1 2013-07-08 00:36:14 <crypt0queen> anyone interested in going in on the group buy for a Bitcoin neon sign? pics --> [ http://imgur.com/a/Z7VWc ] , 4 BTC including shipping to the USA. need 7 more people. :D
2 2013-07-08 00:52:44 <freewil> crypt0queen, how much
3 2013-07-08 00:52:51 <freewil> oh 4 btc
4 2013-07-08 00:52:58 <crypt0queen> yeah includes shipping to the US
5 2013-07-08 00:53:09 <crypt0queen> will need to adjust if international to your country
6 2013-07-08 00:56:11 <crypt0queen> oh, wait, sorry freewil , 3.5 btc @ 77 currently
7 2013-07-08 00:56:16 <crypt0queen> my mistake
8 2013-07-08 00:56:53 <crypt0queen> includes shipping
9 2013-07-08 00:57:03 <crypt0queen> (usa)
10 2013-07-08 00:57:07 <freewil> alright
11 2013-07-08 00:57:17 <freewil> im not interested but ill ask around
12 2013-07-08 00:57:21 <crypt0queen> cheers
13 2013-07-08 01:01:15 <Luke-Jr> crypt0queen: off-topic here, I suggest #bitcoin-otc
14 2013-07-08 01:01:26 <Luke-Jr> neat signs though
15 2013-07-08 01:01:28 <crypt0queen> yeah I asked there already, =)
16 2013-07-08 01:01:30 <crypt0queen> thanks luke
17 2013-07-08 01:01:39 <crypt0queen> will ask again tomorrow so I dont spam the channel there
18 2013-07-08 05:59:56 <warren> Luke-Jr: I found it curious that you conidered nickserv auth to be a good way to verify my identity. that is pretty darn weak compared to other means.
19 2013-07-08 07:31:25 <TD> sipa: good morning. how fast can libsecp256k1 do a point multiplication?
20 2013-07-08 07:31:32 <TD> (private->public)
21 2013-07-08 07:32:51 <phantomcircuit> TD, iirc it's 1000x faster which would mean 1usec
22 2013-07-08 07:32:55 <phantomcircuit> but that seems fast
23 2013-07-08 07:56:16 <gjs278> I found a block on btc guild and I've only genned 10btc with them so far :(
24 2013-07-08 07:59:13 <michagogo> gjs278: Well, that's the trade-off of mining with a pool
25 2013-07-08 07:59:37 <gjs278> time to go 50 straight and not find anything
26 2013-07-08 07:59:47 <michagogo> gjs278: If you're not lucky, you still get BTC
27 2013-07-08 08:00:04 <michagogo> But if you do get lucky, you don't actuallt get lucky
28 2013-07-08 09:59:38 <sipa> TD: 30us or so
29 2013-07-08 10:00:41 <sipa> TD: last i benchmarked, that was "only" 3 times faster than openssl
30 2013-07-08 10:01:33 <Diablo-D3> "only" ;)
31 2013-07-08 10:02:45 <sipa> TD: though that's only for multiplication with the generator (whose multiples are precomputed heavily), not for generic point multiplications
32 2013-07-08 10:08:51 <TD> ok
33 2013-07-08 10:08:54 <TD> that's fine
34 2013-07-08 10:09:09 <TD> so about 33k priv->pubs per second
35 2013-07-08 10:10:16 <sipa> though the current algorithm isn't entirely constant-time, so it may slow down a bit in the future if i get to that
36 2013-07-08 10:10:43 <sipa> (resistence to timing attacks)
37 2013-07-08 10:10:57 <sipa> also, i should probably talk in terms on cpu cycles instead of time
38 2013-07-08 10:11:22 <sipa> close to 100k, i guess
39 2013-07-08 10:19:37 <TD> 100k on what?
40 2013-07-08 10:19:49 <TD> 1,000,000 / 30 is not 100k :-)
41 2013-07-08 10:19:53 <TD> you mean on a quad core system?
