1 2014-08-09 05:38:49 <stapler117> I thought I was going crazy, then gitian worked again. All is well.
2 2014-08-09 09:47:14 <aschildbach> petertodd: Hi Peter, it looks like your testnet seed is down. Could you have a look?
3 2014-08-09 14:52:33 <anotheraltcoin> o/
4 2014-08-09 14:53:26 <anotheraltcoin> I'm trying to make an altcoin, would someone be able to help get the testnet working?
5 2014-08-09 14:55:20 <sipa> not here
6 2014-08-09 14:58:24 <iwilcox> anotheraltcoin: Not #bitcoin either. There's ##altcoin-dev #altcoin ##altcoin
7 2014-08-09 14:59:19 <anotheraltcoin> thank you
8 2014-08-09 15:00:20 <anddam> hi, I remember there was a person in charge of building the OS X application bundle that is shipped in the disk image on bitcoin.org but I cannot recall his name, who is it?
9 2014-08-09 15:01:36 <Diablo-D3> I dont think its luke
10 2014-08-09 15:01:51 <sipa> anddam: gavin built the osx releases until now
11 2014-08-09 15:02:05 <sipa> anddam: though in a future release that will likely be automated
12 2014-08-09 15:06:57 <anddam> thanks
13 2014-08-09 15:08:32 <blazerbit> Hi I'm trying to setup a node on centos. Bitcoind has connections and sync from bootstrap but after bootstrap data is synced I get write erros
14 2014-08-09 15:08:57 <sipa> where do you get those?
15 2014-08-09 15:09:06 <blazerbit> Should I be asking for help here or #bitcoin ?
16 2014-08-09 15:10:31 <anddam> I'm trying to figure why the OS X binary from website is 30 MB while the one built with MacPorts is 7.3 MB
17 2014-08-09 15:10:44 <anddam> can debug symbols account for such a difference?
18 2014-08-09 15:11:06 <sipa> yes, certainly
19 2014-08-09 15:11:26 <sipa> also, isn' t the release binary both 32 and 64 bit?
20 2014-08-09 15:11:39 <sipa> (i'm not up to date with the osx side of things, my info may be outdated)
21 2014-08-09 15:12:00 <anddam> nope, it's x86-64 only
22 2014-08-09 15:12:32 <sipa> ok
23 2014-08-09 15:13:38 <anddam> I mean they both are
24 2014-08-09 15:14:32 <sipa> ok
25 2014-08-09 15:14:40 <sipa> anyway, yes, debug symbols can account for that muchh
26 2014-08-09 15:15:09 <anddam> nice, strip -S -x decreased size from 31489904 to 31489888
27 2014-08-09 15:15:12 <anddam> impressive
28 2014-08-09 15:15:28 <sipa> seems there weren't many debug symbols :)
29 2014-08-09 15:16:05 <anddam> I still see 43k global symbols
30 2014-08-09 15:16:13 <anddam> while the macports built has ~2.5k
31 2014-08-09 15:17:01 <anddam> curiouser and curiouser
32 2014-08-09 15:17:17 <sipa> maybe some flag that exports more symbols by default?
33 2014-08-09 15:17:50 <anddam> I don't know how the website binary was built in first place, hence my initial question
34 2014-08-09 15:17:59 <sipa> maybe cfields knows
35 2014-08-09 15:18:12 <anddam> gavinandresen, cfields: ping
36 2014-08-09 15:18:18 <iwilcox> Maybe the MacPorts build relies on shared Qt, and the website one has it all static?
37 2014-08-09 15:18:26 <fanquake> anddam Is building with macports still fairly easy? last i remember there were either boost or osx 10.6 issues?
38 2014-08-09 15:18:38 <anddam> iwilcox: IIRC you cannot do a static link in darwin
39 2014-08-09 15:19:08 <anddam> anyway the official binary is dynamically linking the Qt framework
40 2014-08-09 15:19:27 <anddam> and of course I'm taking into account just the actual executables, not the whole application bundle
41 2014-08-09 15:19:46 <sipa> maybe some other library? openssl?
