1 2012-06-18 00:00:03 <sipa> yes
2 2012-06-18 00:01:27 <genjix> dk5: line 622 http://gitorious.org/libbitcoin/libbitcoin/blobs/master/src/script.cpp
3 2012-06-18 00:01:47 <dk5> thanks :)
4 2012-06-18 00:02:20 <genjix> the parser is line 981
5 2012-06-18 00:02:35 egecko has joined
6 2012-06-18 00:02:44 TD has quit (Quit: TD)
7 2012-06-18 00:06:09 sirk390 has quit (Read error: Connection reset by peer)
8 2012-06-18 00:06:18 <dk5> sipa: you asked a question on here a long time ago and i was wondering if you ever got a response. in ScanHash_CryptoPP() (main.cpp), the line "if (((unsigned short*)phash)[14] == 0)" checks to see if the hash of the current block being mined has "some" leading zeros. why isn't it 15 instead of 14?
9 2012-06-18 00:06:55 <sipa> dk5: i assume an endianness issue
10 2012-06-18 00:07:22 <dk5> genjix: thanks. looking at it now :)
11 2012-06-18 00:07:57 <sipa> dk5: https://github.com/sipa/bitcoin/blob/ultraprune/src/script.cpp#L1758
12 2012-06-18 00:08:27 <sipa> there are some very instanciated tests for some txout scripts
13 2012-06-18 00:08:50 <gmaxwell> sipa: can I get a quick sanity glance on https://github.com/bitcoin/bitcoin/pull/1477 ? (I'm about to recommend people use testnet, it's better if it works. ;) )
14 2012-06-18 00:09:04 welterde has quit (Ping timeout: 264 seconds)
15 2012-06-18 00:09:30 <dk5> sipa: yeah, that's the way i was going about it. it just seems complicated and i wanted to make sure there wasn't a more general way
16 2012-06-18 00:10:48 <sipa> gmaxwell: merged
17 2012-06-18 00:12:39 <luke-jr> gmaxwell: should I pullreq the block preview relaying?
18 2012-06-18 00:15:02 <gmaxwell> luke-jr: get some more opinions on the new message type? I'm ambivalent about that. Otherwise I think its an okay idea. Actually, where is your latest commit, I'll write a message to the list about it.
19 2012-06-18 00:15:38 <luke-jr> https://github.com/luke-jr/bitcoin/commit/0ce6f590dc2b9cbb46ceecd7320220f55d814bca
20 2012-06-18 00:18:40 Diablo-D3 has joined
21 2012-06-18 00:19:34 <Diablo-D3> MysteryBanshee: hey
22 2012-06-18 00:19:52 <Diablo-D3> MysteryBanshee: to immediately buy it, you need an order for whatever the lowest ask is
23 2012-06-18 00:20:35 copumpkin has joined
24 2012-06-18 00:21:44 welterde has joined
25 2012-06-18 00:22:02 random_cat__ has quit (Ping timeout: 276 seconds)
26 2012-06-18 00:25:32 random_cat__ has joined
27 2012-06-18 00:25:35 maaku has quit (Quit: maaku)
28 2012-06-18 00:31:30 t7 has quit (Remote host closed the connection)
29 2012-06-18 00:38:32 <gmaxwell> luke-jr: Okay, I sent a post.. if you want to add some debugging messages and send me a patch we could spin up some nodes running it and try it out
30 2012-06-18 00:41:06 tcatm_ is now known as tcatm
31 2012-06-18 00:49:05 AntKinGTube has quit (Quit: Leaving)
32 2012-06-18 00:57:59 genjix has quit (Ping timeout: 272 seconds)
33 2012-06-18 00:58:35 <luke-jr> gmaxwell: awesome reply on using master in production btw
34 2012-06-18 00:59:09 <luke-jr> gmaxwell: unfortunately, I'm in the middle of some CGMiner updates, so testing the other stuff may need to wait :/
35 2012-06-18 00:59:50 maaku has joined
36 2012-06-18 00:59:55 <gmaxwell> luke-jr: thats fine.
37 2012-06-18 01:00:40 Joric_ has joined
38 2012-06-18 01:00:40 Joric_ has quit (Changing host)
39 2012-06-18 01:00:40 Joric_ has joined
40 2012-06-18 01:00:51 wizkid057 has quit (Read error: Operation timed out)
41 2012-06-18 01:01:59 wizkid057 has joined
42 2012-06-18 01:02:37 Joric has quit (Ping timeout: 260 seconds)
43 2012-06-18 01:02:50 Joric_ is now known as Joric
44 2012-06-18 01:06:22 rdponticelli_ has joined
45 2012-06-18 01:07:13 rdponticelli has quit (Ping timeout: 246 seconds)
46 2012-06-18 01:07:20 rdponticelli_ is now known as rdponticelli
47 2012-06-18 01:10:42 Joric has quit ()
48 2012-06-18 01:11:27 JZavala has joined
49 2012-06-18 01:14:10 Ummon has joined
50 2012-06-18 01:14:50 Lexa has joined
51 2012-06-18 01:24:13 JZavala has quit (Ping timeout: 244 seconds)
52 2012-06-18 01:30:01 bonks has quit (Excess Flood)
53 2012-06-18 01:31:50 maaku has quit (Quit: maaku)
54 2012-06-18 01:33:40 minimoose has joined
55 2012-06-18 01:39:05 bonks has joined
56 2012-06-18 01:40:42 maaku has joined
57 2012-06-18 02:06:34 wizkid057 has quit (Read error: Operation timed out)
58 2012-06-18 02:08:10 wizkid057 has joined
59 2012-06-18 02:11:58 [7] has quit (Disconnected by services)
60 2012-06-18 02:12:04 TheSeven has joined
61 2012-06-18 02:14:34 winterblack has quit (Remote host closed the connection)
62 2012-06-18 02:18:24 sanchaz has quit (Ping timeout: 276 seconds)
63 2012-06-18 02:24:31 eoss has quit (Remote host closed the connection)
64 2012-06-18 02:29:18 Xunie has joined
65 2012-06-18 02:30:09 minimoose_ has joined
66 2012-06-18 02:30:18 minimoose_ has quit (Client Quit)
67 2012-06-18 02:34:27 minimoose has quit (Ping timeout: 265 seconds)
68 2012-06-18 02:34:50 abracadab is now known as abracadabra
69 2012-06-18 02:55:21 <jgarzik> come on, block. mint!
70 2012-06-18 02:57:20 <luke-jr> that's what she said
71 2012-06-18 02:58:54 Cinder_ has joined
72 2012-06-18 03:00:45 <Cinder_> Looking for a way to filter the download/save file API by data type or downloading server ? Any ideas?
73 2012-06-18 03:02:00 Cinder_ has quit (Client Quit)
74 2012-06-18 03:02:04 <luke-jr> â¦
75 2012-06-18 03:12:18 <Karmaon> luke-jr: why are your periods so small?
76 2012-06-18 03:12:21 <Karmaon> ...
77 2012-06-18 03:12:24 <Karmaon> thats what she said!
78 2012-06-18 03:13:37 egecko has quit (Quit: ~ Trillian Astra - www.trillian.im ~)
79 2012-06-18 03:15:56 osmosis has quit (Quit: Leaving)
80 2012-06-18 03:26:36 <devrandom> /peek sipa
81 2012-06-18 03:29:07 Zarutian has quit (Quit: Zarutian)
82 2012-06-18 03:36:48 Karmaon has quit (Remote host closed the connection)
83 2012-06-18 03:36:49 <jgarzik> huh. looks like somebody is doing sendmany, to send to SD: http://blockexplorer.com/block/000000000000029e73317727abd0519c62753a95adfd18583edb1d63ed58d441
84 2012-06-18 03:37:06 Karmaon has joined
85 2012-06-18 03:37:16 <jgarzik> maybe somebody changed their bot to be more network-friendly
86 2012-06-18 03:37:42 darkee has quit (Ping timeout: 276 seconds)
87 2012-06-18 03:38:29 eoss has joined
88 2012-06-18 03:41:21 <gmaxwell> or they're trying to break it again.
89 2012-06-18 03:45:31 Cory has quit (Read error: No route to host)
90 2012-06-18 03:45:55 <jgarzik> sendmany for SD reduces total number of transactions, which is IMO good. And I -think- a larger TX w/ fee will not penalize non-SD TX's as adversely as sending out a larger number of singleton TX's
91 2012-06-18 03:46:38 <jgarzik> i.e. I think it will not waste space in the free area
92 2012-06-18 03:46:56 <jgarzik> (correct me if I'm wrong...)
93 2012-06-18 03:48:13 eoss has quit (Ping timeout: 246 seconds)
94 2012-06-18 03:57:02 SomeoneWeird has quit (Ping timeout: 260 seconds)
95 2012-06-18 03:57:09 <dk5> is the 20-byte hash in the BIP16 type script different from the 20-byte hash in "OP_DUP OP_HASH160 <pubKeyHash> OP_EQUALVERIFY OP_CHECKSIG"?
96 2012-06-18 03:57:54 <gmaxwell> dk5: it is the hash of a script, not a hash of some public key.
97 2012-06-18 03:58:12 Wack0 has quit (Ping timeout: 260 seconds)
98 2012-06-18 03:58:47 Clipse has quit (Ping timeout: 260 seconds)
99 2012-06-18 03:59:32 <dk5> for <pubKeyHash>, it was the RIPEMD160(SHA256(pubKey)). What is the 20-byte hash in BIP16? RIPEMD160(pubKey+OP_CHECKSIG)?
100 2012-06-18 03:59:58 Wack0 has joined
101 2012-06-18 04:00:12 <gmaxwell> dk5: a complete serialized script.
102 2012-06-18 04:00:21 <dk5> oh
103 2012-06-18 04:00:56 Guest43036 has joined
104 2012-06-18 04:01:59 <dk5> gmaxwell: so I'm misreading "scriptSig: [signature] {[pubkey] OP_CHECKSIG}" and "scriptPubKey: OP_HASH160 [20-byte-hash of {[pubkey] OP_CHECKSIG} ] OP_EQUAL" from the wiki then?
105 2012-06-18 04:03:59 <dk5> gmaxwell: are there any examples of this already in the block chain?
106 2012-06-18 04:04:37 <luke-jr> dk5: in practice, that script is never used in P2SH as implemented today
107 2012-06-18 04:04:45 <luke-jr> dk5: bitcoind only supports multisigs
108 2012-06-18 04:04:57 <gmaxwell> Sure, but I don't recall off the top of my head and I need to run. Luke can probably point you to some.
109 2012-06-18 04:05:15 <dk5> gmaxwell: k, thanks :)
110 2012-06-18 04:05:48 <dk5> luke-jr: but it will be used eventually, no?
111 2012-06-18 04:11:25 <luke-jr> dk5: that's BIP18 or 19
112 2012-06-18 04:11:38 <luke-jr> https://en.bitcoin.it/wiki/BIP_0018
113 2012-06-18 04:16:36 <dk5> luke-jr: the stuff I pasted was on the P2SH wiki. i guess there is overlap. is there an example somewhere for P2SH scripts?
114 2012-06-18 04:17:03 <luke-jr> I forget, sorry.
115 2012-06-18 04:18:22 Detritus has quit (Quit: Konversation terminated!)
116 2012-06-18 04:18:52 Detritus has joined
117 2012-06-18 04:23:00 kakobrekla has quit (Ping timeout: 245 seconds)
118 2012-06-18 04:23:32 ThomasV has joined
119 2012-06-18 04:28:36 brwyatt is now known as brwyatt|Away
120 2012-06-18 04:32:37 ThomasV has quit (Ping timeout: 260 seconds)
121 2012-06-18 04:35:56 osmosis has joined
122 2012-06-18 04:38:07 rdponticelli has quit (Read error: Connection reset by peer)
123 2012-06-18 04:39:20 wizkid057 has quit (Read error: Operation timed out)
124 2012-06-18 04:40:23 wizkid057 has joined
125 2012-06-18 04:54:40 Guest43036 has quit (Changing host)
126 2012-06-18 04:54:40 Guest43036 has joined
127 2012-06-18 04:56:17 Guest43036 is now known as SomeoneWeird
128 2012-06-18 04:58:10 copumpkin has quit (Ping timeout: 240 seconds)
129 2012-06-18 05:00:28 copumpkin has joined
130 2012-06-18 05:02:48 one_zero has joined
131 2012-06-18 05:11:33 sytse has quit (Read error: Operation timed out)
132 2012-06-18 05:13:29 sytse has joined
133 2012-06-18 05:32:45 D34TH has quit (Ping timeout: 276 seconds)
134 2012-06-18 05:38:13 Cory has joined
135 2012-06-18 05:42:09 ThomasV has joined
136 2012-06-18 05:42:40 Aetitiae_ has quit (Ping timeout: 246 seconds)
137 2012-06-18 05:46:28 Cory has quit (Read error: Connection reset by peer)
138 2012-06-18 05:49:23 Cory has joined
139 2012-06-18 05:50:44 onefourone has joined
140 2012-06-18 05:51:06 mmoya has joined
141 2012-06-18 05:51:14 RazielZ has joined
142 2012-06-18 05:52:33 Backburn has joined
143 2012-06-18 05:54:15 ThomasV has quit (Ping timeout: 245 seconds)
144 2012-06-18 05:54:36 wizkid057 has quit (Read error: Operation timed out)
145 2012-06-18 05:55:31 wizkid057 has joined
146 2012-06-18 05:57:35 imsaguy has quit (Ping timeout: 245 seconds)
147 2012-06-18 06:15:31 ovidiusoft has joined
148 2012-06-18 06:23:19 maaku has quit (Quit: maaku)
149 2012-06-18 06:27:50 osmosis has quit (Quit: Leaving)
150 2012-06-18 06:29:54 darkee has joined
151 2012-06-18 06:31:00 <gribble> New news from bitcoinrss: Diapolo opened pull request 1478 on bitcoin/bitcoin <https://github.com/bitcoin/bitcoin/pull/1478>
152 2012-06-18 06:35:58 scottj has joined
153 2012-06-18 06:41:12 <gribble> New news from bitcoinrss: Diapolo opened pull request 1479 on bitcoin/bitcoin <https://github.com/bitcoin/bitcoin/pull/1479>
154 2012-06-18 06:44:55 ThomasV has joined
155 2012-06-18 06:55:05 yellowhat has quit (Remote host closed the connection)
156 2012-06-18 06:58:50 graingert has joined
157 2012-06-18 07:03:24 imsaguy has joined
158 2012-06-18 07:06:44 ThomasV has quit (Quit: Quitte)
159 2012-06-18 07:06:45 <gribble> New news from bitcoinrss: fanquake opened pull request 1480 on bitcoin/bitcoin <https://github.com/bitcoin/bitcoin/pull/1480>
160 2012-06-18 07:08:05 graingert has quit (Ping timeout: 246 seconds)
161 2012-06-18 07:11:09 graingert has joined
162 2012-06-18 07:11:57 Ahimoth has quit (Ping timeout: 244 seconds)
163 2012-06-18 07:12:06 wizkid057 has quit (Read error: Operation timed out)
164 2012-06-18 07:13:07 wizkid057 has joined
165 2012-06-18 07:16:17 graingert_ has joined
166 2012-06-18 07:17:11 graingert__ has joined
167 2012-06-18 07:19:47 graingert_ has quit (Read error: Connection reset by peer)
168 2012-06-18 07:19:58 graingert_ has joined
169 2012-06-18 07:20:04 graingert has quit (Ping timeout: 255 seconds)
170 2012-06-18 07:21:17 graingert has joined
171 2012-06-18 07:21:30 graingert__ has quit (Ping timeout: 256 seconds)
172 2012-06-18 07:22:06 graingert_ has quit (Read error: Connection reset by peer)
173 2012-06-18 07:22:19 graingert_ has joined
174 2012-06-18 07:24:48 graingert__ has joined
175 2012-06-18 07:25:51 graingert has quit (Ping timeout: 264 seconds)
176 2012-06-18 07:26:30 graingert_ has quit (Ping timeout: 245 seconds)
177 2012-06-18 07:26:54 graingert has joined
178 2012-06-18 07:27:10 mcorlett has quit (Changing host)
179 2012-06-18 07:27:10 mcorlett has joined
180 2012-06-18 07:27:44 dvide has joined
181 2012-06-18 07:28:32 Ahimoth has joined
182 2012-06-18 07:29:13 graingert_ has joined
183 2012-06-18 07:29:18 graingert__ has quit (Ping timeout: 265 seconds)
184 2012-06-18 07:31:41 scottj has left ()
185 2012-06-18 07:31:42 graingert has quit (Ping timeout: 276 seconds)
186 2012-06-18 07:43:10 graingert_ has quit (Read error: Connection reset by peer)
187 2012-06-18 07:43:37 graingert has joined
188 2012-06-18 07:44:35 Prattler has joined
189 2012-06-18 07:44:37 <Eliel> any ideas what's creating this spike in this filtered graph now? http://blockchain.info/charts/n-transactions-excluding-popular
190 2012-06-18 07:44:42 graingert_ has joined
191 2012-06-18 07:45:53 <MysteryBanshee> probably me sending so many transactions :P
192 2012-06-18 07:45:54 <MysteryBanshee> j/k
193 2012-06-18 07:47:59 graingert has quit (Ping timeout: 255 seconds)
194 2012-06-18 07:50:27 justmoon has joined
195 2012-06-18 07:50:27 justmoon has quit (Changing host)
196 2012-06-18 07:50:27 justmoon has joined
197 2012-06-18 07:52:49 devrandom has quit (Remote host closed the connection)
198 2012-06-18 07:56:08 random_cat__ has quit (Remote host closed the connection)
199 2012-06-18 07:56:39 <Eliel> MysteryBanshee: you've sent enough to make a spike of 5000?
200 2012-06-18 07:56:56 <Eliel> well, ok, 4000
201 2012-06-18 07:57:10 <MysteryBanshee> lol
202 2012-06-18 07:57:18 <MysteryBanshee> nah I think ive only sent a few hundred
203 2012-06-18 07:57:26 random_cat__ has joined
204 2012-06-18 07:57:39 <MysteryBanshee> probably satoshi dice
205 2012-06-18 07:58:20 devrandom has joined
206 2012-06-18 07:59:32 <Eliel> that graph filters out satoshidice since those belong to the 100 most common addresses
207 2012-06-18 07:59:57 <Eliel> the graph that doesn't filter them out looks like this: http://blockchain.info/charts/n-transactions-per-block
208 2012-06-18 08:17:05 t7 has joined
209 2012-06-18 08:17:14 molecular has quit (Ping timeout: 248 seconds)
210 2012-06-18 08:17:58 molecular has joined
211 2012-06-18 08:18:36 upb has quit (Ping timeout: 244 seconds)
212 2012-06-18 08:20:38 graingert_ has quit (Read error: Connection reset by peer)
213 2012-06-18 08:22:58 sirk390 has joined
214 2012-06-18 08:40:09 wereHamster has quit (Remote host closed the connection)
215 2012-06-18 08:44:15 Nachtwind is now known as n8{work}
216 2012-06-18 08:51:06 wereHamster has joined
217 2012-06-18 08:52:02 coderrr has quit (Quit: Terminated with extreme prejudice - dircproxy 1.0.5)
218 2012-06-18 08:53:15 Ferroh has joined
219 2012-06-18 08:56:30 yellowhat has joined
220 2012-06-18 09:00:57 xxxxxxx has joined
221 2012-06-18 09:09:58 Apexseals has quit (Ping timeout: 250 seconds)
222 2012-06-18 09:18:08 n8{work} has quit (Quit: ZNC - http://znc.in)
223 2012-06-18 09:18:55 Turingi has joined
224 2012-06-18 09:19:00 dlb76 has quit (Ping timeout: 245 seconds)
225 2012-06-18 09:20:24 graingert has joined
226 2012-06-18 09:33:47 Turingi has quit (Read error: Connection reset by peer)
227 2012-06-18 09:34:48 Apexseals has joined
228 2012-06-18 09:36:25 Nachtwind has joined
229 2012-06-18 09:37:14 Zarutian has joined
230 2012-06-18 09:38:24 gribble has quit (Remote host closed the connection)
231 2012-06-18 09:38:51 Joric has joined
232 2012-06-18 09:39:51 danbri has joined
233 2012-06-18 09:41:07 gribble has joined
234 2012-06-18 09:43:02 graingert has quit (Read error: Connection reset by peer)
235 2012-06-18 09:43:19 graingert has joined
236 2012-06-18 09:45:08 wizkid057 has quit (Read error: Operation timed out)
237 2012-06-18 09:46:08 wizkid057 has joined
238 2012-06-18 09:50:21 graingert has quit (Read error: Connection reset by peer)
239 2012-06-18 09:50:40 graingert has joined
240 2012-06-18 09:52:59 graingert_ has joined
241 2012-06-18 09:54:17 graingert__ has joined
242 2012-06-18 09:56:01 graingert__ has quit (Read error: Connection reset by peer)
243 2012-06-18 09:56:44 graingert has quit (Ping timeout: 246 seconds)
244 2012-06-18 09:57:20 graingert_ has quit (Ping timeout: 256 seconds)
245 2012-06-18 10:08:35 graingert_ecs has joined
246 2012-06-18 10:18:59 sanchaz has joined
247 2012-06-18 10:19:40 noagendamarket has quit (Quit: Leaving)
248 2012-06-18 10:23:18 Ummon has quit (Remote host closed the connection)
249 2012-06-18 10:27:42 agricocb has quit (Remote host closed the connection)
250 2012-06-18 10:32:18 graingert_ecs has quit (Remote host closed the connection)
251 2012-06-18 10:43:31 one_zero has quit ()
252 2012-06-18 10:48:55 genjix has joined
253 2012-06-18 10:49:05 leotreasure has joined
254 2012-06-18 10:49:21 <leotreasure> hello
255 2012-06-18 10:50:22 <leotreasure> i'm trying to build the bitcoin daemon on os x lion and get an error: ld: symbol(s) not found for architecture x86_64
256 2012-06-18 10:50:35 <sipa> which symbols?