42 2013-07-08 10:21:16 <sipa> 100k CPU cycles at 3.17GHz is close to 30us
43 2013-07-08 10:21:25 <sipa> no, per core
44 2013-07-08 10:21:39 <TD> oh, sorry
45 2013-07-08 10:21:54 <TD> somehow i added an "r" into "close" so i thought you mean the 33k was off
46 2013-07-08 11:42:50 <BlueMatt> hmm... bitcoin-qt -regtest doesnt load unless I specify the datadir as .bitcoin/regtest manually
47 2013-07-08 11:43:50 <BlueMatt> gmaxwell: wow, surprised he's still posting
48 2013-07-08 11:43:59 <petertodd> BlueMatt: pull-tester still broken? I wanted to see what it thought of my cpfp mempool
49 2013-07-08 11:44:22 <BlueMatt> petertodd: ahh, yes, it is, sorry
50 2013-07-08 11:44:25 <BlueMatt> thanks for reminding me
51 2013-07-08 11:45:10 <petertodd> BlueMatt: np, of course, I'll admit I am being a bit lazy in not setting up a local copy myself... that pull-req is nowhere near actually being merged
52 2013-07-08 11:45:20 <petertodd> ACTION knows 'nuthin about java
53 2013-07-08 11:45:29 <BlueMatt> meh, Im more lazy
54 2013-07-08 11:45:40 <BlueMatt> ACTION makes pull-tester build things when he doesnt feel like doing it locally
55 2013-07-08 11:45:43 <BlueMatt> (mostly win32 stuff)
56 2013-07-08 11:46:37 <pigeons> BlueMatt: why suprised, did he get the usual level of discourse from that forum previously?
57 2013-07-08 11:47:12 <BlueMatt> pigeons: http://en.wikipedia.org/wiki/Hal_Finney_(cypherpunk)
58 2013-07-08 11:47:24 <pigeons> oh yes i know who he is
59 2013-07-08 11:48:06 <pigeons> i was wondering why you were suprised, like maybe he was treated like a typical forum member previously or something
60 2013-07-08 11:48:11 <TD> BlueMatt: that's odd. i recall that working when i first wrote the code. but i only normally use it with bitcoind -regtest
61 2013-07-08 11:48:30 <pigeons> he seems genuinely interested in bitcoin, kind of a child of his idea
62 2013-07-08 11:48:33 <TD> pigeons: he is seriously ill
63 2013-07-08 11:48:41 <TD> pigeons: to the extent that he must type using eye movements
64 2013-07-08 11:48:48 <pigeons> wow, amazing
65 2013-07-08 11:48:57 <TD> that's why we're surprised he's still posting. sadly, his life expectancy is very short
66 2013-07-08 11:49:09 <pigeons> ok thanks
67 2013-07-08 11:49:14 <BlueMatt> pigeons: he has ALS
68 2013-07-08 11:50:30 <BlueMatt> also, on master: when I setgenerate true, I get one block and no more until I setgenerate false and reset it....
69 2013-07-08 11:50:39 <BlueMatt> (on regtest mode, so should work....)
70 2013-07-08 11:51:04 <TD> maybe something broke recently. i was making blocks by repeatedly running setgenerate true this morning, but it's been a couple of weeks since i refreshed my local build
71 2013-07-08 11:52:08 <kinlo> seems there are quite some people mining on testnet
72 2013-07-08 11:52:49 <BlueMatt> TD: setgenerate true should make inf blocks, not just one
73 2013-07-08 11:53:04 <BlueMatt> for now setgenerate true is giving me exactly one block per call...
74 2013-07-08 11:53:08 <TD> yes, that's by design
75 2013-07-08 11:53:11 <TD> it was noted in the original pull req
76 2013-07-08 11:53:23 <TD> you can just do "setgenerate true" repeatedly. no need to do "setgenerate false"
77 2013-07-08 11:53:44 <BlueMatt> wait what?
78 2013-07-08 11:53:49 <BlueMatt> oh, this is new for regtest stuff?
79 2013-07-08 11:53:50 <TD> that way you get exactly one block on demand when you want one
80 2013-07-08 11:53:52 <TD> yeah
81 2013-07-08 11:53:57 <BlueMatt> oh, ok
82 2013-07-08 11:53:59 <TD> does the pull tester rely on "infinite blocks"?
83 2013-07-08 11:54:14 <jgarzik> mornin'
84 2013-07-08 11:54:30 <TD> you can do something like "while true; do ./bitcoind -regtest setgenerate true; usleep $[ 300*1000]; done" on my machine to get lotsablocks
85 2013-07-08 11:55:02 <BlueMatt> TD: pull-tester doesnt care
86 2013-07-08 11:55:14 <BlueMatt> all pull-tester blocks are made in bitcoinj
87 2013-07-08 11:55:32 <TD> ok. i figured the one-block-per-call thing made it easier to do app development. at least i found it helpful so far.
88 2013-07-08 11:55:50 <BlueMatt> just surprised me :p