42 2014-08-09 15:19:48 <anddam> i.e. Qt frameworks are in their separate folder
43 2014-08-09 15:20:33 <rnicoll> Looking at noui.cpp/noui_ThreadSafeMessageBox() and trying to decide if it's safe to actually prompt the user for input from the command line rather than just returning false... any thoughts? This is in the context that in case of LevelDB corruption, the prompt to reindex is presented but no input taken, which feels odd. I'm aware however that prompting from a daemon running in the background may not be useful
44 2014-08-09 15:21:00 <anddam> I just built the same 0.9.2.1 version using mp, it reaches just 7.9 MB
45 2014-08-09 15:21:38 <anddam> sipa: those are very small anyway, but the official binary has @executable_path/../Frameworks/libssl.1.0.0.dylib
46 2014-08-09 15:22:00 <sipa> rnicoll: prompting doesn't make sense for a daemon
47 2014-08-09 15:22:15 <sipa> rnicoll: the message still being shown is weird
48 2014-08-09 15:22:32 <sipa> people sometimes type 'yes' as a result on the command line
49 2014-08-09 15:22:41 <sipa> resulting in a screen full of y's :)
50 2014-08-09 15:23:09 <rnicoll> sipa, was wondering if adding a "-interactive" command line option might help, but then, I'm still just changing the nature of the problem...
51 2014-08-09 15:23:25 <rnicoll> fixing the broken data is easy, you just need to know the right command line option
52 2014-08-09 15:23:57 <rnicoll> thoughts on having it default to rebuilding if run on the command line, rather than failing?
53 2014-08-09 15:24:56 <sipa> i thyink, as someone running the daemon, you should be made aware of the problem, and not silently fix
54 2014-08-09 15:25:12 <sipa> i expect people complaining that 'starting bitcoind takes hours!'
55 2014-08-09 15:25:24 <sipa> if it somehow fails every time
56 2014-08-09 15:25:26 <rnicoll> Change message to "something something run with -reindex to rebuild"?
57 2014-08-09 15:25:41 <sipa> yes
58 2014-08-09 15:25:44 <sipa> the message can certainly be improved
59 2014-08-09 15:26:08 <rnicoll> want to give me a pointer to where to find if it's running as a daemon? :)
60 2014-08-09 15:26:17 <rnicoll> wait, no, got a better answer
61 2014-08-09 15:26:30 <sipa> fDaemon?
62 2014-08-09 15:26:49 <rnicoll> ACTION greps "Looks good, thanks"
63 2014-08-09 15:27:16 <rnicoll> well... one way you want to present a question, the other a statement
64 2014-08-09 15:27:16 <sipa> thoug hi'm not convinced that should make a difference
65 2014-08-09 15:27:58 <sipa> if it's running as a daemon you can't give any message at all
66 2014-08-09 15:28:34 <sipa> and if it's not, i think you should still not assume a user is able to interact
67 2014-08-09 15:29:13 <rnicoll> for the QT version, it makes perfect sense to throw up a dialog to say "Hey, your database appears to be nonsense, want me to fix that?"
68 2014-08-09 15:29:45 <sipa> yup
69 2014-08-09 15:29:47 <rnicoll> however maybe the prompt needs a timeout
70 2014-08-09 15:36:38 <rnicoll> on an aside, anyone else sometimes find they want to actually spend coins, and they've broken every single client they have to hand?
71 2014-08-09 15:36:47 <rnicoll> maybe I shouldn't admit to that
72 2014-08-09 15:36:55 <rnicoll> I'm experimenting, I swear, I'm not that bad :-D
73 2014-08-09 15:59:36 <rnicoll> ACTION wanders back with some clippers and a yak
74 2014-08-09 16:00:01 <rnicoll> So, occurs really we should be able to prod the UI abstraction to determine if it's command-line, shouldn't we?
75 2014-08-09 16:14:38 <rnicoll> going to come back to that, no really easy answers without trampling all over other bits of code
76 2014-08-09 16:14:43 <rnicoll> right, take care all
77 2014-08-09 16:21:23 <topi`> has anyone successfully compiled cudaminer (based on jgarzik's cpuminer) without scrypt.cpp? I cannot compile that since it uses some x86 intrinsics and I'm on arm?