257 2012-06-18 10:51:04 upb has joined
258 2012-06-18 10:51:04 upb has quit (Changing host)
259 2012-06-18 10:51:04 upb has joined
260 2012-06-18 10:51:43 <leotreasure> one sec
261 2012-06-18 10:52:18 <leotreasure> http://pastebin.com/yydXhT0g
262 2012-06-18 10:52:27 <leotreasure> looks like some boost ones
263 2012-06-18 10:53:05 <sipa> and openssl
264 2012-06-18 10:53:15 <sipa> and bdb
265 2012-06-18 10:54:44 <leotreasure> ok, do you have any idea how can i get these symbols to work?
266 2012-06-18 10:57:05 <sipa> by linking against the appropriate libraries
267 2012-06-18 10:57:18 <sipa> unfortunately i have absolutely no idea how to do that on osx
268 2012-06-18 10:57:55 <leotreasure> ok thanks for your help - will look into it
269 2012-06-18 10:58:17 <sipa> did you follow doc/build-osx.txt ?
270 2012-06-18 10:58:21 <leotreasure> yes
271 2012-06-18 10:58:38 <sipa> ok
272 2012-06-18 11:02:41 wizkid057 has quit (Read error: Operation timed out)
273 2012-06-18 11:03:42 wizkid057 has joined
274 2012-06-18 11:17:07 dlb76 has joined
275 2012-06-18 11:18:02 agricocb has joined
276 2012-06-18 11:28:08 erle- has joined
277 2012-06-18 11:41:55 egecko has joined
278 2012-06-18 11:46:24 drizztbsd has joined
279 2012-06-18 11:50:11 Aetitiae has joined
280 2012-06-18 11:53:35 graingert_ has joined
281 2012-06-18 11:55:17 Backburn has quit (Remote host closed the connection)
282 2012-06-18 11:55:24 Clipse has joined
283 2012-06-18 11:55:38 Backburn has joined
284 2012-06-18 12:02:05 cdecker has joined
285 2012-06-18 12:03:11 <MysteryBanshee> hey guys
286 2012-06-18 12:03:19 <MysteryBanshee> how long does it take to brute force a bitcoin private key if you only have a third?
287 2012-06-18 12:03:44 <sipa> 1/3 ?
288 2012-06-18 12:03:47 <Diablo-D3> a third what?
289 2012-06-18 12:04:06 Ummon has joined
290 2012-06-18 12:04:39 <sipa> Diablo-D3: do you know how fast oclvanitygen is compared to bitcoin mining on the same hardware?
291 2012-06-18 12:04:40 <MagicalTux> a third of the key, I guess
292 2012-06-18 12:04:47 <MagicalTux> it still take a very long time
293 2012-06-18 12:05:07 <Diablo-D3> sipa: no, but I do know its a lot faster than cpu vanity genning ;)
294 2012-06-18 12:05:23 <sipa> so 85 out of 256 bits of the secret key are known?
295 2012-06-18 12:06:09 <sipa> oclvanitygen does like 20 Mkey/s on recent GPU cards, it seems
296 2012-06-18 12:07:20 <Diablo-D3> sipa: that seems low
297 2012-06-18 12:07:28 <sipa> so let's make a convervative guess that hardware doing X MH/s can do X/10 Mkey/s (it's probably less)
298 2012-06-18 12:07:45 <sipa> ;;bc,nethash
299 2012-06-18 12:07:45 <gribble> 12321.945276568917
300 2012-06-18 12:08:15 <sipa> so if the entire network switched to trying to brute force your key, it'd do 1 Tkey/s, let's say
301 2012-06-18 12:08:15 <Diablo-D3> I wonder how its hashing those
302 2012-06-18 12:08:25 <Diablo-D3> because theres probably a shitload of improvement to do
303 2012-06-18 12:08:40 <sipa> i doubt that; the bulk of the computation is EC additions
304 2012-06-18 12:09:16 <Diablo-D3> yeah, but how much of it is being done in parallel
305 2012-06-18 12:10:46 <sipa> anyway, to brute force 256-85=171 bits at 1Tkey/s, it'd take some 94847369674933745800730934462790 years
306 2012-06-18 12:11:38 <Diablo-D3> I should look at oclvanitygen anyhow
307 2012-06-18 12:11:45 <Diablo-D3> I wanna change my keys over to vanity keys
308 2012-06-18 12:11:52 <Diablo-D3> s/keys/addresses/
309 2012-06-18 12:16:53 rdponticelli has joined
310 2012-06-18 12:17:04 minimoose has joined
311 2012-06-18 12:19:46 wizkid057 has quit (Read error: Operation timed out)
312 2012-06-18 12:20:30 wizkid057 has joined
313 2012-06-18 12:24:12 datagutt has joined
314 2012-06-18 12:32:19 <xorgate> oclvanitygen does about 30mkeys/s
315 2012-06-18 12:32:24 <xorgate> on my 5870
316 2012-06-18 12:35:07 <sipa> and how many hashes does it do?
317 2012-06-18 12:35:34 <xorgate> cgminer does ~355mhashes/s
318 2012-06-18 12:35:49 <Diablo-D3> so sipa was right
319 2012-06-18 12:35:55 <Diablo-D3> 10 to 1
320 2012-06-18 12:36:04 <drizztbsd> oclvanitygen works better on nvidia
321 2012-06-18 12:36:34 comboy_ has quit (Ping timeout: 256 seconds)
322 2012-06-18 12:38:27 comboy has joined
323 2012-06-18 12:39:13 BurtyBB is now known as BurtyB
324 2012-06-18 12:39:46 <Diablo-D3> drizztbsd: find that hard to believe
325 2012-06-18 12:39:57 <Diablo-D3> the kernel in oclvanitygen is probably shit and needs optimization badly
326 2012-06-18 12:40:08 <drizztbsd> yes, so it works better on nvidia atm
327 2012-06-18 12:40:09 <drizztbsd> :P
328 2012-06-18 12:41:47 <sipa> it relies on fast multiplication
329 2012-06-18 12:42:03 <sipa> hashing only does additions and xors and shift/rotate
330 2012-06-18 12:42:15 <sipa> well, and other bitwise operations
331 2012-06-18 12:42:32 <Diablo-D3> does it use floats?
332 2012-06-18 12:42:35 <sipa> no
333 2012-06-18 12:44:44 <Diablo-D3> then its slower.
334 2012-06-18 12:46:08 <sipa> feel free to optimize it
335 2012-06-18 12:46:30 <sipa> i already found it quite impressive how fast they got this on gpu's
336 2012-06-18 12:47:11 <drizztbsd> I use it on gtx 560 ti :P
337 2012-06-18 12:48:36 Joric has quit ()
338 2012-06-18 12:52:40 <Diablo-D3> sipa: meh
339 2012-06-18 12:52:49 <Diablo-D3> its just more work
340 2012-06-18 12:52:58 <Diablo-D3> and probably whoever owns the project wont accept the patch anyhow
341 2012-06-18 12:54:13 Ummon has quit (Remote host closed the connection)
342 2012-06-18 13:02:29 copumpkin has quit (Quit: Computer has gone to sleep.)
343 2012-06-18 13:02:58 p2k has joined
344 2012-06-18 13:03:47 <Graet> p2k, is author of ecoinpool. we are looking to try optimise pool, if any devs around to answe a few questions. :)
345 2012-06-18 13:03:57 <p2k> heh, hi everyone
346 2012-06-18 13:04:07 slush has joined
347 2012-06-18 13:06:02 <p2k> my question is: if a new block arrives, does bitcoind remove the validated transactions from its transaction pool before forwarding the block to peers?
348 2012-06-18 13:06:37 <sipa> p2k: from the memory pool? almost simtaneously i'd say
349 2012-06-18 13:07:34 <p2k> sipa: "almost" but not immediately? i.e. it's done in another thread?
350 2012-06-18 13:08:18 <sipa> yes, the send is done from the network thread
351 2012-06-18 13:08:59 <sipa> actually, i'll need to look it up
352 2012-06-18 13:11:15 <p2k> the thing is, if it is done in a thread and the block is forwarded while the txn purge process is still running, a pool software might take advantage of that
353 2012-06-18 13:12:28 mrvision has joined
354 2012-06-18 13:12:35 <mrvision> hello
355 2012-06-18 13:13:02 <mrvision> I have a problem. When i try to create a php that connects to the bitcoind server it returns 403 error
356 2012-06-18 13:13:31 <mrvision> does anybody know the solution?
357 2012-06-18 13:13:34 <p2k> sipa: my current idea is: right after a block change, we could for once (and for speed up) assume that all transactions get purged and construct a new block entirely within the pool software
358 2012-06-18 13:15:20 <p2k> sipa: then send this block out via longpolling to the clients. on any subsequent request, we go through the regular process of asking bitcoind for pooled transactions (which may take some time, if bitcoind is still sorting out things)
359 2012-06-18 13:16:20 <mrvision> does anybody know how to connect with the json library to the bitcoind server?
360 2012-06-18 13:18:25 MC1984 has quit (Quit: Leaving)
361 2012-06-18 13:19:37 <xorgate> mrvision there's libs for that
362 2012-06-18 13:20:04 <mrvision> it must be something about the config of my server
363 2012-06-18 13:20:06 <p2k> sipa: worst case is, that a block is solved with only the coinbase transaction where otherwise valid and still pooled transactions could have been integrated. on the other hand, we could prevent stales and thus enable more computing power to be used for securing the network rather than being completely wasted.
364 2012-06-18 13:20:09 <mrvision> it rejects the connection
365 2012-06-18 13:20:26 <mrvision> that's why im looking for someone who has already done it
366 2012-06-18 13:20:48 <mrvision> i'm following a tutorial step by step and it throws me an extrange error
367 2012-06-18 13:20:50 <mrvision> :S
368 2012-06-18 13:24:37 minimoose_ has joined
369 2012-06-18 13:25:44 Turingi has joined
370 2012-06-18 13:25:51 TD has joined
371 2012-06-18 13:27:21 minimoose has quit (Ping timeout: 250 seconds)
372 2012-06-18 13:27:23 minimoose_ is now known as minimoose
373 2012-06-18 13:29:21 copumpkin has joined
374 2012-06-18 13:32:55 <sipa> p2k: i see
375 2012-06-18 13:33:42 rdponticelli_ has joined
376 2012-06-18 13:33:53 <sipa> p2k: but both pruning the mempool and sending out the inv are not expensive
377 2012-06-18 13:34:08 rdponticelli has quit (Ping timeout: 246 seconds)
378 2012-06-18 13:34:16 <sipa> quite sure mempool pruning is done first, by the way
379 2012-06-18 13:35:53 <p2k> sipa: hmm, then there might be some other problem, why we get delays on blockchanges...
380 2012-06-18 13:38:25 james_ is now known as topace
381 2012-06-18 13:38:30 topace has quit (Changing host)
382 2012-06-18 13:38:30 topace has joined
383 2012-06-18 13:40:54 mndrix has joined
384 2012-06-18 13:40:54 mndrix has quit (Changing host)
385 2012-06-18 13:40:54 mndrix has joined
386 2012-06-18 13:40:54 mndrix has quit (Changing host)
387 2012-06-18 13:40:54 mndrix has joined
388 2012-06-18 13:40:57 cdecker has quit (Quit: Leaving.)
389 2012-06-18 13:48:47 gavinandresen has joined
390 2012-06-18 13:50:03 Diapolo has joined
391 2012-06-18 13:50:54 <Graet> https://bitcointalk.org/index.php?topic=88302.0
392 2012-06-18 13:51:17 <Graet> High orphan rate and long confirmation time discussion
393 2012-06-18 13:53:58 rdponticelli_ is now known as rdponticelli
394 2012-06-18 13:57:52 <TD> morning gavin
395 2012-06-18 13:58:03 <gavinandresen> good morning
396 2012-06-18 14:01:13 Zarutian has quit (Quit: Zarutian)
397 2012-06-18 14:01:17 <gavinandresen> sipa here?
398 2012-06-18 14:01:34 <sipa> yes
399 2012-06-18 14:02:02 <gavinandresen> I followed the instructions in Tor.txt to run bitcoin as a hidden service, but I'm not getting connected-- getting nothing but Connection refused messages.
400 2012-06-18 14:02:09 <gavinandresen> (testing pull 1174)
401 2012-06-18 14:02:56 <sipa> where is it trying to connect to?
402 2012-06-18 14:03:04 <sipa> normal ipv4 destinations, or onion addresses?
403 2012-06-18 14:03:07 <gavinandresen> trying connection 69.165.195.172:8333 lastseen=146.9hrs
404 2012-06-18 14:03:15 <gavinandresen> (for example)
405 2012-06-18 14:03:29 <sipa> maybe your tor node hasn't been up long enough to establish connections
406 2012-06-18 14:03:47 <gavinandresen> it was running all weekend
407 2012-06-18 14:03:51 <sipa> ok
408 2012-06-18 14:04:21 <sipa> it may take longer to find a good destination node, as there's no timeout implemented in the proxy code
409 2012-06-18 14:04:33 Ummon_ has joined
410 2012-06-18 14:04:59 <gavinandresen> so I should just let it run for a while?
411 2012-06-18 14:05:06 tower has quit (Ping timeout: 245 seconds)
412 2012-06-18 14:05:43 <sipa> not more than a few minutes
413 2012-06-18 14:05:46 <gavinandresen> hmm, I stopped Tor and I get the same Connection refused messages
414 2012-06-18 14:05:47 <mrvision> does anybody know how to fix this? http://www.bitcoinerr.com/abascal/prueba.php
415 2012-06-18 14:06:00 <gavinandresen> maybe Tor isn't listening on the port I think it is listening on...
416 2012-06-18 14:06:05 <mrvision> i'm trying to create a json object to conect to the bitcoind daemon
417 2012-06-18 14:06:34 <gribble> New news from bitcoinrss: fanquake opened pull request 1481 on bitcoin/bitcoin <https://github.com/bitcoin/bitcoin/pull/1481>
418 2012-06-18 14:06:35 <sipa> mrvision: do you send http user/pass auth?
419 2012-06-18 14:06:38 <gavinandresen> (is there any way for bitcoind to detect that the proxy isn't working and give a better error message?)
420 2012-06-18 14:06:44 <mrvision> sipa yes
421 2012-06-18 14:07:18 <mrvision> sipa i connect putting my ip and the bitcoind port in chrome
422 2012-06-18 14:07:20 egecko has quit (Quit: ~ Trillian Astra - www.trillian.im ~)
423 2012-06-18 14:07:28 <mrvision> and then i put the user and password
424 2012-06-18 14:07:30 <mrvision> it returns this
425 2012-06-18 14:07:31 <mrvision> {"result":null,"error":{"code":-32700,"message":"Parse error"},"id":null}
426 2012-06-18 14:07:45 <sipa> mrvision: yes, you need to send an JSON object in
427 2012-06-18 14:07:47 <Diapolo> We need a proxy-not-working detection, as an enabled proxy also crashes Qt, when proxy is not usable / reachable.
428 2012-06-18 14:07:59 <sipa> gavinandresen: hmm, actually, that wouldn't be hard
429 2012-06-18 14:08:14 <sipa> the ConnectDirectly() call to the proxy shouldn't fail
430 2012-06-18 14:08:43 <sipa> ConnectSocketDirectly()
431 2012-06-18 14:09:59 <sipa> mrvision: no idea then
432 2012-06-18 14:10:19 <mrvision> that's why if you try to cnnect to that ip
433 2012-06-18 14:10:22 <gmaxwell> mrvision: remove characters from your password that it might have a hard time with the escaping.
434 2012-06-18 14:10:31 <gmaxwell> E.g. <>&;
435 2012-06-18 14:10:37 tower has joined
436 2012-06-18 14:10:47 <mrvision> password is something easy for testing like: password
437 2012-06-18 14:10:51 <mrvision> which is not :)
438 2012-06-18 14:10:55 <gmaxwell> No idea then.
439 2012-06-18 14:11:15 <mrvision> ips allowed are: *.*.*.*
440 2012-06-18 14:11:47 <mrvision> if you try to connect here: http://184.154.36.81:8771
441 2012-06-18 14:11:54 <mrvision> you will be prompted for user and pass
442 2012-06-18 14:12:05 <mrvision> so the port is open
443 2012-06-18 14:12:14 <mrvision> and listening
444 2012-06-18 14:13:54 <gavinandresen> sipa: I'm stuck, can't ACK 1174 because I can't get it to connect. Running latest Tor browser bundle on my mac
445 2012-06-18 14:14:53 jurov is now known as away!aktooj@84.245.71.31|jurov
446 2012-06-18 14:15:22 <sipa> gavinandresen: do the connections fail immediately, or does it take a while?
447 2012-06-18 14:15:27 <sipa> (a while = seconds)
448 2012-06-18 14:15:30 <gavinandresen> immediately
449 2012-06-18 14:15:42 <gavinandresen> same behavior if Tor is running or not
450 2012-06-18 14:15:47 <sipa> which command line?
451 2012-06-18 14:15:54 <sipa> or bitcoin.conf options?
452 2012-06-18 14:16:25 <gavinandresen> sipa: https://gist.github.com/2948581
453 2012-06-18 14:18:45 <sipa> i'll try myself in a minute; maybe things broke after the network option changes
454 2012-06-18 14:22:17 Ummon_ has quit (Read error: Connection reset by peer)
455 2012-06-18 14:22:31 Ummon has joined
456 2012-06-18 14:23:25 jurov is now known as jurov|away
457 2012-06-18 14:23:55 Ummon has quit (Read error: Connection reset by peer)
458 2012-06-18 14:24:19 Ummon_ has joined
459 2012-06-18 14:24:35 moartr4dez has quit (Remote host closed the connection)
460 2012-06-18 14:24:35 paraipan has quit (Remote host closed the connection)
461 2012-06-18 14:25:08 moartr4dez has joined
462 2012-06-18 14:25:20 paraipan has joined
463 2012-06-18 14:26:13 leotreasure has quit (Quit: leotreasure)
464 2012-06-18 14:28:59 mrvision has quit (Ping timeout: 245 seconds)
465 2012-06-18 14:29:28 <TheSeven> hm, I thought "move" doesn't allow to push an account negative?
466 2012-06-18 14:29:41 Internet13 has quit (Ping timeout: 244 seconds)
467 2012-06-18 14:29:46 <sipa> it does allow that since 0.3.22 or so
468 2012-06-18 14:29:59 <sipa> since there are other ways to get it negative anyway
469 2012-06-18 14:30:06 <TheSeven> someone should update the docs then...
470 2012-06-18 14:30:32 <sipa> where?
471 2012-06-18 14:30:49 <TheSeven> "Moves from the default account to any other account always succeed; moves from any other account will fail if the account has insufficient funds."
472 2012-06-18 14:30:52 <TheSeven> https://en.bitcoin.it/wiki/Accounts_explained
473 2012-06-18 14:30:58 <sipa> gavinandresen: works perfectly here
474 2012-06-18 14:31:22 <gavinandresen> how should I debug? How do I tell if Tor is listening on 9050 ?
475 2012-06-18 14:31:30 <sipa> netstat -ltnp ?
476 2012-06-18 14:31:39 <sipa> (or osx equivalent)
477 2012-06-18 14:32:09 <sipa> maybe that tor bundle doesn't install a socks proxy by default?
478 2012-06-18 14:32:16 <gavinandresen> Maybe not
479 2012-06-18 14:33:20 <gavinandresen> sigh. I have no idea what netstat -ltnp is on OSX (I keep telling people I don't know nuthin about networking....)
480 2012-06-18 14:33:29 Internet13 has joined
481 2012-06-18 14:34:09 <sipa> netcat 127.0.0.1 9050
482 2012-06-18 14:34:14 <guruvan> probably should install tor and polipo as startup daemons & not use the bundle (also netstat -an | grep "9050")
483 2012-06-18 14:34:14 <sipa> maybe?
484 2012-06-18 14:34:38 Ummon_ has quit (Read error: Connection reset by peer)
485 2012-06-18 14:34:50 <gavinandresen> netstat -an | grep 9050 gets nothing
486 2012-06-18 14:34:50 Ummon_ has joined
487 2012-06-18 14:35:03 <sipa> $ netstat -an | fgrep 9050
488 2012-06-18 14:35:03 <sipa> tcp 0 0 127.0.0.1:9050 0.0.0.0:* LISTEN
489 2012-06-18 14:35:06 <sipa> tcp 0 0 127.0.0.1:9050 127.0.0.1:35409 TIME_WAIT
490 2012-06-18 14:35:42 <gavinandresen> re: install tor and polipo as startup bundles: ummm, if I can't figure it out from the doc/Tor.txt file then very, very few people will be able to figure it out
491 2012-06-18 14:36:08 <sipa> I welcome improvements with OS-specific help :)
492 2012-06-18 14:36:22 <sipa> but for me it's apt-get install tor, and follow the config file
493 2012-06-18 14:36:28 <sipa> *doc
494 2012-06-18 14:37:53 <gavinandresen> "it works on linux" isn't good enough to accept a pull request, in my humble opinon.
