1 2014-09-26 01:34:44 <aburan28> /join #bitcoin
2 2014-09-26 03:00:03 <arowser> mrebola: you can see get_balance https://github.com/bitcoin-abe/bitcoin-abe/blob/master/Abe/DataStore.py#L3046
3 2014-09-26 04:03:19 <bombsite> hello, is there documentation on using the -testnet and -regtest for development?
4 2014-09-26 04:57:10 <phantomcircuit> bombsite, regtest has a rule which accepts all blocks regardless of PoW
5 2014-09-26 04:57:15 <phantomcircuit> testnet has something similar
6 2014-09-26 04:57:22 <phantomcircuit> for 99% of testing regtest works well
7 2014-09-26 05:43:33 <bombsite> phantomcircuit, is the github issue tracker the main one used by bitcoin?
8 2014-09-26 05:47:42 <phantomcircuit> bombsite, yes
9 2014-09-26 05:48:46 <bombsite> phantomcircuit: alright haha, I'm gonna go through and see if I can find some low prio issues
10 2014-09-26 05:48:59 <bombsite> hopefully I can learn this codebase this way
11 2014-09-26 06:23:35 <bombsite> Luke-Jr: or phantomcircuit ping?
12 2014-09-26 07:04:38 <Luke-Jr> bombsite: ?
13 2014-09-26 07:12:54 <bombsite> I'd like to give improving the makefiles a shot, as per this issue https://github.com/bitcoin/bitcoin/issues/3979 . I'm not that experienced with makefiles, but I'd like to learn.
14 2014-09-26 07:13:40 <bombsite> If those files are necessary for d and qt, but not for cli, could we simply refactor where it's pointed to in the Makefile.am?
15 2014-09-26 07:20:31 <sipa> bombsite: it's not that much a problem of makefiles, but of libraries
16 2014-09-26 07:21:07 <bombsite> sipa: Ah that makes sense. so there's not really much that can be fixed by me...
17 2014-09-26 07:21:14 <sipa> we're grouping the objects together in a few libraries, and then link a few together to produce a binary
18 2014-09-26 07:21:30 <bombsite> I'm trying to find a low prio issue that I can try tackling to learn the code base
19 2014-09-26 07:21:41 <sipa> oh please do!
20 2014-09-26 07:21:49 <bombsite> but most of them seem to be beyond my scope
21 2014-09-26 07:22:37 <sipa> feel free to come here and ask quesrions about the codebase
22 2014-09-26 07:22:48 <bombsite> awesome thanks sipa :)
23 2014-09-26 07:22:51 <sipa> i can't immediately point you at a 'simple',8##73 though
24 2014-09-26 07:23:00 <bombsite> it's actually pretty funny though
25 2014-09-26 07:23:04 <sipa> a 'simple' issue
26 2014-09-26 07:23:05 <Luke-Jr> usually autotools is not a good place for beginners :p
27 2014-09-26 07:23:15 <bombsite> I'm good enough to get hired as a software engineer out of college
28 2014-09-26 07:23:20 <bombsite> but still not good enough to contribute
29 2014-09-26 07:23:23 <bombsite> to OS code
30 2014-09-26 07:23:23 <sipa> haha
31 2014-09-26 07:23:31 <sipa> welcome... to the real world
32 2014-09-26 07:24:51 <bombsite> question for you though
33 2014-09-26 07:25:00 <bombsite> what is your general procedure for finding optimizations?
34 2014-09-26 07:25:24 <bombsite> Do you have a way to profile hotspots? or do you just randomly "oh maybe if i do this here"
35 2014-09-26 07:26:17 <wumpus> the first rule of optimization is to profile first, to find the bottleneck
36 2014-09-26 07:26:54 <wumpus> the other way around is extremely discouraged, even if you're experienced, you'll do a lot of extra work and then when you actually measure you realize your change didn't even affect performance
37 2014-09-26 07:27:35 <bombsite> wumpus: That makes sense, but I've also seen some ridiculous optimizations that include fastpaths that basically require a ridiculous amount of understand of the compiler
38 2014-09-26 07:28:04 <bombsite> which I feel like you basically need a phD or lotsa experience to see it
39 2014-09-26 07:28:08 <wumpus> bombsite: oh, sure, DSP and video decoding code is full of those, as well as inner crypto code
40 2014-09-26 07:28:51 <bombsite> is there a way for me to understand/learn/whatever those a bit better without getting a phD?
