:50 <sipa> Adrastia: well, i hope that the wallet implementation and the block validation/relay system can be separate at some point in the future, so you can choose to run either or both, with whatever wallet you like
2162 2013-06-10 23:02:00 <Adrastia> I do not consider it bad behavior.
2163 2013-06-10 23:02:14 <Luke-Jr> Adrastia: it is bad behaviour, whether you understand that or not
2164 2013-06-10 23:02:46 <Adrastia> It's not a question of understanding - I get what you like about it. But you can't just latch on to ONE metric and ignore all the others.
2165 2013-06-10 23:03:04 <sipa> Adrastia: it's all about abstraction really
2166 2013-06-10 23:03:25 <sipa> the way the wallet in the reference client is conceived, is that you have a balance and entry points (addresses) into the wallet
2167 2013-06-10 23:03:42 <Luke-Jr> Adrastia: so you get that by reusing addresses, someone will eventually be able to steal your coins?
2168 2013-06-10 23:03:49 <sipa> the nice thing is that this does not require much intricate understanding about how bitcoin works under the hood
2169 2013-06-10 23:04:16 paracyst has joined
2170 2013-06-10 23:04:34 <Adrastia> "Eventually" pah. The security downgrade from revealing my pubkey is not nearly a big enough hit to make up for the amount of hate the feature engenders in a regular user.
2171 2013-06-10 23:04:34 <sipa> the problem is that when people get confronted with sites like blockchain.info, they start seeing the addresses as the things that hold their coins, which is not entirely wrong, but it breaks the abstraction, and confuses many
2172 2013-06-10 23:05:03 <Luke-Jr> Adrastia: that is just one of two known security issues involved
2173 2013-06-10 23:05:12 <sipa> imho the best argument is that the client should not expose such a function (except in an 'expert mode' or so, like coin control offers) because it requires people to understand more
2174 2013-06-10 23:05:59 <sipa> that, in combination with multiwallet / deterministic wallets / whole-wallet backups
2175 2013-06-10 23:06:16 <sipa> should enable usage where you simply don't need to care about individual addresses anymore
2176 2013-06-10 23:06:18 <Luke-Jr> Adrastia: the other issue is that when quantum-safe cryptographic keys are used, signing multiple messages leaks your private key
2177 2013-06-10 23:06:25 banghouse3 has joined
2178 2013-06-10 23:06:33 <sipa> and maybe we're not there yet
2179 2013-06-10 23:06:47 <sipa> but let's not push a way of working that will complicate things later one
2180 2013-06-10 23:06:49 <sipa> *on
2181 2013-06-10 23:06:52 * sipa zZzZ
2182 2013-06-10 23:07:51 banghouse3 is now known as banghouse_
2183 2013-06-10 23:08:26 <Adrastia> Alright, it's your toy, play with it how you want. Just know that you're not doing any favors for adoption/network health.
2184 2013-06-10 23:09:20 aspect__ is now known as aspect_
2185 2013-06-10 23:09:26 toffoo has joined
2186 2013-06-10 23:09:31 <sipa> Adrastia: btw, coin control will give you that option
2187 2013-06-10 23:09:49 <sipa> which i hope we'll merge soon
2188 2013-06-10 23:10:06 <warren> Coin Control will be a huge help for dealing with p2pool dust. =(
2189 2013-06-10 23:10:42 <sipa> but i'd be sad if people start using coin control in order to keep the ability to micro-manage individual addresses (which doesn't mean they can; i mean that it shouldn't be needed)
2190 2013-06-10 23:10:57 <Luke-Jr> hmm, I got a segfault in CPartialMerkleTree::CalcHash
2191 2013-06-10 23:11:14 <sipa> because it means we're not doing a good job at offering high-level management function
2192 2013-06-10 23:11:41 melvster has quit (Ping timeout: 246 seconds)
2193 2013-06-10 23:11:49 <Luke-Jr> sipa: honestly, I use coin control for every transaction I send now, just to hand-pick dust and priorities to avoid fees and improve my UTXO footprint..
2194 2013-06-10 23:12:08 brwyatt is now known as Away!~brwyatt@brwyatt.net|brwyatt
2195 2013-06-10 23:12:09 <sipa> Luke-Jr: empty vTxid?
2196 2013-06-10 23:12:10 agnostic98 has quit (Ping timeout: 256 seconds)
2197 2013-06-10 23:12:15 <Luke-Jr> sipa: ?
2198 2013-06-10 23:12:23 <Luke-Jr> 2646 pfrom->AddInventoryKnown(inv);
2199 2013-06-10 23:12:47 <Luke-Jr> hmm
2200 2013-06-10 23:12:56 <Luke-Jr> oh, my source is wrong for this binary
2201 2013-06-10 23:12:58 <Luke-Jr> 1 sec
2202 2013-06-10 23:13:03 * sipa zZzZ
2203 2013-06-10 23:13:09 <Luke-Jr> XD
2204 2013-06-10 23:13:46 <Luke-Jr> bah, now it claims it's a comment line
2205 2013-06-10 23:14:11 <toffoo> has anyone else reported LevelDB corruption issues with bitcoin-qt 0.8.2 Mac version ?