495 2012-06-18 14:38:57 <sipa> well, that file contains enough information assuming your tor proxy works
496 2012-06-18 14:39:41 <sipa> it can't reasonably replicate the documentation on how to get tor running in the first place (though some pointers or hints are good)
497 2012-06-18 14:39:53 <sipa> https://www.torproject.org/docs/faq.html.en#TBBSocksPort
498 2012-06-18 14:40:03 <sipa> seems that some setups use a random proxy port
499 2012-06-18 14:40:29 <gavinandresen> ah, that'd do it
500 2012-06-18 14:41:14 <gavinandresen> I'll give that a try when I have a minute, I moved on to testing the signbugs pull request
501 2012-06-18 14:41:39 <sipa> great
502 2012-06-18 14:42:30 graingert has joined
503 2012-06-18 14:42:59 <p2k> gavinandresen: "netstat -ltnp" is "netstat -lnp tcp" on os x and it won't show processes like in the gnu version
504 2012-06-18 14:43:13 <gavinandresen> p2k: thanks
505 2012-06-18 14:45:23 wasabi1 has quit (Read error: Connection reset by peer)
506 2012-06-18 14:46:12 <p2k> and if you really need to know the process, try "lsof -i :1234"
507 2012-06-18 14:47:34 Zarutian has joined
508 2012-06-18 14:51:43 Ummon_ has quit (Remote host closed the connection)
509 2012-06-18 14:53:43 Ummon has joined
510 2012-06-18 14:54:33 Aetitiae has quit (Quit: Leaving)
511 2012-06-18 14:54:37 minimoose_ has joined
512 2012-06-18 14:55:10 graingert_ has quit (Read error: Connection reset by peer)
513 2012-06-18 14:55:34 graingert_ has joined
514 2012-06-18 14:56:46 minimoose has quit (Ping timeout: 245 seconds)
515 2012-06-18 14:56:46 minimoose_ is now known as minimoose
516 2012-06-18 14:58:35 graingert_ has quit (Read error: Connection reset by peer)
517 2012-06-18 14:58:44 graingert_ has joined
518 2012-06-18 14:59:45 graingert_ has quit (Read error: Connection reset by peer)
519 2012-06-18 14:59:48 graingert__ has joined
520 2012-06-18 15:00:36 <MysteryBanshee> sipa: how long does it take to brute for 8 characters of a bitcoin private key at 1 Tkey/s ?
521 2012-06-18 15:00:46 <MysteryBanshee> for=force
522 2012-06-18 15:00:50 <sipa> what is 8 characters? 8 bytes?
523 2012-06-18 15:01:00 Hasbro has quit (Ping timeout: 252 seconds)
524 2012-06-18 15:01:02 <sipa> or 8 base58 encoded characters?
525 2012-06-18 15:01:19 <MysteryBanshee> yeh
526 2012-06-18 15:01:24 <sipa> which one?
527 2012-06-18 15:01:24 <drizztbsd> yes what =.=
528 2012-06-18 15:01:27 <MysteryBanshee> 8 base58 encoded characters
529 2012-06-18 15:01:47 <sipa> MysteryBanshee: 2 minutes
530 2012-06-18 15:01:58 Hasbro has joined
531 2012-06-18 15:02:14 <MysteryBanshee> sipa: seriously?
532 2012-06-18 15:02:20 <upb> 58^8 / (10^12)
533 2012-06-18 15:02:24 <upb> seconds
534 2012-06-18 15:02:32 <drizztbsd> ;;bc,calc 58^8 / (10^12)
535 2012-06-18 15:02:33 <gribble> Error: Something in there wasn't a valid number.
536 2012-06-18 15:02:34 <sipa> ;;calc 58^8 / (10^12)
537 2012-06-18 15:02:35 <gribble> Error: Something in there wasn't a valid number.
538 2012-06-18 15:02:35 <MysteryBanshee> sipa: how many base58 characters do I need for it to take at least a week?
539 2012-06-18 15:02:39 <drizztbsd> fuck gribble
540 2012-06-18 15:03:00 <drizztbsd> MysteryBanshee: lol?
541 2012-06-18 15:03:03 <sipa> MysteryBanshee: 10
542 2012-06-18 15:03:21 <sipa> may i ask what for?
543 2012-06-18 15:03:57 <sipa> please don't go below 128 bits of secret if you're trying to come up with a fancy shorter encoding
544 2012-06-18 15:04:08 D34TH has joined
545 2012-06-18 15:04:09 D34TH has quit (Changing host)
546 2012-06-18 15:04:09 D34TH has joined
547 2012-06-18 15:04:20 <sipa> which is 21 base58 characters
548 2012-06-18 15:04:24 <drizztbsd> 128 over 256?
549 2012-06-18 15:05:06 <sipa> ECDSA/secp256k1 only has 128 bits of security anyway
550 2012-06-18 15:05:38 minimoose_ has joined
551 2012-06-18 15:07:04 Insti has quit (Read error: Connection timed out)
552 2012-06-18 15:07:05 erle- has quit (Quit: erle-)
553 2012-06-18 15:07:11 <MysteryBanshee> sipa: its for a game that I am running
554 2012-06-18 15:07:18 <MysteryBanshee> I dont want people to be able to brute for parts of a private key
555 2012-06-18 15:07:25 Insti has joined
556 2012-06-18 15:07:26 <MysteryBanshee> 10 - thx sipa
557 2012-06-18 15:08:01 <MysteryBanshee> for=force lol
558 2012-06-18 15:08:04 wasabi has joined
559 2012-06-18 15:08:47 <Diapolo> luke-jr: online?
560 2012-06-18 15:08:51 minimoose has quit (Ping timeout: 245 seconds)
561 2012-06-18 15:08:51 minimoose_ is now known as minimoose
562 2012-06-18 15:09:07 <luke-jr> Diapolo: ?
563 2012-06-18 15:11:20 <Diapolo> luke-jr: hi there ... I was working on the GUI part of coin control you said I need to "git push" like I normally do ... this would be a git push origin coincontrol ... I'm not sure.
564 2012-06-18 15:12:11 gavinandresen has quit (Quit: gavinandresen)
565 2012-06-18 15:12:19 darkee has quit (Remote host closed the connection)
566 2012-06-18 15:13:07 darkee has joined
567 2012-06-18 15:13:17 RainbowDashh has joined
568 2012-06-18 15:14:39 p2k has left ()
569 2012-06-18 15:32:43 graingert_ has joined
570 2012-06-18 15:32:44 graingert_ has quit (Read error: Connection reset by peer)
571 2012-06-18 15:35:23 RainbowDashh has quit (Quit: Computer has gone to sleep. DERPY HOOVES SLAMMING THE LAPTOP LID MISTAKE?)
572 2012-06-18 15:37:15 graingert__ has quit (Ping timeout: 276 seconds)
573 2012-06-18 15:42:33 setkeh has quit (Quit: Time For a World Without Govorment Internet Interfearence)
574 2012-06-18 15:42:45 Eliel has quit (Ping timeout: 245 seconds)
575 2012-06-18 15:43:20 Bwild has joined
576 2012-06-18 15:43:23 <Diapolo> luke-jr: I sent a pull-request to luke-jr/coincontrol, take a look, I'll be online later again.
577 2012-06-18 15:43:25 Diapolo has left ()
578 2012-06-18 15:44:23 Eliel has joined
579 2012-06-18 15:48:21 toffoo has joined
580 2012-06-18 15:54:26 wasabi1 has joined
581 2012-06-18 15:54:41 wasabi has quit (Ping timeout: 240 seconds)
582 2012-06-18 15:55:02 ovidiusoft has quit (Quit: leaving)
583 2012-06-18 15:55:53 darkskiez has quit (Remote host closed the connection)
584 2012-06-18 15:58:17 freewil has quit (Quit: Leaving)
585 2012-06-18 15:58:51 <MysteryBanshee> Im having problems with blockchain.info (again)
586 2012-06-18 15:58:52 <MysteryBanshee> anyone else ?
587 2012-06-18 15:59:00 <MysteryBanshee> Everytime I try to send money it crashes my browser
588 2012-06-18 15:59:25 <TD> that sounds like a problem with your browser rather than blockchain.info
589 2012-06-18 15:59:33 <MysteryBanshee> with both firefox and ie?
590 2012-06-18 16:00:03 <sipa> define "crashes my browser" ?
591 2012-06-18 16:00:11 <MysteryBanshee> well causes some kind of error message
592 2012-06-18 16:00:15 <MysteryBanshee> relating to some javascript
593 2012-06-18 16:00:16 <justmoon> sending money means it'll generate a transaction - that's the heftiest job it does, it has to generate a signature
594 2012-06-18 16:00:19 <justmoon> was it working before?
595 2012-06-18 16:00:23 <MysteryBanshee> yes
596 2012-06-18 16:00:36 <justmoon> then I assume ben messed with the javascript
597 2012-06-18 16:00:48 <sipa> that's hardly a crash, and it does sound like a problem with blockchain.info indeed
598 2012-06-18 16:00:48 <justmoon> possibly try and close other programs
599 2012-06-18 16:00:56 <justmoon> maybe it's just timing out
600 2012-06-18 16:00:57 <yellowhat> i'd try ctrl-f5 and try again :)
601 2012-06-18 16:01:34 <MysteryBanshee> tried yello
602 2012-06-18 16:01:42 <MysteryBanshee> blockchain.info is just plain refusing to work and I dont know why
603 2012-06-18 16:02:03 <sipa> queue complaints about centralized services in 1... 2... 3...
604 2012-06-18 16:02:22 <MysteryBanshee> wait, lemme reboot
605 2012-06-18 16:02:27 slush has quit (Read error: Operation timed out)
606 2012-06-18 16:02:30 MysteryBanshee has quit ()
607 2012-06-18 16:03:36 freewil has joined
608 2012-06-18 16:05:57 t7 has quit (Remote host closed the connection)
609 2012-06-18 16:06:33 MysteryBanshee has joined
610 2012-06-18 16:07:25 <MysteryBanshee> its still not working
611 2012-06-18 16:08:24 <MysteryBanshee> im seeing this error in the firefox thingy
612 2012-06-18 16:08:25 <MysteryBanshee> [17:01:54.992] (new ReferenceError("ws is not defined", "https://blockchain.info/Resources/wallet/wallet.min.js", 1)) @ https://blockchain.info/Resources/wallet/wallet.min.js:1
613 2012-06-18 16:08:45 <justmoon> ws is probably WebSocket
614 2012-06-18 16:08:51 <justmoon> wonder if they have a server issue
615 2012-06-18 16:09:12 <justmoon> I'd just file a support request with them
616 2012-06-18 16:13:48 <MysteryBanshee> sent!
617 2012-06-18 16:13:53 <MysteryBanshee> my second support request
618 2012-06-18 16:14:00 <MysteryBanshee> i dont understand why these people take so long to respond
619 2012-06-18 16:14:47 <sipa> anyone an idea how many files windows-based filesystems can handle in a directory, without causing performance issues?
620 2012-06-18 16:15:17 <luke-jr> Diablo-D3: could you try to follow git-standard multiline commit messages in the future please? âº
621 2012-06-18 16:15:40 <jgarzik> sipa: plenty, if it's NTFS (most are). Not so much, if FAT32
622 2012-06-18 16:15:45 <luke-jr> Diablo-D3: that is, "Short summary of change" <blank line> "As many lines of details as you need"
623 2012-06-18 16:16:05 <jgarzik> sipa: but in general... if you are asking that question, you should do something other than store thousands or millions of files :)
624 2012-06-18 16:16:16 rdponticelli has quit (Ping timeout: 276 seconds)
625 2012-06-18 16:16:26 <sipa> jgarzik: i plan to store each block in a separate file, basically
626 2012-06-18 16:16:36 <luke-jr> ^ ++
627 2012-06-18 16:16:58 <sipa> maybe lump them together in files of 10-100, but for now, 1 per file is easier
628 2012-06-18 16:17:14 <luke-jr> sipa: could always use directories like git does too
629 2012-06-18 16:17:16 <jgarzik> sipa: ugh, why?
630 2012-06-18 16:17:30 <sipa> jgarzik: to support pruning by simply deleting the file :)
631 2012-06-18 16:17:50 <jgarzik> sipa: disk space to store blocks is irrelevant
632 2012-06-18 16:18:03 <sipa> ?
633 2012-06-18 16:19:50 jav__ has joined
634 2012-06-18 16:20:54 <sipa> jgarzik: not sure what you're talking about
635 2012-06-18 16:22:18 <jgarzik> sipa: even at SD levels, even at every-block-is-1MB levels, the disk space needed to store the block chain is so cheap that thinking about pruning the raw block chain is pointless
636 2012-06-18 16:23:15 <jgarzik> sipa: nobody stores things in individual files anymore... c.f. git pack files or bdb w/ deletion
637 2012-06-18 16:24:09 <sipa> i'm not going to store entire blocks in bdb
638 2012-06-18 16:24:28 <Prattler> jgarzik, it's not about price, it's about availability and newbie friendliness.
639 2012-06-18 16:25:15 gribble has quit (Remote host closed the connection)
640 2012-06-18 16:26:20 <jgarzik> interesting
641 2012-06-18 16:26:28 <jgarzik> just received two blocks almost simultaneously
642 2012-06-18 16:26:40 <jgarzik> an orphan, and then a valid block that un-orphaned the first
643 2012-06-18 16:26:44 <TheSeven> hm, testnet faucet is acting weird
644 2012-06-18 16:26:45 <luke-jr> jgarzik: I thought git pack files existed for compression more than avoiding individual files?
645 2012-06-18 16:26:50 <jgarzik> 06/18/12 16:21:55 SetBestChain: new best=000000000000089965e3 height=185154 work=358855919114642062569
646 2012-06-18 16:26:51 <jgarzik> 06/18/12 16:21:55 SetBestChain: new best=000000000000019c4a2e height=185155 work=358862718915477312876
647 2012-06-18 16:27:09 <luke-jr> jgarzik: that's⦠interesting
648 2012-06-18 16:27:13 <sipa> is there any consensus about the name for a 210000-block era?
649 2012-06-18 16:27:24 <genjix> the pleistocene
650 2012-06-18 16:27:29 <sipa> genjix: thanks!
651 2012-06-18 16:27:32 <luke-jr> lol
652 2012-06-18 16:27:33 <jgarzik> luke-jr: opening and scanning all those inodes was a huge FS metadata hit
653 2012-06-18 16:27:38 <jgarzik> luke-jr: in the early git versions
654 2012-06-18 16:27:43 <jgarzik> pre-pack era
655 2012-06-18 16:27:49 <luke-jr> jgarzik: ah, didn't know that
656 2012-06-18 16:28:16 <sipa> jgarzik: i might add that these block files will not be frequently accessed (only for block getdata's, rescans and reorgs)
657 2012-06-18 16:28:21 <jgarzik> your FS metadata + your data is likely to be fragmented quite a bit, with a lot of files
658 2012-06-18 16:28:27 <luke-jr> jgarzik: is the reason because filesystem requires userspace-to-kernel calls instead of being entirely in-process, or something else?
659 2012-06-18 16:28:44 <luke-jr> our data is already hugely fragmented :P
660 2012-06-18 16:28:56 <jgarzik> busy web servers, Usenet news servers that stored one-file-per-article and other historical use cases have all proved the one-file-per-datum structure to be ultimately inefficient
661 2012-06-18 16:29:27 <genjix> yeah there's overhead for opening/closing files
662 2012-06-18 16:29:39 <jgarzik> luke-jr: in-process versus in-kernel has nothing to do with data layout on hard disk
663 2012-06-18 16:29:52 <jgarzik> luke-jr: that's process context switching
664 2012-06-18 16:29:53 <genjix> so then you use a single pagefile, and you separate it into different block sizes and make a btree indexer
665 2012-06-18 16:30:04 <genjix> that's bdb
666 2012-06-18 16:30:17 <luke-jr> jgarzik: yes, but I was pondering if there's a reason filesystem use has to suck much more
667 2012-06-18 16:30:34 <jgarzik> but ultimately, the *nix filesystem is simply a database. so, you are overloading the *nix filesystem to behave as a database, yet without all the useful indexing that comes along with it.
668 2012-06-18 16:30:41 <jgarzik> *nix filesystem is not a database
669 2012-06-18 16:30:56 <jgarzik> (even if btrfs or zfs sure does look like one internally)
670 2012-06-18 16:31:01 <sipa> it's an hierarchical database
671 2012-06-18 16:31:14 <jgarzik> theory != practice
672 2012-06-18 16:31:14 * luke-jr personally wishes his filesystem was git <.<
673 2012-06-18 16:32:08 <sipa> jgarzik: ok, so assume i really want to be able to delete blocks (not transactions, i'll always delete a bunch of blocks in one go), what do you suggest?
674 2012-06-18 16:32:25 <genjix> sipa: i suggest a readonly db
675 2012-06-18 16:32:34 <sipa> genjix: delete != readonly
676 2012-06-18 16:32:35 <genjix> beyond a certain point, blocks migrate into RO space
677 2012-06-18 16:32:50 <genjix> yeah but reorgs you can assume never happen 1k blocks back
678 2012-06-18 16:33:05 gribble has joined
679 2012-06-18 16:33:08 <sipa> genjix: i know how i'll deal with reorgs
680 2012-06-18 16:33:17 <jgarzik> sipa: standard solution is either (a) readonly + deletion notations, periodic rewrite or (b) allocate space in pages (1024-byte aligned blocks), and reuse "deleted" space
681 2012-06-18 16:33:22 <luke-jr> genjix: he's not talking about reorgs
682 2012-06-18 16:33:29 <genjix> luke-jr: neither am i
683 2012-06-18 16:33:45 <jgarzik> sipa: bdb, postgresql, mysql trend towards (b)
684 2012-06-18 16:33:52 <jgarzik> sipa: as do Unix filesystems
685 2012-06-18 16:33:53 <genjix> i'm saying you can make a much faster database if it is append only
686 2012-06-18 16:34:00 <gmaxwell> jgarzik: these are archival files that don't need indexing by much of anything. They won't be frequently accessed at all in his scheme.
687 2012-06-18 16:34:03 <genjix> and have a fragmented 'active' index
688 2012-06-18 16:34:16 Ahimoth has quit (Ping timeout: 264 seconds)
689 2012-06-18 16:34:20 <luke-jr> genjix: so what we do right now
690 2012-06-18 16:34:40 <sipa> can someone suggest a nice name for 210000-block periods, and 2016-block periods?
691 2012-06-18 16:34:44 <gmaxwell> jgarzik: Sipa's validation will be satisfied by the txout tree. The blocks are only needed for updating peers and for part of the reorg process.
692 2012-06-18 16:34:46 <genjix> an epoch
693 2012-06-18 16:34:47 <TheSeven> hm, why on earth would the testnet faucet send me 0.11 BTC?
694 2012-06-18 16:34:49 <jgarzik> gmaxwell: yeah, um, that sounds a lot like what we do now. blk????.dat is appendonly, with an actively accessed index
695 2012-06-18 16:34:56 <genjix> 210k blocks = epoch
696 2012-06-18 16:35:00 <gmaxwell> jgarzik: but where old data can't be removed.
697 2012-06-18 16:35:04 <luke-jr> genjix: too ambiguous
698 2012-06-18 16:35:10 <gmaxwell> sipa: I call 2016 blocks 'cycles' or 'difficulty cycles'.
699 2012-06-18 16:35:22 <sipa> era / cycle ?
700 2012-06-18 16:35:25 <jgarzik> gmaxwell: again... why is removal even remotely considered a problem to be given priority?
701 2012-06-18 16:35:45 <genjix> sipa: epoch is a geological term for long time periods.
702 2012-06-18 16:35:52 <genjix> era is colloquial
703 2012-06-18 16:35:56 <genjix> so take your pick
704 2012-06-18 16:35:58 <gmaxwell> jgarzik: because without it we don't get any space savings from txout indexes.
705 2012-06-18 16:35:59 <luke-jr> sipa: for the former, I'd pick a word that expresses "seeding period"
706 2012-06-18 16:36:02 <justmoon> jgarzik: I believe sipa's index fits in memory
707 2012-06-18 16:36:11 <MysteryBanshee> Bitcoin Address: 1ezrqcy23iRPy1mqFPdagoYa4xZDbGEMd
708 2012-06-18 16:36:11 <MysteryBanshee> Private Key (Wallet Import Format): 5JSCbRxWWtKLizLETD5QkSAR7ja2LUZXg2frXQRk7e1yQ3wB(XXX) - Guess last 3 digits
709 2012-06-18 16:36:11 <MysteryBanshee> Now contains 0.28 BTC (Nobody in #btc-mystery has been able to solve it :P)
710 2012-06-18 16:36:40 <sipa> MysteryBanshee: do you really want me to write 5-line bash script to solve that?
711 2012-06-18 16:36:44 <gmaxwell> jgarzik: what sipa's working on will let people run fully validating nodes (but without the whole history) with fairly modest amounts of disk space.
712 2012-06-18 16:36:58 <MysteryBanshee> sipa: please :)
713 2012-06-18 16:37:19 <jgarzik> gmaxwell: then (a) or (b) above
714 2012-06-18 16:37:25 <TheSeven> MysteryBanshee: 3 digits is kinda negligible bruteforce effort, and I bet you could actually calculate it
715 2012-06-18 16:37:38 <MysteryBanshee> 195191 keys
716 2012-06-18 16:37:40 <yellowhat> MysteryBanshee, if you upp the amount to 20BTC i (or someone else) will stop being lazy and collect it. :)
717 2012-06-18 16:37:40 t7 has joined
718 2012-06-18 16:37:47 <MysteryBanshee> haha
719 2012-06-18 16:38:13 <genjix> ill do it for 2 btc
720 2012-06-18 16:38:21 <sipa> jgarzik: well, feel free to implement your own storage layer then; i don't think using separate files will be any performance issue, and it's by far the easiest solution
721 2012-06-18 16:38:33 <luke-jr> sipa: how about seeding cycle vs difficulty cycle?
722 2012-06-18 16:39:17 <luke-jr> can I get some TN3 coins plz? n42rRpKcTpxkqkLBxEoqT3ddDGyLJmGZJJ
723 2012-06-18 16:39:45 <TheSeven> faucet is still TN2, right?
724 2012-06-18 16:40:22 agricocb has quit (Quit: Leaving.)
