1 2015-12-27 01:17:41 <alpalp> Has the idea of making a more "plug-in" architecture ever been discussed?
 2 2015-12-27 03:06:25 <rusty> aj: moving question: re 9999 satoshi fees here.  I'm not seeing them...
 3 2015-12-27 03:10:08 <btcdrak> ping rusty
 4 2015-12-27 03:10:26 <rusty> btcdrak: pong! Merry End-Of-Year Event!
 5 2015-12-27 03:10:38 <btcdrak> rusty: yes!
 6 2015-12-27 03:10:43 <btcdrak> rusty: please check your email :)
 7 2015-12-27 03:46:17 <rusty> btcdrak: sorted.
 8 2015-12-27 03:46:25 <btcdrak> rusty: thanks!
 9 2015-12-27 03:47:32 <rusty> btcdrak: now I can go back to enjoying my holidays.  By optimizing bitcoin-iterate...
10 2015-12-27 03:47:45 <btcdrak> rusty: yay!
11 2015-12-27 03:56:52 <maaku> rusty: pretty sure 9999 fees is a rounding artifact
12 2015-12-27 03:57:17 <rusty> maaku: aj was saying it happened some huge percentage of the time.  AFAICT it doesn't happen more often than chance.
13 2015-12-27 03:57:48 <rusty> maaku: so, either I misunderstood what he said, or one of one of us is wrong.
14 2015-12-27 03:58:07 <maaku> is he talking about total fee or fee-per-kB?
15 2015-12-27 03:58:25 <maaku> if the latter, he might be doing fp math and truncating somewhere
16 2015-12-27 03:59:02 <maaku> that's what i assumed when I saw it at least, because I haven't seen actual 9999 fees either
17 2015-12-27 04:26:27 <aj> maaku: i've just been letting python add things up and subtracting to get fees, totally plausible that i've got an off-by-one somehow. it's the total fee not the fee-per-kB though
18 2015-12-27 04:46:17 <arioBarzan> would anybody shed some light on the rathionallity behind commit/cf0c47b2 ?
19 2015-12-27 04:52:17 <arioBarzan> Does it mean that nobody is using "getwork" anymore?
20 2015-12-27 04:53:54 <maaku> arioBarzan: 'getwork' hasn't been meaningful for years
21 2015-12-27 04:54:28 <maaku> a modern ASIC completely exhausts the search space for 'getwork' in less time than the RPC takes to process
22 2015-12-27 04:55:06 <maaku> 'getwork' was pretty much only useful for testnet, and that is now obsoleted by making regtest more accessible
23 2015-12-27 05:04:26 <jl2012> maaku: what is benefit of using the base32 character set you suggested (except that some value may happen more frequently a the end of the address, which is not important here since the address is very long)
24 2015-12-27 05:06:04 <maaku> jl2012: it's in the doc, but the charset is chosen to be both visually and audibly distinctive (most base32 alphabets don't consider auditory confusion)
25 2015-12-27 05:07:38 <maaku> and i believe that the ordering is selected to make small differences in data result in more distinctive differences in selected codepoint
26 2015-12-27 05:07:56 <maaku> in addition to the better rules regarding string termination
27 2015-12-27 05:44:27 <arioBarzan> maaku: without getwork, how people get "hash data to work on"?
28 2015-12-27 05:48:18 <arioBarzan> maaku: found the answer.
29 2015-12-27 05:52:28 <arioBarzan> It's just replaced with Getblocktemplate. would have been nice if there was a reference on the commit though.
30 2015-12-27 07:01:51 <frankenmint> if a transaction is 'stuck' is this line (mempoolexpiry) the mechanism that relesases it back to a balance for a given relay node?: https://github.com/bitcoin/bitcoin/blob/master/src/main.h#L60
31 2015-12-27 08:38:30 <jl2012> a new segwit address proposal: https://github.com/bitcoin/bips/pull/267 with python code: https://gist.github.com/jl2012/760b0f952715b8b6c608
32 2015-12-27 08:39:17 <jl2012> I used to use perl and this is my first python script
33 2015-12-27 08:50:29 <Luke-Jr> personally I'd prefer Perl :p
34 2015-12-27 08:54:29 <jl2012> I like perl too but seems more people use python so I just follow the trend
35 2015-12-27 09:13:15 <digitsu> perl is an exercise in masochism.
36 2015-12-27 09:13:39 <digitsu> its fun to write BECAUSE you know that nobody can understand it without reading it 10 times over first.
37 2015-12-27 09:14:24 <digitsu> A well written Perl script always gave me the impression of casting a magic spell
38 2015-12-27 09:14:34 <digitsu> the more regex's the better.
39 2015-12-27 09:15:29 <Luke-Jr> pfft, Perl is nice once you know it
40 2015-12-27 09:16:54 <digitsu> I never said it wasn't *nice*
41 2015-12-27 09:17:18 <digitsu> just that it's appeal is based on its obfuscation and terseness
42 2015-12-27 09:17:39 <digitsu> which is, you will agree, usually the opposite of 'ease of comprehension'
43 2015-12-27 14:01:50 <jl2012> is the hashrate really doubled?
44 2015-12-27 15:00:35 <Luke-Jr> jl2012: it appears no
45 2015-12-27 15:01:22 <jl2012> maybe increased certain % + luck
46 2015-12-27 15:05:15 <JackH> blocksize debate going wild again on reddit
47 2015-12-27 15:06:36 <JackH> its interesting though, to see what a coordination there is between people to attack Bitcoin
48 2015-12-27 15:06:43 <JackH> and who is this Peter R anyway?
49 2015-12-27 15:08:46 <Luke-Jr> JackH: off-topic here
50 2015-12-27 15:09:19 <JackH> whops, I am in wrong chan, didnt notice
51 2015-12-27 17:29:22 <to4dy> can't wait for the state of the onion talk :D
52 2015-12-27 17:30:05 <to4dy> wrong chat sry
53 2015-12-27 18:22:10 <kenrestivo> anything open is easily attackable. the flipside is it can be made resilient against it, but it's not easy.
54 2015-12-27 21:13:03 <andytoshi> is there a way i can get all the blocks, in order, from the reference client
55 2015-12-27 21:13:12 <andytoshi> a lot faster that calling `getblockhash` then `getblock`
56 2015-12-27 21:16:01 <alpalp> andytoshi: last time I did that I just made something to read the blk files.
57 2015-12-27 21:16:20 <arubi> andytoshi, somewhat jokingly, you could get them backwards by using just getblockhash :)
58 2015-12-27 21:19:55 <arubi> er, just one time I mean..
59 2015-12-27 23:25:05 <sipa> TIL: when sighash_none or sighash_single is specified, the nSequence values are not sign