2206 2013-06-10 23:14:22 <Luke-Jr> sipa: you're right, vTxid.size()==0
2207 2013-06-10 23:14:44 one_zero has joined
2208 2013-06-10 23:14:45 <dugo> that's using the same address twice, good grief!
2209 2013-06-10 23:14:55 <dugo> wrong chan
2210 2013-06-10 23:15:04 agnostic98 has joined
2211 2013-06-10 23:15:26 <Luke-Jr> dugo: eh, is this conversation in another chan too? XD
2212 2013-06-10 23:15:42 chorao has joined
2213 2013-06-10 23:15:42 chorao has quit (Changing host)
2214 2013-06-10 23:15:42 chorao has joined
2215 2013-06-10 23:16:11 <dugo> Luke-Jr: please stay out of it ;)
2216 2013-06-10 23:16:27 <Luke-Jr> â¦
2217 2013-06-10 23:16:42 fourlights_ has joined
2218 2013-06-10 23:17:27 fourlights has quit (Ping timeout: 252 seconds)
2219 2013-06-10 23:19:48 rdymac has quit (Quit: This computer has gone to sleep)
2220 2013-06-10 23:19:52 digitalmagus has joined
2221 2013-06-10 23:20:28 enikanorov_ has joined
2222 2013-06-10 23:21:01 roconnor has quit (Remote host closed the connection)
2223 2013-06-10 23:22:14 patcon has joined
2224 2013-06-10 23:22:36 akrmn has quit (Quit: Leaving.)
2225 2013-06-10 23:22:39 enikanorov has quit (Ping timeout: 256 seconds)
2226 2013-06-10 23:23:23 akrmn has joined
2227 2013-06-10 23:25:48 BlackPrapor has quit (Ping timeout: 255 seconds)
2228 2013-06-10 23:26:00 defunctzombie is now known as defunctzombie_zz
2229 2013-06-10 23:27:22 robocoin has quit (Ping timeout: 245 seconds)
2230 2013-06-10 23:28:10 agricocb has joined
2231 2013-06-10 23:29:09 asuk has joined
2232 2013-06-10 23:31:12 RoboTeddy has quit (Ping timeout: 248 seconds)
2233 2013-06-10 23:31:13 agnostic98 has quit (Ping timeout: 252 seconds)
2234 2013-06-10 23:31:18 agnostic_ has joined
2235 2013-06-10 23:31:36 agnostic_ has quit (Read error: Connection reset by peer)
2236 2013-06-10 23:31:40 lolcookie__ has quit (Read error: Connection reset by peer)
2237 2013-06-10 23:31:57 asuk has quit (Client Quit)
2238 2013-06-10 23:32:00 lolcookie__ has joined
2239 2013-06-10 23:32:05 agnostic98 has joined
2240 2013-06-10 23:40:16 Thepok has quit (Ping timeout: 248 seconds)
2241 2013-06-10 23:41:31 malexmedia has joined
2242 2013-06-10 23:42:04 <malexmedia> hey all; any ideas why i might not be seeing any connections when connecting through Tor?
2243 2013-06-10 23:43:22 nelisky has quit (Quit: nelisky)
2244 2013-06-10 23:46:04 richcollins has quit (Quit: richcollins)
2245 2013-06-10 23:46:20 ne0futur is now known as n30futur
2246 2013-06-10 23:46:53 n30futur is now known as n3ofutur
2247 2013-06-10 23:47:01 <Luke-Jr> malexmedia: #bitcoin for user chat
2248 2013-06-10 23:47:07 <Luke-Jr> malexmedia: also, read the docs for tor usage first
2249 2013-06-10 23:47:17 <malexmedia> sorry, headed over there
2250 2013-06-10 23:48:55 RoboTeddy has joined
2251 2013-06-10 23:48:59 n3ofutur is now known as ne0futur
2252 2013-06-10 23:50:32 richcollins has joined
2253 2013-06-10 23:52:11 RoboTeddy has quit (Read error: Connection reset by peer)
2254 2013-06-10 23:52:19 digitalmagus has quit (Ping timeout: 256 seconds)
2255 2013-06-10 23:52:26 agnostic98 has quit (Remote host closed the connection)
2256 2013-06-10 23:52:42 RoboTeddy has joined
2257 2013-06-10 23:53:20 KevinT has joined
2258 2013-06-10 23:55:24 agnostic98 has joined
2259 2013-06-10 23:55:24 agnostic98 has quit (Remote host closed the connection)
2260 2013-06-10 23:55:58 agnostic98 has joined
2261 2013-06-10 23:56:06 rdponticelli has quit (Read error: Connection reset by peer)
2262 2013-06-10 23:56:27 AndChat64721 has joined
2263 2013-06-10 23:56:27 RoboTeddy has quit (Read error: Connection reset by peer)
2264 2013-06-10 23:56:43 RoboTeddy has joined
2265 2013-06-10 23:59:38 RoboTeddy has quit (Read error: Connection reset by peer)
2266 2013-06-10 23:59:43 RoboTedd_ has joined