725 2012-06-18 16:41:05 <MysteryBanshee> genjix: i dont have 2btc atm lol, but will do one some time soon, altho it will probably be much harder to brute force :P
726 2012-06-18 16:41:18 <MysteryBanshee> Someones claimed it :)
727 2012-06-18 16:41:51 <jgarzik> sipa: make sure you follow the myriad portability rules required... no more than 14 character filenames, no more than 32k files or dirs in a directory, no larger than 4G single file, use atomically created lockfiles rather than flock, avoid mmap, ...
728 2012-06-18 16:42:30 <yellowhat> MysteryBanshee, your money was claimed
729 2012-06-18 16:43:02 <genjix> jgarzik: or use boost memory maps and boost lockfiles :)
730 2012-06-18 16:43:08 <sipa> jgarzik: what filesystem has a 14-character limit?
731 2012-06-18 16:43:10 <MysteryBanshee> was that you yellowhat?
732 2012-06-18 16:43:11 <genjix> and boost directory
733 2012-06-18 16:43:15 <gmaxwell> sipa: sys v.
734 2012-06-18 16:43:29 <gmaxwell> (old unix had a 14 character limit)
735 2012-06-18 16:43:38 <yellowhat> nope, i started writing a script with bitcoinJ but then i saw someone already claimed it
736 2012-06-18 16:43:46 <jgarzik> and derivatives that still appear on floppy images to this day
737 2012-06-18 16:43:50 <jgarzik> thanks, Sco!
738 2012-06-18 16:44:01 <gmaxwell> I think our source violates that..
739 2012-06-18 16:44:01 * sipa has not had a floppy drive in his computer for 10 years
740 2012-06-18 16:44:22 <jgarzik> yet we still boot with "floppy images"... :)
741 2012-06-18 16:44:38 <sipa> i don't intend to put the blockchain in a boot image
742 2012-06-18 16:47:13 <jgarzik> the general point being that there are reasons why people prefer to run databases on top of filesystems, rather than using the filesystem directly as a database... you may cluster your data together and tune it directly for your application
743 2012-06-18 16:48:03 <sipa> and i argue that this is not database usage, but very simple archive data storage
744 2012-06-18 16:48:05 <jgarzik> but it's cool if we want to relearn all these lessons :)
745 2012-06-18 16:48:41 chrisb__ has joined
746 2012-06-18 16:49:11 mologie has quit (Ping timeout: 240 seconds)
747 2012-06-18 16:51:22 mologie has joined
748 2012-06-18 16:51:55 RainbowDashh has joined
749 2012-06-18 16:55:25 RainbowDashh has quit (Client Quit)
750 2012-06-18 16:57:43 minimoose_ has joined
751 2012-06-18 16:59:30 <Diablo-D3> [12:11:44] <luke-jr> Diablo-D3: could you try to follow git-standard multiline commit messages in the future please? âº
752 2012-06-18 16:59:38 <Diablo-D3> why do you care what I write?
753 2012-06-18 16:59:48 minimoose has quit (Ping timeout: 252 seconds)
754 2012-06-18 16:59:49 minimoose_ is now known as minimoose
755 2012-06-18 17:00:33 <TD> sipa: what are you guys discussing?
756 2012-06-18 17:00:39 <TD> some other form of storage?
757 2012-06-18 17:00:48 <TD> btw leveldb seems to be a win
758 2012-06-18 17:00:58 <luke-jr> Diablo-D3: sorry, that was for Diapolo
759 2012-06-18 17:01:04 <Diablo-D3> ahh
760 2012-06-18 17:01:06 <Diablo-D3> because I was like
761 2012-06-18 17:01:11 <Diablo-D3> since when did you track my shit
762 2012-06-18 17:01:20 <luke-jr> he keeps doing these 500 letter long single-line commit messages
763 2012-06-18 17:01:22 <luke-jr> :p
764 2012-06-18 17:01:33 <Diablo-D3> _wtf_
765 2012-06-18 17:01:36 <Diablo-D3> even I dont do that
766 2012-06-18 17:02:15 <luke-jr> ;;later tell Diapolo could you try to follow git-standard multiline commit messages in the future please? that is, "Short summary of change" <blank line> "As many lines of details as you need"; thanks âº
767 2012-06-18 17:02:15 <gribble> The operation succeeded.
768 2012-06-18 17:02:27 <luke-jr> hmm, wonder if the quotes will survive
769 2012-06-18 17:02:56 <xorgate> i'm confused.. how does a bitcoin app know a certain tx is valid, that an input has enough funds to send to the output?
770 2012-06-18 17:04:02 <TD> xorgate: it knows about every transaction
771 2012-06-18 17:04:11 <TD> so it just looks up the connected outputs and sees how much value there is
772 2012-06-18 17:04:46 <xorgate> so in a way, it scans all the way down until it finds tx that have no previous? the blockreward tx ?
773 2012-06-18 17:05:08 <luke-jr> xorgate: no, it just remembers which transactions were already checked
774 2012-06-18 17:05:26 <xorgate> luke-jr yeh ok, it uses some bookkeeping tricks
775 2012-06-18 17:07:19 <TD> it uses a database
776 2012-06-18 17:07:50 agricocb has joined
777 2012-06-18 17:08:18 <sipa> TD (and others): my idea is this: keep a database (bdb?) with a map (txid -> compressed unspent txout data); that's already done and is about 100 MiB in size right now; in addition, keep all blocks in separate files (one file per block), and when connecting a block, create a file that contains "undo" information (=whatever that block removed from the database), so it can be rolled back in a reorg
778 2012-06-18 17:08:32 <xorgate> sure but right now i'm more interested in the theory than in the implementation
779 2012-06-18 17:08:48 <TD> sipa: that map is basically the existing blkindex.dat, right?
780 2012-06-18 17:08:55 <TD> sipa: why is this better?
781 2012-06-18 17:09:03 <sipa> TD: blkindex doesn't contain the actual txouts
782 2012-06-18 17:09:17 <sipa> TD: it still needs the blockchain to find the prevout transactions
783 2012-06-18 17:09:42 <TD> right, it contains the spent flags but not the rest of the data
784 2012-06-18 17:09:50 <sipa> and it stores the index information horribly inefficient (my db now *with* txout data itself is 6x smaller than blkindex.dat)
785 2012-06-18 17:09:51 <TD> txouts are largely uncompressible
786 2012-06-18 17:09:55 <TD> they're hashes and keys
787 2012-06-18 17:10:05 <TD> yes bdb is inefficient
788 2012-06-18 17:10:05 <sipa> no, they're amounts and scripts
789 2012-06-18 17:10:08 <TD> so far it seems leveldb can beat it easily
790 2012-06-18 17:10:13 <TD> yes, and scripts are mostly keys and hashes
791 2012-06-18 17:10:34 <sipa> TD: i got the entire txout data serialized in 64 MiB
792 2012-06-18 17:10:46 BNCatDIGISHELL has quit (Ping timeout: 265 seconds)
793 2012-06-18 17:10:57 <TD> and then what?
794 2012-06-18 17:10:57 <sipa> by using custom serialization for amounts and frequent scripts
795 2012-06-18 17:11:02 <TD> ok
796 2012-06-18 17:11:46 <sipa> i mean, i'm not arguing that obviously keys and hashes are not compressible, but compared to a naive serialization you can easily win 30-40%
797 2012-06-18 17:12:02 <sipa> by removing all the rest
798 2012-06-18 17:12:39 <sipa> TD: anyway, the ultimate point is that you can remove the block files when they are deep enough
799 2012-06-18 17:12:53 <sipa> they're only needed for rescans, reorgs and getdata's
800 2012-06-18 17:13:07 att has joined
801 2012-06-18 17:13:25 <TD> right
802 2012-06-18 17:13:41 <TD> but that's just a variant on pruning, no?
803 2012-06-18 17:13:45 <sipa> yes
804 2012-06-18 17:14:03 <gribble> New news from bitcoinrss: luke-jr opened issue 1482 on bitcoin/bitcoin <https://github.com/bitcoin/bitcoin/issues/1482>
805 2012-06-18 17:14:19 <sipa> though the smaller working set (txout database) should speed up verification as well, imho
806 2012-06-18 17:14:19 <TD> ok
807 2012-06-18 17:14:38 <sipa> as there are no accesses to the blockchain necessary during normal verification
808 2012-06-18 17:14:48 <TD> yeah
809 2012-06-18 17:14:55 <TD> if it can be held entirely in RAM, that does sound like a big win
810 2012-06-18 17:16:04 <sipa> so instead of pruning transactions when they're fully spent and deep enough, you prune separate txouts immediately, but keep recent blocks + undo information separately
811 2012-06-18 17:17:39 moartr4dez has left ()
812 2012-06-18 17:17:52 <TD> so
813 2012-06-18 17:18:05 <TD> on my leveldb setup whilst re-scanning the chain i see it spending about 20% of its time waiting for io
814 2012-06-18 17:18:11 <TD> 15-20%, normally
815 2012-06-18 17:18:28 <sipa> did you do a compare of a -loadblock with bdb and leveldb?
816 2012-06-18 17:18:36 Bwild has quit (Quit: leaving)
817 2012-06-18 17:18:42 <TD> i'm doing it now, with various settings
818 2012-06-18 17:18:45 <sipa> ok
819 2012-06-18 17:18:46 <TD> ah, hmm
820 2012-06-18 17:19:03 <sipa> on my laptop, it takes 31 minutes to load 172k blocks
821 2012-06-18 17:19:04 <TD> the 15-20% iowait time seems to be the background leveldb thread. the main thread spends almost no time in iowait and is using 100% of a cpu core
822 2012-06-18 17:19:25 <TD> so i guess that thread is basically maxed on sig verification
823 2012-06-18 17:19:40 <TD> hmm your laptop is much faster than my workstation, it seems. i wonder if LVM adds significant overhead
824 2012-06-18 17:19:50 <TD> so far i'm seeing 96 minutes for BDB with 185127 blocks
825 2012-06-18 17:19:55 <TD> though the last 10,000 blocks are a killer
826 2012-06-18 17:19:59 <TD> so that may be the difference
827 2012-06-18 17:20:17 <TD> times to verify at 182k are often 300-1000msec
828 2012-06-18 17:20:25 <MysteryBanshee> no block for 23 minutes
829 2012-06-18 17:20:26 <MysteryBanshee> for real?
830 2012-06-18 17:20:34 <sipa> ;;bc,tblb
831 2012-06-18 17:20:36 <gribble> Error: There's really no reason why you should have underscores or brackets in your mathematical expression. Please remove them.
832 2012-06-18 17:20:37 <TD> in contrast leveldb seems to manage in 50m minutes
833 2012-06-18 17:20:53 <TD> bdb generates a 611mb blkindex.dat and leveldb generates 397mb of data
834 2012-06-18 17:21:02 <sipa> TD: i7 2.2 GHz, 8 GiB RAM, bdb running on a userspace-encrypted filesystem (though i load the old blk0001.dat in ram beforehand)
835 2012-06-18 17:21:11 <TD> ah
836 2012-06-18 17:21:19 <TD> i wonder how much IO i lose to reading in the old blk0001.dat
837 2012-06-18 17:21:21 <TD> mine isn't in ram
838 2012-06-18 17:21:47 <sipa> it doesn't make a difference of more than a few minutes
839 2012-06-18 17:21:59 <TD> that makes sense
840 2012-06-18 17:22:17 <TD> one thing that confuses me is adding a 10 bit per key bloom filter makes little difference, though it's supposed to save a lot of seeks. i suspect the filter is getting saturated.
841 2012-06-18 17:22:23 <TD> the disk usage certainly goes up with one
842 2012-06-18 17:22:36 <TD> do you know how many txns there are currently, off hand?
843 2012-06-18 17:23:01 <TD> the other possibility is that it's not bottlenecked on IO at the moment. the saturated CPU core would seem to imply that
844 2012-06-18 17:23:19 <TD> maybe disabling sig verification is the next step, to make it purely io bound
845 2012-06-18 17:23:21 <sipa> 750k unspent txs, 1.7M unspent txouts
846 2012-06-18 17:24:42 minimoose_ has joined
847 2012-06-18 17:25:17 <TD> oh yeah, the other metric
848 2012-06-18 17:25:28 <TD> with bdb i seem to get a steady 3-5 megabytes/sec to disk, as seen by iotop
849 2012-06-18 17:25:33 <TD> with leveldb that's more like a sustained 25mb
850 2012-06-18 17:25:41 <sipa> interesting
851 2012-06-18 17:26:02 <TD> though that includes compaction traffic of course. ldb optimizes for reducing seeks over throughput
852 2012-06-18 17:26:04 <TD> at least in theory :)
853 2012-06-18 17:27:20 <MysteryBanshee> Uhm, why is the time between blocks so retarted all the time?
854 2012-06-18 17:27:33 <MysteryBanshee> there was like 10-11 blocks in an hour and now nothing for 30 minutes
855 2012-06-18 17:27:42 m00p has joined
856 2012-06-18 17:28:10 minimoose has quit (Ping timeout: 265 seconds)
857 2012-06-18 17:28:10 minimoose_ is now known as minimoose
858 2012-06-18 17:28:17 <sipa> MysteryBanshee: it's a poisson process
859 2012-06-18 17:28:41 <sipa> ;;bc,tblb 1h
860 2012-06-18 17:28:43 <gribble> 4 days, 18 hours, 26 minutes, and 50 seconds
861 2012-06-18 17:28:57 <sipa> on average you get one block every 10 minutes
862 2012-06-18 17:29:09 <sipa> but once every 4 days for example, one will take over an hour
863 2012-06-18 17:29:17 <sipa> ;;bc,tblb 30m
864 2012-06-18 17:29:19 <gribble> 4 hours, 9 minutes, and 14 seconds
865 2012-06-18 17:29:33 <drizztbsd> ;;bc,tblb 1d
866 2012-06-18 17:29:35 <gribble> Error: float division
867 2012-06-18 17:29:59 <drizztbsd> no blocks over 1 day, cool :P
868 2012-06-18 17:30:16 <sipa> ;;bc,tblb 3h
869 2012-06-18 17:30:17 <gribble> 7526 years, 24 weeks, 3 days, 13 hours, 30 minutes, and 19 seconds
870 2012-06-18 17:30:19 <TD> sipa: threading verification .... AFAICT the right place is CBlock::ConnectBlock, add a CS that protects mapQueuedChanges and then push each tx onto a worker pool queue
871 2012-06-18 17:30:24 <drizztbsd> LOL
872 2012-06-18 17:30:29 maaku has joined
873 2012-06-18 17:30:35 <drizztbsd> how can you calculate this?
874 2012-06-18 17:31:00 <MysteryBanshee> sipa: nobody is gonna tolerate waiting 4 hours 9 minutes for a transaction to go through
875 2012-06-18 17:31:14 <drizztbsd> MysteryBanshee: use litecoin :E
876 2012-06-18 17:31:14 <TD> sipa: once all txns are verified you opush mapQueuedChanges all at once
877 2012-06-18 17:31:19 <MysteryBanshee> :P
878 2012-06-18 17:31:21 <drizztbsd> or green address
879 2012-06-18 17:31:41 <TD> sipa: may need to check sigop count serially ...
880 2012-06-18 17:31:58 <sipa> TD: no need for a CS; mapQueuedChanges is only accessed within the block connection code, which runs in a single thread
881 2012-06-18 17:32:03 <TD> yeah
882 2012-06-18 17:32:04 <TD> but i mean
883 2012-06-18 17:32:09 <TD> if you wanted to thread tx verification
884 2012-06-18 17:32:13 <MysteryBanshee> green addresses will fix this?
885 2012-06-18 17:32:14 <TD> you'd have to change that, right? :)
886 2012-06-18 17:32:29 <MysteryBanshee> i thought it was impossible to send btc from an address that has 0 confirmations
887 2012-06-18 17:32:29 <TD> fetchinputs/connectinputs would run in parallel for different transactions in a block
888 2012-06-18 17:32:42 <sipa> MysteryBanshee: the protocol allows it, the client doesn't
889 2012-06-18 17:32:58 <sipa> MysteryBanshee: green addresses are not a scalable solution in my opinion; people would need to know all green addresses for each trusted business
890 2012-06-18 17:33:17 <TD> sipa: i guess the issue is you need to connect serially
891 2012-06-18 17:33:21 <TD> hrmmm
892 2012-06-18 17:33:34 <sipa> TD: yes, and transactions can depend on previous transactions in the same block
893 2012-06-18 17:33:37 <TD> right
894 2012-06-18 17:33:54 <sipa> tx verification is certainly threadable, but it's not trivial imho
895 2012-06-18 17:34:06 <TD> maybe the right design is to have all this code run serially, but the act of actually checking scripts be done on the thread pool. after doing all the connecting work, you wait for all scripts to report in as completed
896 2012-06-18 17:34:15 <TD> if any report in as failed, the block is invalid anyway ....
897 2012-06-18 17:34:24 <sipa> TD: that sounds like a nice design
898 2012-06-18 17:34:37 <TD> though, that said, the sig cache is supposed to make block verification super cheap right
899 2012-06-18 17:34:52 <sipa> uhu
900 2012-06-18 17:34:52 <TD> it's only when you use -loadblock that this looks like a bottleneck. for any long running node block handling should be totally IO bound, no?
901 2012-06-18 17:35:48 <justmoon> sipa: are you sure txs can spend others from the same block?
902 2012-06-18 17:35:53 <sipa> justmoon: yes
903 2012-06-18 17:35:59 <justmoon> sipa: 100% sure?
904 2012-06-18 17:36:01 <sipa> justmoon: yes
905 2012-06-18 17:36:16 <justmoon> ok that's kinda weird
906 2012-06-18 17:36:27 <justmoon> bitcoinjs verified in parallel, I don't remember running into that
907 2012-06-18 17:36:28 <sipa> justmoon: otherwise a 0-conf spend would not be valid
908 2012-06-18 17:36:33 <justmoon> although it was a long time ago
909 2012-06-18 17:36:38 <justmoon> verifies*
910 2012-06-18 17:37:02 Bwild has joined
911 2012-06-18 17:37:17 <TD> justmoon: it's valid in the protocol. satoshis block creation code won't create such blocks
912 2012-06-18 17:37:29 <justmoon> ah see, that makes more sense
913 2012-06-18 17:37:29 <TD> this sort of thing is why implementing bitcoin is such a PITA
914 2012-06-18 17:37:31 <sipa> sure it does
915 2012-06-18 17:37:34 <TD> it does?
916 2012-06-18 17:37:45 <sipa> when two dependent transactions get mined at the same time
917 2012-06-18 17:37:45 <TD> i thought i had looked at it and there was some issue that prevented it from doing that
918 2012-06-18 17:37:47 <genjix> justmoon: yes they can
919 2012-06-18 17:37:50 <justmoon> sipa: example tx or gtfo :P
920 2012-06-18 17:37:54 <genjix> but they must be chronological in the block
921 2012-06-18 17:37:54 RainbowDashh has joined
922 2012-06-18 17:38:25 <justmoon> ah nevermind
923 2012-06-18 17:38:37 <justmoon> yeah I'm looking for transactions in the same block first
924 2012-06-18 17:38:43 <justmoon> then looking at the actual database
925 2012-06-18 17:38:44 <MysteryBanshee> Whos idea was using a poissen distribution?
926 2012-06-18 17:38:44 <sipa> i'm sure deepbit's payouts will be an excellent example
927 2012-06-18 17:38:50 <sipa> MysteryBanshee: math's
928 2012-06-18 17:38:56 <MysteryBanshee> It might take 10 minutes on average, but what if someone wants to send funds in an emergency and all of a sudden it takes 2 hours:
929 2012-06-18 17:38:56 <[Tycho]> hello.
930 2012-06-18 17:38:57 <MysteryBanshee> ?
931 2012-06-18 17:38:58 <justmoon> so I just forgot implementing it
932 2012-06-18 17:39:02 <genjix> lol
933 2012-06-18 17:39:13 <genjix> no one picked a poisson
934 2012-06-18 17:39:25 <sipa> MysteryBanshee: bitcoin transactions are instantaneous; confirming takes a bit of time
935 2012-06-18 17:39:28 <justmoon> * forgot that I implemented it
936 2012-06-18 17:39:37 <sipa> MysteryBanshee: visa transactions take months to confirm
937 2012-06-18 17:39:53 <MysteryBanshee> yes but the money can be used instantly by the receiving party
938 2012-06-18 17:40:02 <TD> MysteryBanshee: you can use unconfirmed funds instantly
939 2012-06-18 17:40:13 <TD> MysteryBanshee: the current software doesn't make it easy, but nothing technically stops you
940 2012-06-18 17:40:25 <jgarzik> genjix: no one chose a fish? :)
941 2012-06-18 17:40:31 <TD> MysteryBanshee: accepting unconfirmed funds requires you trust the sender, that's all. credit cards are the same
942 2012-06-18 17:41:41 <MysteryBanshee> its just this waiting for a block thing is starting to become a pain
943 2012-06-18 17:41:44 <TD> justmoon: the guys who wrote the ETH paper want to come to the next meetup
944 2012-06-18 17:41:51 <TD> MysteryBanshee: what do you want to do with the funds you're receiving?
945 2012-06-18 17:41:55 <MysteryBanshee> isnt it possible to keep everything the same but have the blocks be 2 minutes?
946 2012-06-18 17:42:11 <MysteryBanshee> TD: I need to send them to someone but my ewallet doesnt allow sending unconfirmed
947 2012-06-18 17:42:15 <justmoon> TD: this is the artist I was talking about last time: http://www.brigittedaetwyler.ch/works
948 2012-06-18 17:42:17 <TD> who are you sending them to?
949 2012-06-18 17:42:32 <MysteryBanshee> TD: Someone who got a puzzle right
950 2012-06-18 17:42:40 <TD> MysteryBanshee: OK, so they trust you, right?