78 2014-08-09 16:50:06 <cfields> anddam: pong
79 2014-08-09 16:51:43 <cfields> anddam: we build with mostly static libs, except for qt, which is included in our dmg
80 2014-08-09 16:52:08 <cfields> anddam: so if macports builds more linuxy, relying on system shared libs, their bitcoin binaries would end up being much smaller
81 2014-08-09 16:53:03 <cfields> but after #4592, the osx binaries will be full static, which shrinks them down a bunch
82 2014-08-09 16:57:10 <cfields> full static build comes to 25M
83 2014-08-09 16:57:28 <copumpkin> is there a way in bitcoin-qt to mark an address as "don't use by default in transactions"?
84 2014-08-09 16:57:44 <copumpkin> i.e., I made an address for signature purposes and I don't want people to be able to send dust to it and figure out what I'm up to
85 2014-08-09 16:58:25 <pigeons> copumpkin: in the coincontrol dialog you can "lock" inputs, of course that's after someone sends dust to it
86 2014-08-09 16:59:02 <copumpkin> oh I see
87 2014-08-09 16:59:11 <copumpkin> didn't notice the lock option in the menu
88 2014-08-09 16:59:13 <copumpkin> thanks!
89 2014-08-09 16:59:22 <copumpkin> so I guess I'd ask someone to send dust to it pre-emptively
90 2014-08-09 16:59:26 <copumpkin> and then get it locked
91 2014-08-09 16:59:45 <sipa> you don't lock an address
92 2014-08-09 16:59:48 <sipa> you lock an output
93 2014-08-09 16:59:54 <copumpkin> oh
94 2014-08-09 17:00:11 <copumpkin> I guess that makes sense from a protocol standpoint, but doesn't really suit my use case then
95 2014-08-09 17:00:11 <pigeons> just use coincontrol everytime you send and dont let it chosse and you'll always know what inputs you are sending
96 2014-08-09 17:00:25 <copumpkin> yeah, I was hoping to avoid having to use coincontrol all the time
97 2014-08-09 17:01:02 <sipa> bitcoin-qt should ignore dust already
98 2014-08-09 17:01:34 <copumpkin> as in it just gets a low priority?
99 2014-08-09 17:02:46 <sipa> i don't know the precise way it works currently
100 2014-08-09 17:02:56 <sipa> but i can't imagine that it would pick dust as input
101 2014-08-09 17:03:08 <sipa> as that means it would need to add more fee than the value of the dust coins...
102 2014-08-09 18:02:52 <andytoshi> sipa: does secp256k1 accept the same crazy crap that openssl does? is it (supposed to be) safe to use it in consensus code?
103 2014-08-09 18:09:40 <sipa> andytoshi: no guarantee that it accepts the exact same set
104 2014-08-09 18:10:01 <sipa> though last i tried it correctly parses everything in mainnet and testnet
105 2014-08-09 18:12:18 <sipa> andytoshi: i want a consensus rules enforcing strict DER before enabling it
106 2014-08-09 18:13:12 <andytoshi> is that something that's gonna happen?
107 2014-08-09 18:14:22 <sipa> doesn't sound hard
108 2014-08-09 18:14:43 <sipa> there haven't been non-DER signatures in the main chain for ages
109 2014-08-09 18:14:53 <sipa> but it's only enforced by standardness, not consensus
110 2014-08-09 18:15:17 <andytoshi> it'd be neat if we could have a retroactive fork :P
111 2014-08-09 18:15:26 <andytoshi> though i'm sure as soon as there is a concrete plan some joker will go ruin it..
112 2014-08-09 18:16:13 <sipa> i hate to go through the whole increase-version, wait-until-75%, enforce-in-v3, wait-until-95%, require-v3 steps
113 2014-08-09 18:16:21 <sipa> so i want to combine it with Bip62
114 2014-08-09 18:18:10 <andytoshi> is there a schedule for bip62 becoming active?
115 2014-08-09 18:18:22 <andytoshi> is nversion=3 out there now?