41 2014-09-26 07:28:53 <wumpus> but in general, for most people, most of the code you'll write in practice, you don't need those tricks (and you'll avoid them like the plague so that other people understand your code)
42 2014-09-26 07:29:14 <wumpus> it's still good to have a deep understanding of CPUs and compilers of course :)
43 2014-09-26 07:29:54 <bombsite> I should quit my job and go back and get my masters lol
44 2014-09-26 07:30:33 <bombsite> wumpus: what's your method for learning a new codebase?
45 2014-09-26 07:31:16 <wumpus> bombsite: making small changes
46 2014-09-26 07:31:20 <bombsite> Do you draw diagrams? or step through it with gdb? or interesting ways
47 2014-09-26 07:31:26 <bombsite> hmm...
48 2014-09-26 07:32:03 <wumpus> ie, find some simple thing that annoys you or crashes, debug into it, try to fix it
49 2014-09-26 07:32:42 <wumpus> drawing diagrams or stepping with gdb can be part of the process
50 2014-09-26 07:33:58 <bombsite> alright haha.. I think I'm gonna spend my friday and saturday just stepping through the code and trying to understand it
51 2014-09-26 07:34:20 <bombsite> gonna sleep. night :)
52 2014-09-26 07:34:37 <wumpus> goodnight
53 2014-09-26 07:54:33 <chichov> what library is bitcoin using to load config files?
54 2014-09-26 07:54:46 <Luke-Jr> none
55 2014-09-26 07:55:01 <Luke-Jr> unless you mean Bitcoin-Qt
56 2014-09-26 07:55:03 <Luke-Jr> then it's Qt
57 2014-09-26 07:55:24 <Luke-Jr> but that's not used for the .bitcin config files
58 2014-09-26 07:55:27 <Luke-Jr> .bitcoin*
59 2014-09-26 07:59:15 <chichov> yep, qt
60 2014-09-26 07:59:16 <wumpus> libsatoshi
61 2014-09-26 07:59:29 <wumpus> :P
62 2014-09-26 07:59:35 <chichov> heh, no?
63 2014-09-26 07:59:50 <wumpus> the great utility library called util.cpp
64 2014-09-26 08:00:12 <wumpus> ReadConfigFile()
65 2014-09-26 08:00:14 <chichov> alright, thanks
66 2014-09-26 08:00:59 <Luke-Jr> wumpus: it would actually be useful to have that part abstracted - I reimplement it in BFGMiner :p
67 2014-09-26 08:01:13 <chichov> ah yea, a boost config file is used
68 2014-09-26 08:01:22 <timothy> Luke-Jr: I used to maintain bfgminer on archlinux :P
69 2014-09-26 08:01:30 <Luke-Jr> timothy: used to? :o
70 2014-09-26 08:01:31 <wumpus> Luke-Jr: hah, there's also various implementations of it in Python
71 2014-09-26 08:01:46 <timothy> Luke-Jr: I don't mine anymore since mining with GPU is no more profitable :)
72 2014-09-26 08:01:49 <chichov> thanks
73 2014-09-26 08:01:53 <timothy> and it's almost useless
74 2014-09-26 08:02:57 <Luke-Jr> timothy: get some ASICs? :p
75 2014-09-26 08:03:26 <timothy> I accept donations :P also an old usb semi-broken asic :P
76 2014-09-26 10:43:11 <timothy> another reason to pass to cloudflare http://blog.cloudflare.com/keyless-ssl-the-nitty-gritty-technical-details/
77 2014-09-26 10:54:48 <timothy> where is the tar of new version?
78 2014-09-26 11:00:20 <Happzz> are there some criteria under which a transaction is most likely to never be confirmed?
79 2014-09-26 11:00:34 <Happzz> what happens to such a transaction that essentially never confirms?
80 2014-09-26 11:00:51 <Happzz> s/most likely/likely
81 2014-09-26 11:12:11 <wumpus> nodes will forget about them sooner or later
82 2014-09-26 11:15:00 <wumpus> sooner if the transaction is non-standard (it won't even be accepted in the mempool), but if valid otherwise and just not feasible for miners to mine it will be forgotten from mempools as nodes restart
83 2014-09-26 11:27:56 <Happzz> wumpus how do i know if a transaction is that unfeasible
84 2014-09-26 11:28:06 <Happzz> like, what criterias
85 2014-09-26 11:29:27 <wumpus> primarily whether the fee is enough for the size (in kb) of the transaction
86 2014-09-26 11:32:44 <wumpus> if no fee, or too little fee is attached there are some other constraints that determine the priority as a free transaction, such as how old the input coins are
87 2014-09-26 14:04:41 <sea> Hi all, I'm looking for an expert in the bitcoin protocol.
88 2014-09-26 14:04:55 <timothy> like gribble?