951 2012-06-18 17:42:47 <MysteryBanshee> yeh
952 2012-06-18 17:42:50 <MysteryBanshee> a little
953 2012-06-18 17:42:57 <genjix> jgarzik: a reference to jesus?
954 2012-06-18 17:43:08 <TD> justmoon: i have a feeling i've met her, actually
955 2012-06-18 17:43:10 <TD> the name rings a bell
956 2012-06-18 17:43:13 <TD> i may have sold her bitcoins
957 2012-06-18 17:43:18 <justmoon> TD: oh, mysterious :D
958 2012-06-18 17:43:37 <justmoon> TD: should we set a date? I'm tending towards doing it in zurich and pretty soon
959 2012-06-18 17:43:48 <TD> MysteryBanshee: technically nothing stops you from sending the funds on immediately. the e-wallet service you're using is imposing its own rules on you. but there's no specific requirement
960 2012-06-18 17:43:57 <justmoon> TD: July 7th maybe?
961 2012-06-18 17:44:08 <TD> MysteryBanshee: in future i think people carrying around long chains of unconfirmed transactions will become more common. the software just doesn't do a good job of supporting it today.
962 2012-06-18 17:44:11 <TD> justmoon: sounds ok
963 2012-06-18 17:44:45 <justmoon> TD: I'll be sure to mention your name then, maybe it rings a bell for her as well ;)
964 2012-06-18 17:44:52 <TD> MysteryBanshee: so see this as a temporary growing pain of bitcoin, not a fundamental flaw
965 2012-06-18 17:45:21 <TD> MysteryBanshee: anyway, there was a block 2 mins ago. happy now?
966 2012-06-18 17:45:31 <sipa> ;;bc,tslb
967 2012-06-18 17:45:35 <gribble> Error: unexpected EOF while parsing (<string>, line 1)
968 2012-06-18 17:45:43 * sipa hurts gribble
969 2012-06-18 17:49:10 <jgarzik> genjix: no, French
970 2012-06-18 17:49:26 <justmoon> TD: do you want to ask the eth guys if they have time before we announce the date?
971 2012-06-18 17:49:27 <genjix> *whoosh*
972 2012-06-18 17:49:29 <drizztbsd> ;;bc,tslb
973 2012-06-18 17:49:30 <gribble> Time since last block: 7 minutes and 8 seconds
974 2012-06-18 17:49:33 <MysteryBanshee> not only did i wait 40 minutes for a block, my transaction didnt even make it into the block
975 2012-06-18 17:49:34 <MysteryBanshee> GRRRRRRRRR
976 2012-06-18 17:49:36 <genjix> justmoon / TD, if you're in berlin http://bitcoin-hackathon.com/
977 2012-06-18 17:49:37 <TD> justmoon: i'll CC you
978 2012-06-18 17:49:39 <genjix> you can stay with me
979 2012-06-18 17:49:40 <drizztbsd> MysteryBanshee: fee
980 2012-06-18 17:49:54 <genjix> (sipa too since you're close)
981 2012-06-18 17:49:55 <drizztbsd> low priority transaction WITHOUT a fee required many blocks
982 2012-06-18 17:50:13 <MysteryBanshee> drizztbsd: its a 5 btc transaction it doesnt need a fee
983 2012-06-18 17:50:30 <drizztbsd> I never say it needs a fee
984 2012-06-18 17:50:32 <TD> fee has nothing to do with monetary value
985 2012-06-18 17:50:41 <MysteryBanshee> seriously that is BS, why didnt my client ask me for one?
986 2012-06-18 17:50:42 <TD> whether a fee helps or not is unintuitive
987 2012-06-18 17:50:46 <drizztbsd> some pools only accept transaction with fee
988 2012-06-18 17:50:55 <TD> MysteryBanshee: there's a lot of discussion right now around the right way to handle fees
989 2012-06-18 17:51:00 <TD> because we agree with you. the way it works now is BS
990 2012-06-18 17:51:03 <drizztbsd> and other pools give priority on transaction with fees
991 2012-06-18 17:51:08 <TD> satoshidice has shown this problem earlier than anticipated
992 2012-06-18 17:51:15 <MysteryBanshee> ok from now on im sending a 0.1337 fee with every transaction
993 2012-06-18 17:51:17 <MysteryBanshee> just to piss off the miners
994 2012-06-18 17:51:19 <TD> lol
995 2012-06-18 17:51:23 <drizztbsd> lol
996 2012-06-18 17:51:29 <TD> a fee for transactions you want to confirm very fast, will help at this time
997 2012-06-18 17:51:36 <TD> also, making it easier to spend unconfirmed coins
998 2012-06-18 17:51:43 <TD> as long as you know the risks of course
999 2012-06-18 17:51:50 <drizztbsd> spamdice waste many bytes of the block
1000 2012-06-18 17:51:54 <drizztbsd> :P
1001 2012-06-18 17:52:00 <TD> if you sent your unconfirmed money to somebody else, and the person who sent to you double spent, your puzzle prize would roll back too
1002 2012-06-18 17:52:27 <genjix> i think the block subsidy is far too high
1003 2012-06-18 17:53:06 <genjix> maybe we should make the next block have 18 million BTC and then rely on fees
1004 2012-06-18 17:53:26 <MysteryBanshee> I wasnt joking about the 0.1337 fee btw
1005 2012-06-18 17:53:28 <MysteryBanshee> :P
1006 2012-06-18 17:53:49 <MysteryBanshee> this is taking ageeees
1007 2012-06-18 17:53:54 BNCatDIGISHELL has joined
1008 2012-06-18 17:54:08 Ummon has quit (Read error: Connection reset by peer)
1009 2012-06-18 17:54:47 <MysteryBanshee> I sent another 2 BTC including a 0.1337 fee and its still not going in the block
1010 2012-06-18 17:55:00 <MysteryBanshee> Why is it not working?
1011 2012-06-18 17:55:03 <drizztbsd> wait for next block!
1012 2012-06-18 17:55:12 <MysteryBanshee> urgh!
1013 2012-06-18 17:56:58 <MysteryBanshee> and why does blockexplorer.com not work anymore??
1014 2012-06-18 17:57:02 <MysteryBanshee> why does nothign seem to work?
1015 2012-06-18 17:57:05 <MysteryBanshee> I cant take this any more
1016 2012-06-18 17:57:07 <MysteryBanshee> Argh!!!!
1017 2012-06-18 17:57:53 * sipa pours some valium into MysteryBanshee
1018 2012-06-18 17:58:03 * MysteryBanshee thanks sipa
1019 2012-06-18 17:58:30 <TD> heh
1020 2012-06-18 17:58:33 <TD> try blockchain.info
1021 2012-06-18 17:58:39 <genjix> too flashy
1022 2012-06-18 18:00:15 <TheSeven> hm, any way to tell bitcoind to spend funds from a particular address other than creating a new wallet around it?
1023 2012-06-18 18:00:18 danbri_ has joined
1024 2012-06-18 18:01:39 <TheSeven> oh, and what does bitcoind do if 1. someone is soloing, 2. it has an encrypted wallet, 3. that wallet is locked and 4. the keypool runs empty?
1025 2012-06-18 18:03:28 danbri has quit (Ping timeout: 260 seconds)
1026 2012-06-18 18:05:17 <gmaxwell> TheSeven: reuses keys.
1027 2012-06-18 18:05:49 <gmaxwell> TheSeven: as for your prior questionâ see the coinbaser patches.
1028 2012-06-18 18:05:55 <jeremias> http://bitcoin-hackathon.com/ bitcoin hackathon, 13-15 in berlin, maybe someone here might be interested joining
1029 2012-06-18 18:06:03 wizkid057 has quit (Read error: Operation timed out)
1030 2012-06-18 18:06:50 <genjix> sipa: ^
1031 2012-06-18 18:06:50 <gmaxwell> TheSeven: If that feature interest you it could use some more people to give it love and attention.
1032 2012-06-18 18:06:59 wizkid057 has joined
1033 2012-06-18 18:07:00 <MysteryBanshee> sipa: you get my pm?
1034 2012-06-18 18:07:53 <justmoon> jeremias: genjix just spammed it a couple minutes ago :P
1035 2012-06-18 18:08:03 <justmoon> genjix: I'll go if mike goes :)
1036 2012-06-18 18:08:07 <justmoon> TD: ^
1037 2012-06-18 18:08:39 <TD> might be interesting
1038 2012-06-18 18:08:46 <genjix> justmoon: well you can all stay with me. i have a double bed, 2 spare matresses + lots of floor
1039 2012-06-18 18:09:02 <TD> i have friends in berlin already i could stay with
1040 2012-06-18 18:09:12 <TD> those ideas are all wildly ambitious though :)
1041 2012-06-18 18:09:27 <justmoon> TD: come on, we'll have a blast!
1042 2012-06-18 18:09:42 <justmoon> TD: have you ever met andreas in person?
1043 2012-06-18 18:09:47 <TD> no
1044 2012-06-18 18:09:54 <justmoon> see? come on, let's do it :)
1045 2012-06-18 18:10:05 <genjix> he made bitcoin for android. it's a very well designed app i think
1046 2012-06-18 18:10:23 <genjix> i never used a mobile but already in minutes i was understanding it
1047 2012-06-18 18:10:30 <genjix> and he was explaining the design decisions
1048 2012-06-18 18:10:34 <jeremias> i have also space for one guy
1049 2012-06-18 18:10:40 <genjix> that was very cool
1050 2012-06-18 18:10:48 <jeremias> or maybe two if you are willing to sleep in a double bed
1051 2012-06-18 18:11:30 MC1984 has joined
1052 2012-06-18 18:16:24 drizztbsd has quit (Remote host closed the connection)
1053 2012-06-18 18:22:00 <yellowhat> jeremias, i will be at the hackathon :)
1054 2012-06-18 18:25:02 MysteryBansheeX has joined
1055 2012-06-18 18:25:18 rdponticelli has joined
1056 2012-06-18 18:25:46 <genjix> yellowhat: what's your email?
1057 2012-06-18 18:26:48 MysteryBanshee has quit (Ping timeout: 260 seconds)
1058 2012-06-18 18:27:06 MysteryBansheeX is now known as MysteryBanshee
1059 2012-06-18 18:27:35 toffoo has quit ()
1060 2012-06-18 18:28:17 ovidiusoft has joined
1061 2012-06-18 18:30:44 terry has quit (Quit: Connection closed for inactivity)
1062 2012-06-18 18:34:10 <sipa> genjix: 13-15 juli?
1063 2012-06-18 18:34:21 gavinandresen has joined
1064 2012-06-18 18:34:30 <sipa> in that case, too bad, i'll be on holyday in iceland :)
1065 2012-06-18 18:34:34 <sipa> holiday
1066 2012-06-18 18:35:22 rdponticelli has quit (Ping timeout: 276 seconds)
1067 2012-06-18 18:36:44 <gavinandresen> sipa: ok if I pull 1159, "Use block time for wallet transactions found in rescan" ?
1068 2012-06-18 18:37:05 <sipa> gavinandresen: luke has another version of that, which may be better
1069 2012-06-18 18:37:26 <sipa> i have little opinion about which is better
1070 2012-06-18 18:37:44 minimoose_ has joined
1071 2012-06-18 18:37:51 ThomasV has joined
1072 2012-06-18 18:39:19 <gavinandresen> yours I understand, is a small amount of code, and clearly seems better than the behavior we have now. I'm pulling it.
1073 2012-06-18 18:39:40 <gavinandresen> ... even more spiffier code can come later.
1074 2012-06-18 18:39:53 <sipa> gavinandresen: i think it's insufficiently tested
1075 2012-06-18 18:39:54 <genjix> sipa: ah pity. enjoy your holiday tho
1076 2012-06-18 18:40:07 rdponticelli has joined
1077 2012-06-18 18:40:47 minimoose has quit (Ping timeout: 245 seconds)
1078 2012-06-18 18:40:48 minimoose_ is now known as minimoose
1079 2012-06-18 18:41:31 <gavinandresen> sipa: ok, if it's not tested enough then I'll leave it.
1080 2012-06-18 18:42:33 <gavinandresen> sipa: any objections to me pulling 1399, Improve parsing of IPv6 addresses ?
1081 2012-06-18 18:42:41 <sipa> no
1082 2012-06-18 18:44:42 <gmaxwell> Thoughts on this? https://github.com/bitcoin/bitcoin/pull/1393 it's been surprisingly quiet. There was a big forum thread on it, and sipa had a prior version of it.
1083 2012-06-18 18:45:22 <gavinandresen> "Status: I have not tested this at all yet, other than writing and running tests for the ordering"
1084 2012-06-18 18:45:27 <gavinandresen> has that changed?
1085 2012-06-18 18:46:43 <gmaxwell> I haven'tâ didn't want to bother if someone was going to step up and say they didn't like the intended behavior.
1086 2012-06-18 18:47:02 TD has quit (Quit: TD)
1087 2012-06-18 18:48:23 <sipa> i've frequently seen reports about people complaining about reported transaction times
1088 2012-06-18 18:48:55 <gavinandresen> my only concern is changing the way transactions are timestamped is the kind of little simple-seeming change that might eat up tons of time if there are services that assume the old behavior and break
1089 2012-06-18 18:49:09 <sipa> but almost no discussion about these patches
1090 2012-06-18 18:49:12 <gavinandresen> .... and fixing it just isn't anywhere near the top of my priority list
1091 2012-06-18 18:50:04 RainbowDashh has quit (Quit: Computer has gone to sleep. DERPY HOOVES SLAMMING THE LAPTOP LID MISTAKE?)
1092 2012-06-18 18:51:24 MC1984 has quit (Quit: Leaving)
1093 2012-06-18 18:51:53 m00p has quit (Ping timeout: 260 seconds)
1094 2012-06-18 18:54:10 theorb has joined
1095 2012-06-18 18:54:37 theorbtwo has quit (Remote host closed the connection)
1096 2012-06-18 18:54:51 theorb is now known as theorbtwo
1097 2012-06-18 18:57:05 <TheSeven> bitcoind can't deal with multiple wallets but just one blockchain copy, right?
1098 2012-06-18 18:58:02 <TheSeven> what about an extended importprivkey command that doesn't import the key but instead sends the funds from a key to a new one from the keypool?
1099 2012-06-18 18:58:29 <luke-jr> sweepprivkey
1100 2012-06-18 18:58:43 <TheSeven> luke-jr: is there such a thing? :)
1101 2012-06-18 18:58:47 <TheSeven> or a patch for that?
1102 2012-06-18 18:58:53 <luke-jr> dunno
1103 2012-06-18 18:59:46 <yellowhat> google is creepy. really. marked the days 13-15 july to add the hackathon to my schedule and it proposed: "Wochenendtrip nach Berlin" "Weekend trip to Berlin"
1104 2012-06-18 19:03:32 ThomasV has quit (Quit: Quitte)
1105 2012-06-18 19:04:52 <sipa> what was CTxDB::ReadOwnerTxes ever used for?
1106 2012-06-18 19:05:04 <sipa> maybe it predates the wallet logic
1107 2012-06-18 19:06:07 theymos has joined
1108 2012-06-18 19:07:20 <gavinandresen> that'll teach me to pull before running unit tests... test/netbase_tests.cpp:89: error in "netbase_lookupnumeric": check TestParse("128.5.1", "128.5.0.1:65535") failed
1109 2012-06-18 19:07:49 <sipa> gavinandresen: hmm, interesting
1110 2012-06-18 19:07:53 datagutt has quit (Quit: Computer has gone to sleep.)
1111 2012-06-18 19:07:55 <sipa> seems OS depending
1112 2012-06-18 19:08:15 <sipa> we don't rely on that behaviour, so maybe i shouldn't have put it in a unit test
1113 2012-06-18 19:08:22 <gavinandresen> testing on my mac
1114 2012-06-18 19:08:26 <gmaxwell> TD: good job with the leveldb testing.
1115 2012-06-18 19:08:27 kreal has joined
1116 2012-06-18 19:08:47 <kreal> Can I get bitcoind to list over 10 transactions without providing an account?
1117 2012-06-18 19:08:54 <gmaxwell> TD[gone]: those were the kind of improvements jgarzik was expecting from swithing to hash indexing from btree but hadn't realized.
1118 2012-06-18 19:08:59 <kreal> bitcoind listtransactions 1000 <-- ?
1119 2012-06-18 19:09:02 <gmaxwell> kreal: use account ""
1120 2012-06-18 19:09:13 <gmaxwell> (or "*" ? )
1121 2012-06-18 19:09:17 <kreal> hmm *
1122 2012-06-18 19:09:17 <gavinandresen> "*"
1123 2012-06-18 19:09:18 wizkid057 has quit (Quit: brb)
1124 2012-06-18 19:09:31 <jgarzik> leveldb says 'ordered' thus it is btree?
1125 2012-06-18 19:09:31 <kreal> awesome thanks alot.
1126 2012-06-18 19:10:06 <jgarzik> DB_HASH got bigger and seemed slower. DB_BTREE w/ 3 databases saved 10% on space, and shaved at least 20 seconds off -loadblock
1127 2012-06-18 19:10:33 <sipa> jgarzik: do you happen to know how large blkhash.dat got?
1128 2012-06-18 19:10:51 <jgarzik> sipa: should be a pastebin with numbers in the pullreq
1129 2012-06-18 19:11:03 <jgarzik> sipa: it was not very large
1130 2012-06-18 19:11:31 <jgarzik> sipa: at the time I thought we should remove it from bdb, and simply use a flat file
1131 2012-06-18 19:12:16 <theymos> gmaxwell: What do you think of etotheipi's proposal ( https://bitcointalk.org/index.php?topic=88208.0 )? It doesn't seem much better than a plain Merkle tree of unspent outputs to me, but I thought I remembered you saying you liked it.
1132 2012-06-18 19:12:52 <jgarzik> so, a bit counter-intuitive that DB_HASH was not a win, but according to the documentation, setting the hash fill factor and bucket packing is important, and I did not do either
1133 2012-06-18 19:13:07 <jgarzik> splitting blkindex.dat into 3 databases was definitely a minor win
1134 2012-06-18 19:13:56 <jgarzik> a pity that TD[gone] did not post a link to his bitcoind+leveldb changes
1135 2012-06-18 19:14:13 <jgarzik> that might have been useful for my idea to test with KyotoCabinet
1136 2012-06-18 19:14:30 <gmaxwell> theymos: I don't think it's much better than a plain merkle tree of unspent outputs. I still don't think anyone has solved the unbalancing problem. amiller says he has but I haven't convinced myself of this.
1137 2012-06-18 19:14:47 <amiller> gmaxwell, have you looked at my code at all
1138 2012-06-18 19:15:01 <theymos> gmaxwell: OK, thanks.
1139 2012-06-18 19:15:08 <amiller> i'm not sure what to work on to help this get clarified
1140 2012-06-18 19:15:46 <gmaxwell> amiller: I just haven't had the free cycles to look into it yet. "read my code" is an okay thing to say, but it's not the lowest effort thing to do.
1141 2012-06-18 19:16:47 <gmaxwell> In any case, the biggest challenge in that whole space is securely preventing unbalancing. If thats now solved⦠great. Me being convinced of that doesn't have to happen quickly if it's true.
1142 2012-06-18 19:16:48 <sipa> you need a 160-bit key; use a btree with fanout of 16 or 256, with leaf nodes being the data, and the hashes of the children stored in the inner nodes
1143 2012-06-18 19:16:58 <sipa> so you have 5 or 10 levels whatever happens
1144 2012-06-18 19:17:47 <sipa> when updating a leaf, you update the 5 or 10 parent nodes by dropping the last 4-8 bits of the key, and update the corresponding data element, and hash it again, and recurse
1145 2012-06-18 19:18:23 <gmaxwell> yea, iff the key is always a hash I think you can make unbalancing require long partial preimage attacks, which is probably fine.
1146 2012-06-18 19:18:48 <gmaxwell> But if you want to make the key something else (e.g. perhaps some kinds of addresses, or domain names for the namecoin like use case) then this is far less obviously safe to me.
1147 2012-06-18 19:20:33 <amiller> the balanced binary tree approach is far simpler than the bucket/hash approach
1148 2012-06-18 19:20:41 <amiller> it also gives you a worst-case guarantee of O(log N)
1149 2012-06-18 19:21:23 O2made has joined
1150 2012-06-18 19:21:54 <gmaxwell> amiller: how can you give me a provable No-transaction-found without giving me the whole tree in a balanced tree case?
1151 2012-06-18 19:21:55 <amiller> take a binary search tree, have each parent node also contain a hash for its children, now it's a merkle search tree
1152 2012-06-18 19:22:14 <amiller> gmaxwell, i give you the paths through the tree to two elements a0 and c0
1153 2012-06-18 19:22:18 <amiller> you can prove that they are adjacent
1154 2012-06-18 19:22:26 <amiller> if they're adjacent, then there's no b0 in there
1155 2012-06-18 19:22:40 O2made has quit (Client Quit)
1156 2012-06-18 19:23:37 <amiller> i give you enough information that you can simulate performing the query yourself
1157 2012-06-18 19:23:40 <amiller> beginning with the data for the root node
1158 2012-06-18 19:24:05 <gmaxwell> Okay, that makes sense to me.
1159 2012-06-18 19:24:21 <[Tycho]> Who is Piotr Piasecki ?