116 2014-08-09 18:18:28 <sipa> not yet; we first need an implementation
117 2014-08-09 18:18:38 <sipa> bip62 isn't implemented anywhere; it's still a draft
118 2014-08-09 18:20:06 <andytoshi> ok, good to know. for now i'll just use secp256k1 :)
119 2014-08-09 19:06:40 <alebedev> are there any PHP developers here that can do a job unrelated to Bitcoin
120 2014-08-09 19:07:04 <alebedev> PM me
121 2014-08-09 19:07:04 <sipa> alebedev: not here
122 2014-08-09 19:07:06 <alebedev> what channel
123 2014-08-09 19:07:08 <alebedev> should i go to
124 2014-08-09 19:07:12 <sipa> i don't care; not here
125 2014-08-09 19:26:13 <raistlinthewiz> is there way to determine the category of a raw-transaction? like what we can see in pure gettransaction.details.category
126 2014-08-09 19:27:00 <sipa> ?
127 2014-08-09 19:27:40 <sipa> what is gettransaction.details.categofy?
128 2014-08-09 19:28:18 <raistlinthewiz> http://pastie.org/9458399
129 2014-08-09 19:28:23 <sipa> ah, you mean the details.category field op the rpc command
130 2014-08-09 19:28:24 <raistlinthewiz> 'category' : 'immature',
131 2014-08-09 19:28:30 <raistlinthewiz> yes
132 2014-08-09 19:28:48 <sipa> in general that is not possible for raw transactions
133 2014-08-09 19:28:49 <raistlinthewiz> is it also possible to check this with rawtransaction.vouts?
134 2014-08-09 19:29:00 <sipa> as it is about how thetransaction interacts with your specific wallet
135 2014-08-09 19:29:31 <sipa> though you can determine that a transaction is a generation transactions, and whether it's immature or not
136 2014-08-09 19:29:36 <sipa> and its confirmation count
137 2014-08-09 19:29:40 <sipa> but not whether it's a send or receive
138 2014-08-09 19:30:15 <raistlinthewiz> so i can only determine if it's immature or not with gettransaction rpc call
139 2014-08-09 19:30:52 <sipa> gettransaction can do everything
140 2014-08-09 19:30:55 <sipa> getrawtransaction can't
141 2014-08-09 19:31:21 <sipa> gettransaction queries your _wallet_ for that transaction, and returns information about the effect that transaction has on your wallet
142 2014-08-09 19:31:40 <sipa> getrawtransaction queries the transaction database, and returns wallet-independent information
143 2014-08-09 20:42:20 <josiah> Hi Folks
144 2014-08-09 20:47:38 <josiah> need help with testnet
145 2014-08-09 20:48:02 <sipa> we can't help without knowing what the problem is...
146 2014-08-09 20:48:27 <josiah> ignore that
147 2014-08-09 20:48:49 <josiah> my testnet is still syncing
148 2014-08-09 21:12:35 <anddam> cfields: back, still there?
149 2014-08-09 21:14:16 <anddam> cfields: reading #4592 now
150 2014-08-09 21:26:49 <anddam> cfields: are you building using gcc rather than clang?
151 2014-08-09 23:29:14 <josiah> hey guys how do i view the full transaction data
152 2014-08-09 23:29:31 <josiah> gettransaction only give me part of it
153 2014-08-09 23:30:56 <josiah> I want to get access to the source of the input value from the preivous transaction
154 2014-08-09 23:32:53 <amiller> josiah, a transaction contains enough information about the inputs (the txhash and the output number) to do gettransaction again on the earlier transaction
155 2014-08-09 23:34:51 <josiah> @amiller thanks
156 2014-08-09 23:35:08 <josiah> but I can't seem to glean that information from looking at the transaction data
157 2014-08-09 23:35:34 <josiah> I can only see the current transactions id
158 2014-08-09 23:35:49 <josiah> don't see the out number either
159 2014-08-09 23:39:13 <amiller> josiah, try bitcoind gettransaction <txid> 1
160 2014-08-09 23:39:18 <amiller> josiah, the 1 at the end indicates "verbose" mode
161 2014-08-09 23:39:56 <josiah> Thanks found another way
162 2014-08-09 23:40:18 <josiah> getrawtransaction