89 2014-09-26 14:05:05 <sea> Well, they should ideally be here and not idle. Is gribble here?
90 2014-09-26 14:05:24 <timothy> I'm joking, gribble is a bot
91 2014-09-26 14:05:26 <RagnarDanneskjol> haha
92 2014-09-26 14:06:04 <sea> Hrm. Well anyway, I'd like a human who is reasonably certain that they understand the protocol well. An academic would be best. Are there any academics reading this?
93 2014-09-26 14:06:23 <wumpus> better to just ask your question
94 2014-09-26 14:07:15 <sea> I've developed a new family of attacks against the bitcoin protocol which will require a major overhaul and I'm looking for an independent expert to review my work before it's published.
95 2014-09-26 14:07:29 <sea> I've got a few people already but I want someone from here as well.
96 2014-09-26 14:08:29 <wumpus> ping @gmaxwell
97 2014-09-26 14:13:14 <b-itcoinssg> sea: are you still there?
98 2014-09-26 14:13:18 <sea> I'm here
99 2014-09-26 14:14:07 <b-itcoinssg> I'm a scientist and I review peer reviewed papers, I'd be really interested in reading yours my e-mail is bitcoinssg@gmail.com and my twitter is @bitcoinssg
100 2014-09-26 14:14:14 <sea> Oh good!
101 2014-09-26 14:23:38 <helo> i promise i'm not a malware writer hoping to get my foot in the door before any fixes can be put in place
102 2014-09-26 15:00:59 <willermo> hello folks, i trying to dump all the bitcoin transactions to a DB, but bitcoind return "No information available about transaction" for certain transactions
103 2014-09-26 15:01:12 <willermo> for example this tx 0437cd7f8525ceed2324359c2d0ba26006d92d856a9c20fa0241106ee5a597c9
104 2014-09-26 15:02:00 <jgarzik> willermo, turn on txindex=1 in config, and -reindex
105 2014-09-26 15:02:13 <jgarzik> willermo, spent transactions are not kept in that index otherwise
106 2014-09-26 15:06:04 <willermo> jgarzik: ok thanks i will try
107 2014-09-26 18:26:02 <midnightmagic> sea: you did catch what helo said? Be careful. There are a number of academics who will pre-announce, and attempt to profit from any vulnerabilities you think you've found.
108 2014-09-26 18:29:52 <midnightmagic> doh
109 2014-09-26 19:10:31 <coryfields> maraoz: ping
110 2014-09-26 19:11:55 <coryfields> maraoz: it'd be great if you could help get the deps working in 10.6, i'd be more than happy to step through it with you if you're willing to fight with it for a bit
111 2014-09-26 19:33:42 <maraoz> coryfields: I'd love to, but I gotta go now. maybe next week?
112 2014-09-26 19:33:58 <coryfields> sure
113 2014-09-26 19:39:25 <BlueMatt> whats with the segfaults on master last night?
114 2014-09-26 19:41:15 <sipa> ACTION knows nothing
115 2014-09-26 19:41:31 <BlueMatt> damn, I was hoping someone would say they've been fixed, then I wouldnt have to debug :p
116 2014-09-26 19:47:13 <sipa> what are you even talking about?
117 2014-09-26 19:51:07 <BlueMatt> dunno, built master last night and tried to do a loadblock with gui and it constantly segfaulted
118 2014-09-26 19:51:19 <BlueMatt> 0.9.3 worked, however
119 2014-09-26 19:51:28 <sipa> interesting
120 2014-09-26 19:51:55 <BlueMatt> well, sometimes segfaulted, double-free'd, etc
121 2014-09-26 19:52:41 <sipa> ewww
122 2014-09-26 19:53:17 <BlueMatt> indeed
123 2014-09-26 20:02:12 <coryfields> BlueMatt: tests pass for that commit?
124 2014-09-26 20:02:41 <BlueMatt> coryfields: which commit?
125 2014-09-26 20:03:18 <coryfields> BlueMatt: whatever commit in master you were building
126 2014-09-26 20:03:35 <BlueMatt> didnt have time to try, I was off to bed....
127 2014-09-26 20:03:39 <BlueMatt> will debug a bit later today
128 2014-09-26 20:59:40 <abrkn> BlueMatt: check priv if available
129 2014-09-26 22:33:07 <gmaxwell> Did anyone actually competent see sea's message at all.
130 2014-09-26 22:35:54 <gwillen> gmaxwell: well, sea's still online, just not in here
131 2014-09-26 23:17:41 <lechuga_> anyone actually competent
132 2014-09-26 23:17:57 <lechuga_> good grief