1160 2012-06-18 19:27:22 <amiller> anything operation you can perform on a regular structure, you can do _securely_ with a structure augmented with merkle hashes
1161 2012-06-18 19:27:53 <jgarzik> never heard of him
1162 2012-06-18 19:28:22 <amiller> everywhere you have a link or a pointer, you store the hash of the data right next to it. So every time you would follow a link, you know what the data is in advance, starting with the root
1163 2012-06-18 19:29:51 <amiller> the blockchain for example, is a linked list, augmented with merkle hashes
1164 2012-06-18 19:30:51 m00p has joined
1165 2012-06-18 19:33:27 <luke-jr> rcorreia: ping
1166 2012-06-18 19:34:31 jurov is now known as away!aktooj@84.245.71.31|jurov
1167 2012-06-18 19:34:43 darkee has joined
1168 2012-06-18 19:36:58 D34TH has quit (Quit: Leaving)
1169 2012-06-18 19:37:07 RazielZ has quit (Ping timeout: 276 seconds)
1170 2012-06-18 19:37:46 darkee has quit (Ping timeout: 276 seconds)
1171 2012-06-18 19:37:59 TD has joined
1172 2012-06-18 19:38:01 <theymos> [Tycho]: ThePiachu, I think.
1173 2012-06-18 19:38:16 <[Tycho]> Cool.
1174 2012-06-18 19:38:17 RazielZ has joined
1175 2012-06-18 19:38:20 <[Tycho]> But who is he ?
1176 2012-06-18 19:38:22 MiningBuddy has joined
1177 2012-06-18 19:39:09 genjix has quit (Ping timeout: 265 seconds)
1178 2012-06-18 19:39:32 <theymos> A forum user. Posts on the Bitcoin Stack Exchange a lot. Recently published a Bitcoin-related paper (which I haven't read yet).
1179 2012-06-18 19:39:52 Raziel_ has joined
1180 2012-06-18 19:40:20 <luke-jr> ThePiachu runs that pool
1181 2012-06-18 19:40:25 <luke-jr> for mining vanity addresses
1182 2012-06-18 19:40:58 MiningBuddy- has quit (Ping timeout: 248 seconds)
1183 2012-06-18 19:41:00 <[Tycho]> Cool.
1184 2012-06-18 19:42:45 erle- has joined
1185 2012-06-18 19:42:57 wizkid057 has joined
1186 2012-06-18 19:43:37 RazielZ has quit (Ping timeout: 276 seconds)
1187 2012-06-18 19:44:32 maaku has quit (Quit: maaku)
1188 2012-06-18 19:45:14 <luke-jr> ;;later tell rcorreia is -n for uint64 well-defined behaviour?
1189 2012-06-18 19:45:14 <gribble> The operation succeeded.
1190 2012-06-18 19:46:09 <luke-jr> ;;later tell rcorreia also is uint64=negative int64 well-defined behaviour pre-C++11 too?
1191 2012-06-18 19:46:10 <gribble> The operation succeeded.
1192 2012-06-18 19:49:53 <theymos> -n wraparound for unsigned primitive types is standard.
1193 2012-06-18 19:57:21 <gmaxwell> luke-jr: wraparound on unsigned is well specified.
1194 2012-06-18 19:57:30 <gmaxwell> (it even has to work on ones compliment machines)
1195 2012-06-18 19:58:07 <gmaxwell> On unsigned wraparound is unspecified and the compiler is free to produce code which behaves differently in that case.
1196 2012-06-18 19:58:16 <gmaxwell> gah _SIGNED_
1197 2012-06-18 19:58:42 <luke-jr> gmaxwell: I'm confused more now :P
1198 2012-06-18 19:59:15 <theymos> Conversion from signed to unsigned is defined unless the signed number is negative IIRC.
1199 2012-06-18 19:59:47 <gmaxwell> Unsigned is well specified to wrap. Signed's behavior on over/underflow is unspecified.
1200 2012-06-18 19:59:53 <gmaxwell> Conversion is _implementation defined_
1201 2012-06-18 20:00:17 <gmaxwell> At least in C89/C99, usually C++ is the same for this stuff.
1202 2012-06-18 20:00:29 <luke-jr> hmm
1203 2012-06-18 20:01:15 <BlueMattBot> Project Bitcoin-Test build #314: FAILURE in 34 min: http://jenkins.bluematt.me/job/Bitcoin-Test/314/
1204 2012-06-18 20:01:16 <BlueMattBot> * pieter.wuille: Improve parsing of IPv6 addresses
1205 2012-06-18 20:01:16 <BlueMattBot> * pieter.wuille: Add netbase tests
1206 2012-06-18 20:01:17 <BlueMattBot> * fanquake: Further updates to build instructions
1207 2012-06-18 20:01:17 <BlueMattBot> * fanquake: OS X not OSX
1208 2012-06-18 20:01:19 <Diablo-D3> gmaxwell: yes but
1209 2012-06-18 20:01:20 <luke-jr> I'm not sure what "wrap" means here; uint= -uint ?
1210 2012-06-18 20:01:23 <Diablo-D3> its guaranteed to be sane
1211 2012-06-18 20:01:26 <gmaxwell> (unspecified means that the program is permitted to trigger false vacuum collapse and end the universe. though in practice it just rarely causes overflow tests to get optimized out. :) )
1212 2012-06-18 20:01:44 <Diablo-D3> luke-jr: ever write -1 for an uint?
1213 2012-06-18 20:01:48 <luke-jr> Diablo-D3: no.
1214 2012-06-18 20:01:50 <Diablo-D3> it wraps to max_uint
1215 2012-06-18 20:02:23 <luke-jr> I just want to verify 5849bd4 (in master now) is well-defined under older C++ compilers, basically :p
1216 2012-06-18 20:02:23 <gmaxwell> Diablo-D3: -1 unit is an implementation defined conversion. (the -1 literal is a signed value)
1217 2012-06-18 20:02:49 <Diablo-D3> gmaxwell: yes, and the result is guaranteed to make sense for that platform.
1218 2012-06-18 20:03:02 <gmaxwell> Diablo-D3: depends on the compiler implementation.
1219 2012-06-18 20:03:21 graingert has quit (Ping timeout: 246 seconds)
1220 2012-06-18 20:03:38 graingert has joined
1221 2012-06-18 20:03:38 <Diablo-D3> gmaxwell: let me rephrase
1222 2012-06-18 20:03:51 <Diablo-D3> if the platform is fucktarded, no one is using it\
1223 2012-06-18 20:07:28 egecko has joined
1224 2012-06-18 20:08:40 <luke-jr> gmaxwell: so 'uint a = (int)-5' is NOT well-defined? or is?
1225 2012-06-18 20:08:56 <gmaxwell> luke-jr: it's implementation defined.
1226 2012-06-18 20:09:16 <gmaxwell> In the C standards there are three states, specified, implementation defined, and undefined.
1227 2012-06-18 20:09:18 <luke-jr> hmm, so the comment is wrong :|
1228 2012-06-18 20:09:52 <sipa> so it's guaranteed to result in sane behaviour on every platform, but the result will not necessarily be identical between platforms?
1229 2012-06-18 20:09:57 <gmaxwell> The third case you never want to use because the compiler is permitted to "miscompile" code so that it behaves differently in cases where undefined things would happen.
1230 2012-06-18 20:10:44 <sipa> undefined == byzantine failure is permitted
1231 2012-06-18 20:10:58 <gmaxwell> And actually happens sometimes.
1232 2012-06-18 20:11:22 <theymos> Yeah, I believe that 5849bd4 is wrong and the original is correct.
1233 2012-06-18 20:11:28 <gmaxwell> "why did the compiler remove all that code!" "it could tell that it would only happen after an undefined operation"
1234 2012-06-18 20:11:36 <luke-jr> ok, so "uint a = (int)-5; uint b = -a;" should be safe?
1235 2012-06-18 20:12:13 <luke-jr> theymos: I think 5849bd4 is trying to address the case where n is a negative number, that the signed type cannot represent in positive
1236 2012-06-18 20:13:00 toffoo has joined
1237 2012-06-18 20:13:16 <theymos> luke-jr: Yes, but it seems to result in implementation-defined behavior because it converts a negative signed number to unsigned. The original version ensures that the signed number is positive before converting to unsigned, which is defined.
1238 2012-06-18 20:13:41 <gmaxwell> What theymos says.
1239 2012-06-18 20:13:52 <luke-jr> hmm
1240 2012-06-18 20:14:07 <gmaxwell> -a is an operation entirely with signed values.
1241 2012-06-18 20:14:19 <gmaxwell> then it's cast to uint, but it's always (we hope!) postive there.
1242 2012-06-18 20:14:22 <theymos> There is an issue when sn is the signed minimum value, since it doesn't have an inverse.
1243 2012-06-18 20:14:31 <luke-jr> so basically, we want to revert 5849bd4 and special-case the minimum value
1244 2012-06-18 20:14:31 <theymos> (in 2's complement.)
1245 2012-06-18 20:14:50 <luke-jr> is every other negative value guaranteed to have a positive representation? :/
1246 2012-06-18 20:14:59 <gmaxwell> Though generally I don't think we should put in special effort to avoid twos compliment common implementation defined behaviors. We absolutely should not use undefined behavior.
1247 2012-06-18 20:15:09 <TheSeven> sendmany with one target == sendfrom, right?
1248 2012-06-18 20:15:29 <gmaxwell> (even when the undefined behavior works)
1249 2012-06-18 20:15:36 <sipa> TheSeven: yes
1250 2012-06-18 20:15:39 <TheSeven> i.e. sendfrom/sendtoaddress are just subsets of sendmany...
1251 2012-06-18 20:15:54 <theymos> luke-jr: If possible, it seems to me that an exception should be thrown when sn is the minimum value an int64 can hold.
1252 2012-06-18 20:16:10 <theymos> luke-jr: And other than that, the original version of that code should be restored.
1253 2012-06-18 20:16:49 <luke-jr> theymos: why an exception?
1254 2012-06-18 20:18:11 <theymos> Because the input is invalid. std::domain_error or something. (Maybe Bitcoin doesn't use exceptions like this. It's just what I would do if I was writing that code.)
1255 2012-06-18 20:20:57 drizztbsd has joined
1256 2012-06-18 20:20:59 BlueMatt has joined
1257 2012-06-18 20:20:59 BlueMatt has quit (Changing host)
1258 2012-06-18 20:20:59 BlueMatt has joined
1259 2012-06-18 20:21:31 <luke-jr> theymos: I don't see how it's invalidâ¦
1260 2012-06-18 20:22:13 PK has joined
1261 2012-06-18 20:23:47 justmoon has quit (Quit: Leaving)
1262 2012-06-18 20:26:44 <luke-jr> how about this: n = -(sn + 1); ++n;
1263 2012-06-18 20:27:31 wizkid057 has quit (Read error: Operation timed out)
1264 2012-06-18 20:28:07 <theymos> Then you'll have implementation-defined overflow when sn is its max value.
1265 2012-06-18 20:28:20 <theymos> (You're right about exceptions.)
1266 2012-06-18 20:28:23 wizkid057 has joined
1267 2012-06-18 20:28:53 <luke-jr> theymos: that code only runs if sn is negative.
1268 2012-06-18 20:29:07 <gavinandresen> RE: signed/unsigned arithmetic stuff: please discuss this with the author of the pull, he seems to be an expert
1269 2012-06-18 20:29:47 <TheSeven> sipa: it isn't... comment-to is missing from sendmany
1270 2012-06-18 20:30:02 <sipa> TheSeven: right
1271 2012-06-18 20:30:55 <gmaxwell> theymos: overflow of signed values in C is unspecified, not implementation defined. (though its uncommon to see byzantine failure from it)
1272 2012-06-18 20:31:07 Diapolo has joined
1273 2012-06-18 20:31:12 MC1984 has joined
1274 2012-06-18 20:31:13 <theymos> luke-jr: OK, I think that's safe. (I was going to suggest using stringstreams.)
1275 2012-06-18 20:31:41 <luke-jr> gavinandresen: yeah, the discussion started with a ping to him ;)
1276 2012-06-18 20:32:26 <luke-jr> gavinandresen: shall I wait to make it a pullreq until he's responded?
1277 2012-06-18 20:33:08 <luke-jr> (as opposed to having the pullreq open for him to respond to there)
1278 2012-06-18 20:33:11 <gavinandresen> yes. Otherwise you're playing whack-a-mole
1279 2012-06-18 20:33:36 <Diapolo> Is there a reason, why no unit (e.g. BTC) is displayed in transaction list? Just asking, because I'm working there currently ...
1280 2012-06-18 20:34:16 gavinandresen has quit (Quit: gavinandresen)
1281 2012-06-18 20:34:19 <luke-jr> Diapolo: Spesmilo did this. It isn't very pretty.
1282 2012-06-18 20:34:43 <Diapolo> It could be pretty easily changed ... wanted yes / no?
1283 2012-06-18 20:34:49 <luke-jr> probably not.
1284 2012-06-18 20:35:04 <luke-jr> probably user group A would go up in arms against it
1285 2012-06-18 20:35:12 <luke-jr> then when it's reverted, user group B would go up in arms that they liked it
1286 2012-06-18 20:35:18 maaku has joined
1287 2012-06-18 20:35:30 <Diapolo> ^^ I know that war (thinks of the progressbar)
1288 2012-06-18 20:35:38 <gmaxwell> Diapolo: generally we should avoid adding more configurable options unless there is a strong need. It's more code to test, more gratitious difference in behavior that might confuse people.
1289 2012-06-18 20:36:14 <Diapolo> gmaxwell: I didn't want to make this selectable, just asking if a hard-coded change makes any sense.
1290 2012-06-18 20:36:34 <Diapolo> no problem with leaving this untouched
1291 2012-06-18 20:37:05 <gmaxwell> Ah I thought you were suggesting making it an option with the yes /no. I don't have any general opinion otherwise.
1292 2012-06-18 20:37:17 <luke-jr> Diapolo: I *do* think it would look nicer, if digits after 2 decimal places were displayed in a slightly smaller font size.
1293 2012-06-18 20:37:29 <Diapolo> it seems to be the only place in the GUI that displays amounts without unit
1294 2012-06-18 20:37:59 <luke-jr> also, I find often that I'd like to see a total balance based only on my filtered search results ;)
1295 2012-06-18 20:38:04 <Diapolo> luke-jr: that sounds like a different story ^^
1296 2012-06-18 20:38:07 <luke-jr> hehe
1297 2012-06-18 20:38:22 <luke-jr> you seemed bored ;)
1298 2012-06-18 20:38:35 <Diapolo> but I agree ... and must admit the code in that part of the client looks a little hacked ...
1299 2012-06-18 20:38:44 theymos has quit (Remote host closed the connection)
1300 2012-06-18 20:38:45 <gmaxwell> Diapolo: go make coin control awesome so we can merge it.
1301 2012-06-18 20:39:23 <Diapolo> gmaxwell: I was working on that one today ...
1302 2012-06-18 20:39:35 <gmaxwell> I saw you mention it earlier thats why I suggested it.
1303 2012-06-18 20:39:43 <Diapolo> but as I didn't yet write a good commit message luke refuses to comment ^^
1304 2012-06-18 20:40:25 <luke-jr> Diapolo: I'm updating stable first :p
1305 2012-06-18 20:40:44 <Diapolo> gmaxwell: you can try it out ... https://github.com/Diapolo/bitcoin/tree/coincontrol
1306 2012-06-18 20:40:49 <gmaxwell> haha. Well then! The git style is "<=70ch summary\n\nBlather no one reads". now you know.
1307 2012-06-18 20:41:07 <Diapolo> luke-jr: I didn't want you to update just take a look and try it first :). Consider it work-in-progress.
1308 2012-06-18 20:41:51 <gmaxwell> Diapolo: I vaguely recall sipa had some structural complaint about how the input restriction was actually applied.
1309 2012-06-18 20:42:09 <luke-jr> yeah, it's pretty sucky atm
1310 2012-06-18 20:42:23 <Diapolo> that's true the internals need rework by skilled devs
1311 2012-06-18 20:42:25 <luke-jr> ;;later tell rcorreia Can you comment on https://github.com/luke-jr/bitcoin/commit/0f5a2a82d93f55e1f5ad1c5401b8b5ceeb38b0df ?
1312 2012-06-18 20:42:26 <gribble> The operation succeeded.
1313 2012-06-18 20:42:50 <gmaxwell> Diapolo: It can be reviewed by skilled devs. Telling you what you did wrong is probably a better use of people's time than actually fixing it themselves.
1314 2012-06-18 20:42:54 <Diapolo> I can polish, not really implement coincontrol, sorry :).
1315 2012-06-18 20:42:56 maaku has quit (Quit: maaku)
1316 2012-06-18 20:43:18 <luke-jr> Diapolo: this part counts a polish imo
1317 2012-06-18 20:43:22 maaku has joined
1318 2012-06-18 20:43:47 <luke-jr> Diapolo: basically, the restriction should be made into an argument for the pertinent functions rather than a value set on the object
1319 2012-06-18 20:43:59 <Diapolo> I have a hard time understanding the code sipa didn't like ... I had some input to make GUI wise, think nothing wrong with that?
1320 2012-06-18 20:44:12 <luke-jr> (or you could make a "CTransactionMaker" class that you set options on then ask to make a CTransaction)
1321 2012-06-18 20:44:48 <Diapolo> I'm no atomic-science-phsyician -_- at least not yet.
1322 2012-06-18 20:44:56 <gmaxwell> Diapolo: nothing wrong with that, but the feature won't be merged if more people aren't willing to work on itâ and that includes more than just the GUI. (though perhaps if you'll handle the GUI someone will do the rest)
1323 2012-06-18 20:44:56 <luke-jr> Diapolo: you're getting there ;)
1324 2012-06-18 20:45:06 <luke-jr> Diapolo: I've had to wince a lot less at your pullreqs lately ;)
1325 2012-06-18 20:45:39 <Diapolo> luke-jr: that is true :) and it's a lot of code in the client, so I'm getting better ...
1326 2012-06-18 20:46:55 <Diapolo> did I say I hate that #ifdef Q_WS_MAC stuff for layouting ... damn it's more ugly than #ifdefs for WIN32 API wise ...
1327 2012-06-18 20:48:25 JZavala has joined
1328 2012-06-18 20:51:20 maaku has quit (Quit: maaku)
1329 2012-06-18 20:51:47 minimoose_ has joined
1330 2012-06-18 20:52:25 RainbowDashh has joined
1331 2012-06-18 20:53:50 minimoose__ has joined
1332 2012-06-18 20:54:04 minimoose__ has quit (Client Quit)
1333 2012-06-18 20:54:08 minimoose has quit (Ping timeout: 256 seconds)
1334 2012-06-18 20:54:30 jurov is now known as jurov|away
1335 2012-06-18 20:54:47 maaku has joined
1336 2012-06-18 20:56:12 minimoose_ has quit (Ping timeout: 245 seconds)
1337 2012-06-18 20:56:46 <Diapolo> luke-jr: As working on too many patches in parallel is rather ineffective, did you try out the sign/verify thing again? I would like to see this in 0.7. Wumups didn't comment yet though.
1338 2012-06-18 20:57:08 <luke-jr> Diapolo: that's on my list XD
1339 2012-06-18 20:57:57 <gribble> New news from bitcoinrss: Diapolo opened pull request 1483 on bitcoin/bitcoin <https://github.com/bitcoin/bitcoin/pull/1483>
1340 2012-06-18 20:58:36 <Diapolo> luke-jr: ^^ I put so much effort in that commit-message :-P.
1341 2012-06-18 20:58:48 <Diapolo> (for sign/verify)
1342 2012-06-18 20:59:12 <luke-jr> heh
1343 2012-06-18 21:03:54 <sipa> TD[gone]: how did you solve the reading-uncommitted-changes thing?
1344 2012-06-18 21:05:26 <TD> you can iterate over a WriteBatch
1345 2012-06-18 21:05:43 <TD> so in the CTxDB::Read method i just do that if there's an active batch
1346 2012-06-18 21:05:49 <sipa> i see
1347 2012-06-18 21:06:18 <sipa> i was hoping you unentangled the db-update code and the block connection logic
1348 2012-06-18 21:06:54 darkee is now known as !~darkee@gateway/tor-sasl/darkee|darkee
1349 2012-06-18 21:08:36 <sipa> which doesn't seem so hard: we already maintain a mapQueuedChanges; all is necessary is doing that at a higher level than just one block
1350 2012-06-18 21:10:38 <xorgate> ubuntu
1351 2012-06-18 21:10:41 <xorgate> oops
1352 2012-06-18 21:11:05 <luke-jr> https://github.com/bitcoin/bitcoin/pull/1484/files âº
1353 2012-06-18 21:12:15 <luke-jr> anyone want to test that it's deterministic across multiple people? ;)
1354 2012-06-18 21:12:33 <luke-jr> (it should be: I tested by rebuilding the whole thing from scratch twice)
1355 2012-06-18 21:13:08 <gribble> New news from bitcoinrss: luke-jr opened pull request 1484 on bitcoin/bitcoin <https://github.com/bitcoin/bitcoin/pull/1484>
1356 2012-06-18 21:14:28 JZavala has quit (Ping timeout: 250 seconds)
1357 2012-06-18 21:15:36 att has quit (Ping timeout: 255 seconds)
1358 2012-06-18 21:15:58 <TD> sipa: no, i wanted a minimally invasive patch
1359 2012-06-18 21:16:11 <sipa> understandable
1360 2012-06-18 21:16:11 <TD> sipa: the code is still too messy for a pull req .... we need to decide how to manage the new dependency
1361 2012-06-18 21:16:30 <TD> but i'll upload it at some point. it should be fairly clear that it's correct
1362 2012-06-18 21:17:48 <sipa> TD: i'd like to change more than that anyway; if the block-update process can be free of database updates until the connection is guaranteed to be safe, it should be more trivial to move to another db engine too
1363 2012-06-18 21:18:24 <TD> it wasn't really hard to do the swap
1364 2012-06-18 21:19:05 <TD> the tedious part would be the conversion code
1365 2012-06-18 21:19:19 <TD> even with a great machine and sig checking disabled, it still takes >10 mins
1366 2012-06-18 21:19:26 <TD> so you need a ui, progress bars, explanation of what's going on, etc
1367 2012-06-18 21:19:40 <sipa> sure, but we'll need that for whatever database change we do
1368 2012-06-18 21:19:49 <sipa> even jeff's database file split needed that
1369 2012-06-18 21:20:14 <luke-jr> TD: >10 mins sounds good compared to jgarzik's split :p
1370 2012-06-18 21:20:51 <sipa> that just did a loadblocks, which means also redoing signature checks
1371 2012-06-18 21:21:51 <luke-jr> sipa: btw, can do you 0.4.7rc2 and 0.5.6rc2 gitian builds?
1372 2012-06-18 21:22:02 JZavala has joined
1373 2012-06-18 21:22:13 <luke-jr> like to make these stable and on SF ASAP since they're the only 0.4 and 0.5 that actually work now :/
1374 2012-06-18 21:22:40 <sipa> jgarzik: i just did a loadblock to 1-file-per-block in 23 minutes
1375 2012-06-18 21:22:47 <sipa> for 172k blocks
1376 2012-06-18 21:22:55 <sipa> previous record was 31 minutes
1377 2012-06-18 21:24:12 <TD> yeah so bdb without sig checking for me was 40+ mins
1378 2012-06-18 21:24:16 <TD> with leveldb, 1
1379 2012-06-18 21:24:17 <TD> 12
1380 2012-06-18 21:24:32 m00p has quit (Ping timeout: 252 seconds)
1381 2012-06-18 21:24:56 <sipa> yes, my "standard" benchmark is only to block 172k, so it doesn't include SD
1382 2012-06-18 21:25:25 sytse has quit (Ping timeout: 272 seconds)
1383 2012-06-18 21:25:35 jav__ has quit (Quit: Verlassend)
1384 2012-06-18 21:27:34 PK has quit ()
1385 2012-06-18 21:28:35 jrmithdobbs has quit (Ping timeout: 272 seconds)
1386 2012-06-18 21:28:35 forrestv has quit (Ping timeout: 272 seconds)
1387 2012-06-18 21:28:46 luke-jr_ has joined
1388 2012-06-18 21:29:05 sirk390 has quit (Quit: Leaving.)
1389 2012-06-18 21:29:09 <sipa> luke-jr: currently too busy doing benchmarks
1390 2012-06-18 21:29:13 luke-jr has quit (Ping timeout: 272 seconds)
1391 2012-06-18 21:29:14 dub has quit (Ping timeout: 272 seconds)
1392 2012-06-18 21:29:14 darkski3z has quit (Ping timeout: 272 seconds)
1393 2012-06-18 21:29:29 <sipa> i can let it run overnight though
1394 2012-06-18 21:30:28 chrisb__ has quit (Quit: Leaving)
1395 2012-06-18 21:31:31 agricocb has quit (Quit: Leaving.)
1396 2012-06-18 21:32:04 <sipa> deleting the blockchain takes ages though, when it's split in 186000 files
1397 2012-06-18 21:32:10 <luke-jr_> Diapolo: left comments for sign/verify
1398 2012-06-18 21:32:12 <luke-jr_> sipa: wtf? loadblocks takes hours here
1399 2012-06-18 21:33:00 <sipa> luke-jr_: what hardware?
1400 2012-06-18 21:33:31 <luke-jr_> sipa: i5-2400
1401 2012-06-18 21:33:40 <sipa> RAM / disk?
1402 2012-06-18 21:33:51 <luke-jr_> 16 GB Mushkin RAM
1403 2012-06-18 21:33:59 <sipa> :o
1404 2012-06-18 21:34:08 jrmithdobbs has joined
1405 2012-06-18 21:34:32 <luke-jr_> Device Model: WDC WD20EARS-00MVWB0
1406 2012-06-18 21:34:34 <gmaxwell> luke-jr_: soudns broken.
1407 2012-06-18 21:34:36 <Diapolo> luke-jr: wil check tomorrow :) thx
1408 2012-06-18 21:34:41 <Diapolo> bye bye
1409 2012-06-18 21:34:43 Diapolo has left ()
1410 2012-06-18 21:34:47 <luke-jr_> gmaxwell: ?
1411 2012-06-18 21:34:54 <sipa> seriously... deleting the blockchain seems to take almost as long as building it :S
1412 2012-06-18 21:35:06 darkski3z has joined
1413 2012-06-18 21:35:11 * sipa blames his userspace-encrypted filesystem
1414 2012-06-18 21:35:30 <gmaxwell> luke-jr_: loadblocks takes like an hour on a spining disk system for me.
1415 2012-06-18 21:35:56 <luke-jr_> I *think* this HD is 5400 RPM, but Google doesn't seem to confirm :/
1416 2012-06-18 21:36:19 <luke-jr_> hmm, maybe because I'm going from/to the same disk?
1417 2012-06-18 21:36:35 <sipa> luke-jr_: i'm also doing it from/to the same disk
1418 2012-06-18 21:37:18 <luke-jr_> hmm
1419 2012-06-18 21:37:31 <luke-jr_> I'm 32-bit if that matters
1420 2012-06-18 21:37:32 <Diablo-D3> not that RPM speeds mean anything anymore
1421 2012-06-18 21:37:42 <JFK911> means a whole lot
1422 2012-06-18 21:37:43 <luke-jr_> Diablo-D3: yeah, I noticedâ¦
1423 2012-06-18 21:38:01 <luke-jr_> JFK911: pretty sure my new 5400 RPM disks outperform my older 7200 RPM
1424 2012-06-18 21:38:04 <Diablo-D3> your drive sucks, btw
1425 2012-06-18 21:38:10 <JFK911> luke-jr_: not for rotational latency
1426 2012-06-18 21:38:22 <Diablo-D3> western digital caviar greens are shit
1427 2012-06-18 21:38:22 <luke-jr_> Diablo-D3: yeah, I was going to use it for backupsâ¦
1428 2012-06-18 21:38:28 <Diablo-D3> luke-jr_: thats even worse
1429 2012-06-18 21:38:30 <JFK911> thats the biggest problem with spindles now
1430 2012-06-18 21:38:32 <luke-jr_> but then it benchmarked better -.-
1431 2012-06-18 21:38:52 <Diablo-D3> yes, but WD greens _fail more_
1432 2012-06-18 21:38:56 <Diablo-D3> its a known defect in the design
1433 2012-06-18 21:39:02 <JFK911> interesting
1434 2012-06-18 21:39:04 <JFK911> i have a WD10EACS
1435 2012-06-18 21:39:05 <Diablo-D3> they obsessively park heads and it wears the drives out
1436 2012-06-18 21:39:08 <JFK911> and its very old
1437 2012-06-18 21:39:19 <Diablo-D3> the exact identical caviar black, save physical drive, doesnt do it
1438 2012-06-18 21:39:20 <JFK911> its outlasted my black WD5000AAKS disks
1439 2012-06-18 21:39:28 <JFK911> Diablo-D3: they aren't identical.
1440 2012-06-18 21:39:37 <Diablo-D3> JFK911: the 2TB ones are
1441 2012-06-18 21:39:40 <Diablo-D3> WD makes exactly one
1442 2012-06-18 21:39:51 <Diablo-D3> the difference is testing and the firmware
1443 2012-06-18 21:40:00 coblee_ has joined
1444 2012-06-18 21:40:07 <JFK911> there are a bunch of different motors and head stacks
1445 2012-06-18 21:40:09 <Diablo-D3> the enterprise ones pass stricter testing than the blacks than the greens
1446 2012-06-18 21:40:16 <Diablo-D3> JFK911: yes, but not for the 2TB
1447 2012-06-18 21:40:16 <JFK911> had you done some research into the DCM labels on the WD drives
1448 2012-06-18 21:40:19 <JFK911> maybe youd notice that
1449 2012-06-18 21:40:35 <luke-jr_> Diablo-D3: I "fixed" the obsessive parking
1450 2012-06-18 21:40:45 <JFK911> you'll find that every character in the DCM represents one major assembly
1451 2012-06-18 21:40:52 <Diablo-D3> luke-jr_: how?
1452 2012-06-18 21:40:58 <Diablo-D3> JFK911: yes, and WD lies out their ass
1453 2012-06-18 21:41:02 <Diablo-D3> I dont trust a thing they say
1454 2012-06-18 21:41:06 <Diablo-D3> they are on my shitlist forever.
1455 2012-06-18 21:41:23 <Diablo-D3> luke-jr_: and btw, I think yours is 7200rpm
1456 2012-06-18 21:41:26 <luke-jr_> hdparm -S 242 /dev/hd2011a
1457 2012-06-18 21:41:27 <JFK911> Diablo-D3: sure, they lie to their own manufacturing and engineering groups
1458 2012-06-18 21:41:40 <JFK911> Diablo-D3: and that's why data recovery guys hunt for compatible parts solely by the DCM mark on the label
1459 2012-06-18 21:41:41 <Diablo-D3> luke-jr_: I would never buy a 7200rpm drive
1460 2012-06-18 21:41:46 <Diablo-D3> not for 2TB anyways
1461 2012-06-18 21:41:50 <Diablo-D3> four platter drives tend to fry
1462 2012-06-18 21:41:56 <luke-jr_> Diablo-D3: hdparm -S 242 /dev/hd2011a
1463 2012-06-18 21:42:03 <Diablo-D3> luke-jr_: that wont fix it
1464 2012-06-18 21:42:06 <Diablo-D3> people have tried
1465 2012-06-18 21:42:09 <JFK911> how many platters are my 7K1000?
1466 2012-06-18 21:42:11 <JFK911> they are heavy
1467 2012-06-18 21:42:16 <JFK911> and old revs (not .B .C)
1468 2012-06-18 21:42:19 <Diablo-D3> JFK911: depends on the capacity
1469 2012-06-18 21:42:24 <JFK911> 1tb
1470 2012-06-18 21:42:28 <luke-jr_> Diablo-D3: well, if the difference is firmware, where's the one that's better? :P
1471 2012-06-18 21:42:29 <JFK911> old drives, still working
1472 2012-06-18 21:42:38 <JFK911> im sure they are at least three
1473 2012-06-18 21:42:40 <JFK911> possibly five
1474 2012-06-18 21:42:44 <Diablo-D3> luke-jr_: its locked to the hardware, you cant flash it with a black
1475 2012-06-18 21:42:51 <luke-jr_> lame
1476 2012-06-18 21:43:01 <Diablo-D3> WD wants your money, and doesnt want to provide customer service to get it
1477 2012-06-18 21:43:02 <JFK911> but.. tha hardware is the same...
1478 2012-06-18 21:43:08 coblee has quit (Ping timeout: 244 seconds)
1479 2012-06-18 21:43:08 coblee_ is now known as coblee
1480 2012-06-18 21:43:10 <JFK911> why "cant you copy the flash"
1481 2012-06-18 21:43:11 <Diablo-D3> JFK911: yup, thats the hilarity
1482 2012-06-18 21:43:26 <Diablo-D3> their flash tool probably checks what the serial number is
1483 2012-06-18 21:43:29 <JFK911> what im laughing at is your level of understanding of disk spindles
1484 2012-06-18 21:43:40 <Diablo-D3> JFK911: yes, we know you're a troll
1485 2012-06-18 21:43:50 <sipa> haha
1486 2012-06-18 21:44:13 <Diablo-D3> luke-jr_: at any rate, when I built my backup drive, I went with a 5400rpm 2TB samsung
1487 2012-06-18 21:44:14 <sipa> watching Diablo-D3 furiously defending his point is always funny
1488 2012-06-18 21:44:22 <JFK911> so Diablo-D3
1489 2012-06-18 21:44:28 <JFK911> do you even know where the firmware is stored in a disk
1490 2012-06-18 21:44:40 <JFK911> whats the name for it?
1491 2012-06-18 21:44:48 <Diablo-D3> luke-jr_: its 4 platter, but 5400rpm 3.5" 4 platters dont seem to fry
1492 2012-06-18 21:44:57 <JFK911> and can you explain how you might read or write it?
1493 2012-06-18 21:45:29 <JFK911> or maybe some other information that is stored along with the firmware?
1494 2012-06-18 21:45:30 <BlueMattBot> Project Bitcoin-Test build #315: STILL FAILING in 30 min: http://jenkins.bluematt.me/job/Bitcoin-Test/315/
1495 2012-06-18 21:45:31 dub has joined
1496 2012-06-18 21:45:54 <JFK911> these are generic spindle disk questions
1497 2012-06-18 21:46:29 <Diablo-D3> luke-jr_: they have 2TB SSDs btw ;) ;) ;) ;) ;) ;) ;)
1498 2012-06-18 21:48:16 <JFK911> anyway if you are interested in disks i can help you move from assumptions based on observations of a technically-interested-but-not-quite-engineering layperson, to actual facts
1499 2012-06-18 21:48:31 <JFK911> you might even repair one or two also
1500 2012-06-18 21:49:14 <JFK911> they are challenging little complexes and lots of fun
1501 2012-06-18 21:53:13 maaku has quit (Quit: maaku)
1502 2012-06-18 21:53:27 <luke-jr_> JFK911: I have a dead WD drive. :/
1503 2012-06-18 21:54:03 maaku has joined
1504 2012-06-18 21:54:20 <JFK911> the thing about repairing disks is its big money if you're good
1505 2012-06-18 21:54:29 <Diablo-D3> luke-jr_: LOL wow
1506 2012-06-18 21:54:30 D34TH has joined
1507 2012-06-18 21:54:34 <JFK911> but it's difficult to learn without a pile of broken modern stuff
1508 2012-06-18 21:54:37 <Diablo-D3> what did I say WDs do? I said they die.
1509 2012-06-18 21:54:47 <JFK911> all disks die
1510 2012-06-18 21:54:57 <JFK911> whats this, a broken ford on the side of the road
1511 2012-06-18 21:54:59 <JFK911> all fords suck!!
1512 2012-06-18 21:55:04 <luke-jr_> lol
1513 2012-06-18 21:55:06 da2ce7 has quit (Ping timeout: 245 seconds)
1514 2012-06-18 21:55:16 <luke-jr_> I'd really like to get at some data on that drive btw
1515 2012-06-18 21:55:18 <JFK911> dont matter that they sell more cars than anyone else
1516 2012-06-18 21:55:34 <JFK911> well, don't practice DR on valuable parts
1517 2012-06-18 21:55:36 <Diablo-D3> luke-jr_: there are companies that specialize in data recovery
1518 2012-06-18 21:55:43 <Diablo-D3> its expensive
1519 2012-06-18 21:55:47 <Diablo-D3> but it teaches you a lesson
1520 2012-06-18 21:55:52 <Diablo-D3> multiple backups in multiple locations
1521 2012-06-18 21:55:54 <JFK911> if you really want it, probably your best bet is to pony up $1000 and send it away
1522 2012-06-18 21:56:05 <luke-jr_> Diablo-D3: any that guarantee privacy even in face of laws requiring mandatory reporting?
1523 2012-06-18 21:56:19 Raziel_ has quit (Quit: Leaving)
1524 2012-06-18 21:56:24 <JFK911> what has mandatory reporting besides child porn
1525 2012-06-18 21:56:45 <luke-jr_> a convicted child pornographer used that PC at one point for about a week
1526 2012-06-18 21:56:53 <Diablo-D3> luke-jr_: they'll turn you in for your child porn, yes
1527 2012-06-18 21:56:55 <luke-jr_> I'm not taking that chance
1528 2012-06-18 21:57:01 <JFK911> then put the disk on ebay
1529 2012-06-18 21:57:21 <JFK911> lol why are you hanging around kidporno people
1530 2012-06-18 21:57:30 <luke-jr_> JFK911: that was before he was convicted obviously
1531 2012-06-18 21:57:33 <luke-jr_> or even suspected
1532 2012-06-18 21:57:39 <Diablo-D3> luke-jr_: seriously, you knew better not to let your priest use your box
1533 2012-06-18 21:57:41 <JFK911> did you tell him the church doesnt abuse people
1534 2012-06-18 21:57:56 <luke-jr_> â¦
1535 2012-06-18 21:58:09 <luke-jr_> he's a jew, so your troll attempt fails
1536 2012-06-18 21:58:36 <JFK911> oh so you tried to convert him
1537 2012-06-18 21:58:44 <Diablo-D3> a.... jew?! and you didnt try to sway his faith?!
1538 2012-06-18 21:58:54 sirk390 has joined
1539 2012-06-18 21:59:14 <Diablo-D3> SAY NO TO GOLDMAN SACHS! THE MERREL LYNCH GOLDEN BULL IS A FALSE IDOL!
1540 2012-06-18 21:59:29 <JFK911> but that bull has a shekel medallion around his neck
1541 2012-06-18 21:59:50 <luke-jr_> he claimed he acknowledged Catholicism was the true religion, but "didn't have time" to convert or something stupid
1542 2012-06-18 21:59:57 <Diablo-D3> didnt have the time?!
1543 2012-06-18 21:59:59 <JFK911> haaahahah
1544 2012-06-18 22:00:00 <Diablo-D3> no wonder!
1545 2012-06-18 22:00:10 <Diablo-D3> he was busy fapping to underage boy penis!
1546 2012-06-18 22:00:23 <JFK911> well normally a jew wouldnt consider converting
1547 2012-06-18 22:00:24 agricocb has joined
1548 2012-06-18 22:00:26 <luke-jr_> thankfully he's in jail for life now
1549 2012-06-18 22:00:28 <JFK911> sounds like luke almost convinced him
1550 2012-06-18 22:00:36 <Diablo-D3> its so tiny... zoom in! zoom in! zoom in! zoom in! now reverse that reflection and sharpen!
1551 2012-06-18 22:00:45 maaku has quit (Quit: maaku)
1552 2012-06-18 22:00:54 <Diablo-D3> wait, sorry, got memes crossed.
1553 2012-06-18 22:01:46 <JFK911> So the only disk make I don't have any broken of, is Hitachi
1554 2012-06-18 22:01:54 <JFK911> probably I got good models.
1555 2012-06-18 22:02:08 <JFK911> because some old hitachis commonly broke, the glass platter ones
1556 2012-06-18 22:02:25 <JFK911> and i heard some things about 7k2000 or 7k3000 models
1557 2012-06-18 22:02:40 <JFK911> I have broken WD disks but many in service
1558 2012-06-18 22:02:49 <JFK911> They don't die as reliably as Seagate.
1559 2012-06-18 22:03:00 <JFK911> And I have never had a WD give me corrupt data back like Seagate has.
1560 2012-06-18 22:03:23 <JFK911> Diablo-D3: remember the 90's when every WD broke after some months?
1561 2012-06-18 22:03:29 minimoose has joined
1562 2012-06-18 22:03:51 <Diablo-D3> hitachis are fucking tanks
1563 2012-06-18 22:03:56 <Diablo-D3> but they sold their drive division
1564 2012-06-18 22:04:03 <Diablo-D3> and seagate? owns samsung now too
1565 2012-06-18 22:04:08 <JFK911> well, WD still makes the 7K1000
1566 2012-06-18 22:04:13 <JFK911> I'd buy more of them
1567 2012-06-18 22:04:21 <Diablo-D3> all drives are now made by two companies
1568 2012-06-18 22:04:24 <Diablo-D3> I have problem with this
1569 2012-06-18 22:04:38 <JFK911> What about Toshiba?
1570 2012-06-18 22:04:44 <Diablo-D3> well, Im hoping WD quits making drives and just expands ibm^Whitachi gst
1571 2012-06-18 22:04:45 <xorgate> noobq here, i'm in folder bitcoin/src, why do i have to start bitcoind using './bitcoind' instead of 'bitcoind' ?
1572 2012-06-18 22:04:52 <Diablo-D3> toshiba is ultrashit
1573 2012-06-18 22:05:02 <Diablo-D3> and they dont make any enterprise drives
1574 2012-06-18 22:05:03 <JFK911> Maybe Toshiba got the shittiest WD factories
1575 2012-06-18 22:05:04 dvide has quit ()
1576 2012-06-18 22:05:10 <Diablo-D3> xorgate: because that is how sh works
1577 2012-06-18 22:05:15 <Diablo-D3> xorgate: has nothing to do with bitcoin
1578 2012-06-18 22:05:30 <Diablo-D3> xorgate: sh, by default, does not have your current dir in your path for security reasons
1579 2012-06-18 22:05:31 <xorgate> ok
1580 2012-06-18 22:06:02 minimoose has quit (Client Quit)
1581 2012-06-18 22:06:04 <Diablo-D3> and the best part? that sh behavior is older than the oldest person in here
1582 2012-06-18 22:06:18 <JFK911> actually i'm older than sh
1583 2012-06-18 22:06:21 <xorgate> i'm just a big linux noob im afraid
1584 2012-06-18 22:06:22 <JFK911> and vi
1585 2012-06-18 22:06:28 <JFK911> evven ed
1586 2012-06-18 22:06:31 <Diablo-D3> JFK911: er, how old are you?
1587 2012-06-18 22:06:49 <JFK911> old enough that i worked for a computer manufacturer on the east coast
1588 2012-06-18 22:06:55 <xorgate> he stems from before jan 1 1970
1589 2012-06-18 22:07:08 <Diablo-D3> xorgate: nothing existed before the epoch damnit
1590 2012-06-18 22:07:16 ovidiusoft has quit (Ping timeout: 264 seconds)
1591 2012-06-18 22:07:29 <JFK911> do you remember when all computers came from NY and MA?
1592 2012-06-18 22:07:40 <Diablo-D3> before the silicon valley? Im aware of it
1593 2012-06-18 22:07:53 <sipa> gmaxwell: i never expected this... switching to separate files per block reduced my 172k benchmark from 31 minutes to 19 minutes...
1594 2012-06-18 22:07:59 forrestv has joined
1595 2012-06-18 22:07:59 <Diablo-D3> ibm's shit used to be all in upper state ny or some shit
1596 2012-06-18 22:08:15 <JFK911> Diablo-D3: not just ibm
1597 2012-06-18 22:08:23 <JFK911> Boston used to be the capital
1598 2012-06-18 22:08:27 <JFK911> up thru the late 80s
1599 2012-06-18 22:09:07 <JFK911> think dec, prime
1600 2012-06-18 22:09:08 <JFK911> thinking machines
1601 2012-06-18 22:09:09 <JFK911> etc etc
1602 2012-06-18 22:09:13 <Diablo-D3> yeah old fucking school
1603 2012-06-18 22:09:18 <Diablo-D3> infact, older than old school
1604 2012-06-18 22:09:28 <Diablo-D3> they rode dinosuars to work to build ... dinosaurs.
1605 2012-06-18 22:09:32 * Diablo-D3 shoots self for pun
1606 2012-06-18 22:09:50 <Diablo-D3> JFK911: you know, I wonder if n0n00dz4u committed suicide
1607 2012-06-18 22:09:53 <JFK911> AT&T too, in NJ.
1608 2012-06-18 22:10:28 <JFK911> Silicon Valley has always been behind
1609 2012-06-18 22:10:50 <JFK911> If we didn't have California, we'd have thrown out text terminals and microcoded CPU sometime during the 80's
1610 2012-06-18 22:11:04 <JFK911> now look what we have today. hypervisors and form/field interfaces
1611 2012-06-18 22:11:16 <Diablo-D3> fucking steve jobs!
1612 2012-06-18 22:11:19 <JFK911> all east coast concepts.
1613 2012-06-18 22:11:23 jine has quit (Ping timeout: 260 seconds)
1614 2012-06-18 22:11:30 jine has joined
1615 2012-06-18 22:11:32 <JFK911> ibm came up with all that stuff.
1616 2012-06-18 22:11:42 <Diablo-D3> dude
1617 2012-06-18 22:11:48 <Diablo-D3> ibm invented.... everything
1618 2012-06-18 22:11:50 <Diablo-D3> in the 60s
1619 2012-06-18 22:12:02 <Diablo-D3> I actually cant think of anything novel that didnt come from them
1620 2012-06-18 22:12:15 <Diablo-D3> atomic instructions? ibm in the 60s.
1621 2012-06-18 22:12:23 <Diablo-D3> s or htm? ibm 60s.
1622 2012-06-18 22:12:26 <JFK911> the way your web browser is used today, is so much like a 3270 or other field/screen terminal.
1623 2012-06-18 22:12:48 <Diablo-D3> threads? pipes? everything is a file? ibm
1624 2012-06-18 22:13:09 <JFK911> real easy to use a virtual punch card with your old apps
1625 2012-06-18 22:13:09 <Diablo-D3> (threads was a fucking stupid idea though)
1626 2012-06-18 22:13:10 <JFK911> ;)
1627 2012-06-18 22:13:31 <Diablo-D3> shipping fully loaded machines to customers and remotely enabling features to upgrade
1628 2012-06-18 22:13:34 <Diablo-D3> that was really clever
1629 2012-06-18 22:13:36 <Diablo-D3> best thing ibm did
1630 2012-06-18 22:13:46 <Diablo-D3> it was essentially a computation vending machine
1631 2012-06-18 22:13:58 <Diablo-D3> some guy had to swing by periodically to reload it as redundant parts failed
1632 2012-06-18 22:17:58 wizkid057 has quit (Read error: Operation timed out)
1633 2012-06-18 22:18:41 wizkid057 has joined
1634 2012-06-18 22:20:40 <sipa> TD: cool, protobuf also uses base-128 encoded numbers
1635 2012-06-18 22:20:53 <TD> it uses varints, yes
1636 2012-06-18 22:20:58 <TD> google runs off varints
1637 2012-06-18 22:21:13 <sipa> though there is a trivial optimization possible, that it doesn't seem to do :)
1638 2012-06-18 22:21:23 <TheSeven> https://theseven.bounceme.net:8338/ user: webuser, password: webpass
1639 2012-06-18 22:21:26 RainbowDashh has quit (Quit: Computer has gone to sleep. DERPY HOOVES SLAMMING THE LAPTOP LID MISTAKE?)
1640 2012-06-18 22:21:29 <TheSeven> almost finished, just sendmany missing
1641 2012-06-18 22:21:38 <TheSeven> and txn fees, I wonder how I should handle those
1642 2012-06-18 22:21:39 <TD> in the code or the format?
1643 2012-06-18 22:21:41 <sipa> for all but the first byte, you know it will never be 0
1644 2012-06-18 22:21:53 <sipa> so you can use them to encode 1-128 instead of 0-127
1645 2012-06-18 22:22:37 <TD> how do you know it won't be zero?
1646 2012-06-18 22:22:54 <sipa> because in that case, the previous byte wouldn't have had its high bit set
1647 2012-06-18 22:23:39 <TD> oh, i think i see. it's late ....
1648 2012-06-18 22:23:47 <TD> it may have been done that way to be more efficient on old processors
1649 2012-06-18 22:23:58 maaku has joined
1650 2012-06-18 22:24:00 <TD> this format was designed years ago. maybe some quirk of old pentiums or something
1651 2012-06-18 22:24:06 <TD> now it can't be changed ....
1652 2012-06-18 22:24:10 <sipa> maybe; this way it only need bitshifts and no addition/multiplications
1653 2012-06-18 22:24:55 RainbowDashh has joined
1654 2012-06-18 22:25:28 rdponticelli has quit (Ping timeout: 276 seconds)
1655 2012-06-18 22:26:17 MC1984 has quit (Quit: Leaving)
1656 2012-06-18 22:26:18 <TheSeven> anyone here interested in my js bitcoin ui thing?
1657 2012-06-18 22:26:22 <TD> i think i read that most of googles cpu time is spent on varint decoding
1658 2012-06-18 22:26:28 <sipa> haha :D
1659 2012-06-18 22:26:34 <TD> so trading off space vs cpu isn't necessarily a good choice
1660 2012-06-18 22:26:38 <TheSeven> someone willing to do some basic security auditing on the server code?
1661 2012-06-18 22:30:23 <MysteryBanshee> Im sure this has been asked before but...
1662 2012-06-18 22:30:29 <MysteryBanshee> Are multisig transactions implemented yet?
1663 2012-06-18 22:30:35 <sipa> yes and no
1664 2012-06-18 22:30:51 <sipa> the network supports them and will verify them
1665 2012-06-18 22:31:02 <sipa> addresses exist to represent multisig destinations
1666 2012-06-18 22:31:26 <sipa> and clients can spend coins sent to multisig address if they have *all* relevant keys
1667 2012-06-18 22:31:36 RainbowDashh has quit (Ping timeout: 256 seconds)
1668 2012-06-18 22:32:00 <MysteryBanshee> oh
1669 2012-06-18 22:32:03 <sipa> the big missing thing (though it's being worked on): a scheme for agreeing on a transaction between multiple parties that share keys
1670 2012-06-18 22:32:16 helo has quit (Ping timeout: 252 seconds)
1671 2012-06-18 22:33:52 helo has joined
1672 2012-06-18 22:34:23 TD has quit (Quit: TD)
1673 2012-06-18 22:34:33 maqr has joined
1674 2012-06-18 22:38:05 <TheSeven> someone want some TN2 coins?
1675 2012-06-18 22:38:20 <TheSeven> or should I just throw them at the faucet?
1676 2012-06-18 22:38:35 Turingi has quit (Quit: Leaving)
1677 2012-06-18 22:42:17 RainbowDashh has joined
1678 2012-06-18 22:46:18 <xorgate> grep
1679 2012-06-18 22:46:30 <xorgate> man i should really check which app has focus
1680 2012-06-18 22:50:23 <sipa> :)
1681 2012-06-18 22:51:14 <TuxBlackEdo> who uses grep with no arguments anyways?
1682 2012-06-18 22:53:18 Ummon_ has joined
1683 2012-06-18 22:54:24 <xorgate> the noob who needs to know what the arguments are :)
1684 2012-06-18 22:54:47 copumpkin has quit (Quit: Computer has gone to sleep.)
1685 2012-06-18 22:56:01 <xorgate> so im browsing around the source.. i notice a lot of code like '5 * 60' to check for 5 minutes, would it not make sense to do things like that in constants
1686 2012-06-18 22:56:11 JZavala has quit (Ping timeout: 240 seconds)
1687 2012-06-18 22:57:15 <TheSeven> hm, bitcoind sending coins from an address to itself? http://blockexplorer.com/testnet/address/n4cXKDD8CdQxbvn2AKcUM43S8gqSqX5fZ7
1688 2012-06-18 22:59:26 RainbowDashh has quit (Quit: Computer has gone to sleep. DERPY HOOVES SLAMMING THE LAPTOP LID MISTAKE?)
1689 2012-06-18 23:00:48 da2ce7 has joined
1690 2012-06-18 23:01:09 RainbowDashh has joined
1691 2012-06-18 23:04:59 <luke-jr_> xorgate: you mean 300 or FIVE_MINUTES ?
1692 2012-06-18 23:05:41 <gmaxwell> TheSeven: you mean "random bitcoin-testnet user sending coins to his own addresses"?
1693 2012-06-18 23:05:44 <xorgate> luke-jr_ no i mean 60 * WHATEVER_IT_IS_IN_MINUTES
1694 2012-06-18 23:06:01 <luke-jr_> xorgate: 60 * FIVE ? O.o
1695 2012-06-18 23:06:38 <xorgate> grrr the example is that it checks if peers have a time different than my local time, the limit is 5 minutes
1696 2012-06-18 23:06:52 <luke-jr_> ok?
1697 2012-06-18 23:08:25 <xorgate> so it would be nicer if that was not just '5' somewhere in the file, but a constant at the top named something like 'MAX_PEER_TIME_OFFSET' or whatever
1698 2012-06-18 23:08:31 att has joined
1699 2012-06-18 23:09:00 <luke-jr_> maybe/.
1700 2012-06-18 23:09:18 <gmaxwell> xorgate: that just means you need to go hunt around to resolve the macro while reading the code.. it makes sense where its used in multiple places and needs to be consistent.
1701 2012-06-18 23:09:35 <TheSeven> gmaxwell: still doesn't make much sense to include them in the txn at all
1702 2012-06-18 23:10:24 eoss has joined
1703 2012-06-18 23:10:24 eoss has quit (Changing host)
1704 2012-06-18 23:10:24 eoss has joined
1705 2012-06-18 23:11:47 <xorgate> gmaxwell maybe in the future it is used elsewhere as well, then someone's gonna have to hunt down the places where this '5' is used. It seems like good form to me to not put constant values like that in code somewhere
1706 2012-06-18 23:12:28 <gmaxwell> xorgate: ... if its used elsewhere then it should potentially be converted into a macro then..
1707 2012-06-18 23:12:34 <TheSeven> is it true that there are plans for a web frontend shipped with bitcoin-qt?
1708 2012-06-18 23:12:47 <TheSeven> i just heard some rumors...
1709 2012-06-18 23:13:49 <gmaxwell> TheSeven: Have you confused this with #apple? Bitcoin development isn't generally some secretive org with leaks where you here "rumors..." about new features.
1710 2012-06-18 23:14:08 <gmaxwell> Can you repeat exactly what you heard? Becaues I have no clue what you're talking about.
1711 2012-06-18 23:14:24 <vragnaroda> gmaxwell: *hear :p
1712 2012-06-18 23:14:43 <gmaxwell> vragnaroda: hey, I got the past tense form right.
1713 2012-06-18 23:14:50 <vragnaroda> :)
1714 2012-06-18 23:14:57 <gmaxwell> "spread spectrum spelling"
1715 2012-06-18 23:16:02 copumpkin has joined
1716 2012-06-18 23:17:35 sirk390 has quit (Quit: Leaving.)
1717 2012-06-18 23:20:05 hnz has joined
1718 2012-06-18 23:20:57 hnz_ has quit (Ping timeout: 245 seconds)
1719 2012-06-18 23:25:55 erle- has quit (Quit: erle-)
1720 2012-06-18 23:28:57 AntKinGTube has joined
1721 2012-06-18 23:29:01 <sipa> gmaxwell: may we should start spreading rumours for planned features, and see what people talk about before starting to implement them :)
1722 2012-06-18 23:29:19 <gmaxwell> haha
1723 2012-06-18 23:30:31 <luke-jr_> O.o
1724 2012-06-18 23:31:06 <luke-jr_> how about a CGMiner/BFGMiner API frontend tab? <.<
1725 2012-06-18 23:31:24 <sipa> or be very vague, and watch the forum people argue how the hell we'd pull something as amazing as that rumour... maybe we'll get very nice ideas
1726 2012-06-18 23:32:00 t7 has quit (Remote host closed the connection)
1727 2012-06-18 23:32:02 <gmaxwell> haha. sadly, I think ideas are not something we're short of... there are a lot of neat ideas. What we need is a machine to add hours to the day.
1728 2012-06-18 23:32:15 <sipa> "bitcoin 0.8 will support USD"
1729 2012-06-18 23:32:22 <gmaxwell> At least thats what I need, dunno about the rest of you.
1730 2012-06-18 23:32:23 <gmaxwell> hahahahahhaa
1731 2012-06-18 23:32:27 brwyatt is now known as Away!~brwyatt@brwyatt.net|brwyatt
1732 2012-06-18 23:32:43 <jgarzik> more hours in the day, and more testers of my personal hacks
1733 2012-06-18 23:32:45 <sipa> "can you send me some testnet USD?"
1734 2012-06-18 23:33:21 <luke-jr_> I need more people pulling <.<
1735 2012-06-18 23:33:28 <luke-jr_> and gitian builders
1736 2012-06-18 23:33:43 <gmaxwell> Should have named testnet USD.
1737 2012-06-18 23:34:26 <sipa> "0.7 will ship USD3 support"
1738 2012-06-18 23:34:31 <gmaxwell> luke-jr_: I did a whole new ubuntu install to get gitian working .. and still no joy. I don't know why it worked out of the box for anyone else.
1739 2012-06-18 23:34:54 <luke-jr_> sipa: I bet the first 90% of replies will think it says USB3
1740 2012-06-18 23:35:03 <sipa> gmaxwell: have you seen my issue i reported on devrandom/gitian repo?
1741 2012-06-18 23:35:16 <luke-jr_> gmaxwell: maybe try an older commit ;)
1742 2012-06-18 23:35:18 <sipa> gmaxwell: i had to change some line in vmbuilder code
1743 2012-06-18 23:35:23 <gmaxwell> No. Was it related to stupidity in selecting the loopback device?
1744 2012-06-18 23:35:39 <sipa> something about umounting the lo device
1745 2012-06-18 23:35:46 <gmaxwell> That sounds like it.
1746 2012-06-18 23:35:50 * luke-jr_ still uses KVM-based gitian, even though he pestered devrandom for LXCâ¦
1747 2012-06-18 23:36:06 * luke-jr_ grubles at whoever broke the next-test build
1748 2012-06-18 23:36:24 <gmaxwell> luke-jr_: Yea, I'm trying the LXC. What I'd like to do is have jenkins doing gitian builds of every commit.
1749 2012-06-18 23:37:10 <luke-jr_> gmaxwell: too bad GitHub is proprietary; we could run our own and modify it to have binaries of every pullreq for testersâ¦
1750 2012-06-18 23:37:43 <gmaxwell> thats a good idea, it wouldn't be tooo hard to script that. But we couldn't get the links onto he github site.
1751 2012-06-18 23:38:36 <gmaxwell> I .. might also be a bit inclined to make those builds handicapped to be testnet only. :)
1752 2012-06-18 23:38:43 MC1984 has joined
1753 2012-06-18 23:38:48 <luke-jr_> gmaxwell: yeah, that might be difficult ;)
1754 2012-06-18 23:38:56 <sipa> LXC didn't work for me, last time i tried
1755 2012-06-18 23:40:40 <AntKinGTube> anybody have a copy of a compiled version of the litecoin miner conf script I can borrow?
1756 2012-06-18 23:40:44 <luke-jr_> UML is probably the best solution?
1757 2012-06-18 23:40:44 <gmaxwell> sipa: so I'm confused that your comment about file per block... I'm confused that it wasn't followed up by a comment saying you were wrong and it wasn't really faster.
1758 2012-06-18 23:40:55 <luke-jr_> lol
1759 2012-06-18 23:41:09 <luke-jr_> AntKinGTube: this is #bitcoin-dev, not #scamcoin
1760 2012-06-18 23:41:15 <gmaxwell> sipa: what all did you do? Is it also eliminating bdb index lookups because the block hash is embodied in the file name?
1761 2012-06-18 23:41:25 <AntKinGTube> yeah yeah yeah
1762 2012-06-18 23:41:29 <gmaxwell> AntKinGTube: #litecoin
1763 2012-06-18 23:41:41 <sipa> gmaxwell: i just used one file per block, and different entries in blkindex.dat
1764 2012-06-18 23:41:48 <AntKinGTube> thanks!
1765 2012-06-18 23:41:56 <sipa> gmaxwell: but i suspect my encrypted home may play an important part
1766 2012-06-18 23:42:04 <gmaxwell> sipa: maybe FS readhead was doing dumb things that made life harder?
1767 2012-06-18 23:42:06 genjix has joined
1768 2012-06-18 23:42:19 <gmaxwell> sipa: it _shouldn't_ have, because it's jut a block translator (assuming you're using dmcrypt)
1769 2012-06-18 23:42:30 Ahimoth has joined
1770 2012-06-18 23:42:35 <sipa> gmaxwell: no, that'd be an encrypted block device, and be in kernel space
1771 2012-06-18 23:42:47 <sipa> this is ecrypt something, and in userspace
1772 2012-06-18 23:42:50 <gmaxwell> oh!
1773 2012-06-18 23:42:51 <gmaxwell> perhaps.
1774 2012-06-18 23:43:06 <gmaxwell> although if it does.. thats interesting too. but not interesting for bitcoin.
1775 2012-06-18 23:43:17 <sipa> but yes, i did a full sync to 172k blocks in 18 minutes
1776 2012-06-18 23:43:37 <gmaxwell> With ECDSA disabled, I assume?
1777 2012-06-18 23:43:40 <sipa> no
1778 2012-06-18 23:43:49 <sipa> but the last checkpoint is at 168k
1779 2012-06-18 23:43:50 <TheSeven> gmaxwell: someone said "could be useful. though i read something the core bitcoin-qt will have a web frontend shipped with it by default too"
1780 2012-06-18 23:43:55 <gmaxwell> hm. thats faster than I sync to that I think.
1781 2012-06-18 23:43:59 <TheSeven> so I don't really know much about it either
1782 2012-06-18 23:44:25 <sipa> gmaxwell: it's currently in my "info" branch if you want to test
1783 2012-06-18 23:44:48 <gmaxwell> TheSeven: I think someone was confused there. We have on or vague list of goals seperating the frontend+wallet from the backend so you could run multiple wallets with bitcoind. This would make it easier to create some kind of web front end but I haven't see anyone talk about that.
1784 2012-06-18 23:46:23 <gmaxwell> I don't personally think a webfrontend is very exciting over the ability to run a very thin (wallet only) client app remotely to access your own or someone you trust's bitcoind. But my preferences are weird.
1785 2012-06-18 23:47:15 <sipa> BlueMatt: it seems jenkins builds without USE_IPV6
1786 2012-06-18 23:48:08 <luke-jr_> sipa: how, without #1431?
1787 2012-06-18 23:48:15 luke-jr_ is now known as luke-jr
1788 2012-06-18 23:48:38 <sipa> luke-jr: obviously, as that isn't merged
1789 2012-06-18 23:48:59 <sipa> but jenkins uses its own makefile i think
1790 2012-06-18 23:49:00 <luke-jr> sipa: no, I'm wondering how they could be, since I don't think it's possible to build without IPv6 until that's merged :P
1791 2012-06-18 23:49:03 <luke-jr> oh
1792 2012-06-18 23:49:13 <sipa> ic
1793 2012-06-18 23:52:57 <MC1984> how feasible is it to purhase some sort of small embedded (raspi or whatever) system to have as your own chainserver, then connect your desktop/phone/whatever thru the internet
1794 2012-06-18 23:54:08 <luke-jr> MC1984: not possible at all right now
1795 2012-06-18 23:54:13 <luke-jr> MC1984: maybe ask in a few months
1796 2012-06-18 23:54:18 Prattler has quit (Remote host closed the connection)
1797 2012-06-18 23:54:33 <jgarzik> ?
1798 2012-06-18 23:54:46 jo__ has joined
1799 2012-06-18 23:54:48 <jgarzik> you can run bitcoind on embedded, as long as you have the filesystem storage space
1800 2012-06-18 23:54:48 <luke-jr> MC1984: or maybe see what Electrum offers
1801 2012-06-18 23:55:10 <luke-jr> jgarzik: I think he meant connect Bitcoin-Qt to an off-site blockchain store
1802 2012-06-18 23:55:21 <MC1984> yeah it seems like the future is in chainservers which sucks ass itself
1803 2012-06-18 23:55:26 slush has joined
1804 2012-06-18 23:55:38 <MC1984> but it would be nice to atleast mainatin some decentralisation there
1805 2012-06-18 23:55:57 <MC1984> or we endup witht he facebook of chainservers, or something
1806 2012-06-18 23:56:14 <sipa> what exactly do you understand under chainserver?
1807 2012-06-18 23:56:26 <sipa> does it maintain your wallet as well?
1808 2012-06-18 23:56:36 <MC1984> no
1809 2012-06-18 23:57:12 <MC1984> just serves you the chain to you personal devices
1810 2012-06-18 23:58:22 <sipa> you may want to have a look at electrum then
1811 2012-06-18 23:58:40 <luke-jr> [23:51:15] <luke-jr> MC1984: or maybe see what Electrum offers <-- :p
1812 2012-06-18 23:59:11 <MC1984> yes like electrum but without everyone leeching off his server
1813 2012-06-18 23:59:21 <luke-jr> MC1984: I think the server might be open source too