1 2015-05-28 01:46:11 <Bitswim> Is anyone familiar with pybitcointools able to help me add my blockchain.info api key to get it to use that so my requests stop timing out? I will tip you generously for your time!
  2 2015-05-28 02:08:32 <poutine> try in #bitcoin
  3 2015-05-28 03:08:26 <Rozal> Whose in charge of bitcoin?
  4 2015-05-28 03:10:01 <bsm117532> Someone who passed grammar school.
  5 2015-05-28 03:11:09 <Squidicuz> ...is that a requirement?
  6 2015-05-28 03:49:14 <Rozal> Lol
  7 2015-05-28 03:49:30 <Rozal> Whouse in charge
  8 2015-05-28 03:56:39 <Squidicuz> hose in charge
  9 2015-05-28 09:23:54 <wumpus> harding: yes, I should really be siging the release announcements that contain a magnet URI
 10 2015-05-28 09:24:33 <wumpus> it's easy to put up a fake torrent (whether the SHA1 hash matches or not)
 11 2015-05-28 09:26:13 <wumpus> Rozal: the people who run full nodes, by deciding what software they use (this is a #bitcoin question, by the way)
 12 2015-05-28 09:39:40 <kinlo> wumpus: are we still hosting the downloads on sourceforge?
 13 2015-05-28 09:40:48 <wumpus> kinlo: there appear to be reasonably recent executables on sourceforge, I do not know who puts there there, though (must be either gavinandresen or jgarzik)
 14 2015-05-28 09:41:53 <wumpus> (or maybe sourceforge themselves, it's rumored that that happens too)
 15 2015-05-28 09:42:39 <wumpus> in any case I do not have access there
 16 2015-05-28 09:58:56 <maaku> gmaxwell: is it too soon to ask for a bip number reservation for nSequence relative lock-time?
 17 2015-05-28 12:01:44 <petertodd> maaku: number or not, you definitely need to do a demo application of the idea to clarify what's actually meant to be going on here, and what the advantages are exactly
 18 2015-05-28 12:02:36 <petertodd> maaku: note how there is both a demo written for CLTV, and at least some use-cases are in the "so obvious we have consensus without running code" category
 19 2015-05-28 12:17:43 <michagogo> Hm. Anyone want to respond to this? http://lists.alioth.debian.org/pipermail/pkg-bitcoin-devel/Week-of-Mon-20150525/000880.html
 20 2015-05-28 12:18:32 <michagogo> (Re: Debian having Bitcoin Core packaged as unstable-only and the maintainer considering changing that)
 21 2015-05-28 12:19:22 <hearn> there is an open letter floating around somewhere that asks linux distros not to package bitcoin core
 22 2015-05-28 12:47:29 <gavinandresen> wumpus: I haven’t uploaded binaries to sourceforge in a very long time. Also, if you need/want admin access to the sourceforge account, just send me your sourceforge login.
 23 2015-05-28 13:19:12 <bad_duck> about sourceforge "hijacking" project -> https://sourceforge.net/blog/gimp-win-project-wasnt-hijacked-just-abandoned/
 24 2015-05-28 13:46:06 <kuchbhi_llele> hey guys, can you please explain me, how does get blocks command work.
 25 2015-05-28 13:46:51 <kuchbhi_llele> ir can you please provide me link to an explanation,
 26 2015-05-28 13:46:56 <kuchbhi_llele> or*
 27 2015-05-28 13:50:30 <dhill> kuchbhi_llele: maybe http://godoc.org/github.com/btcsuite/btcd/wire#MsgGetBlocks may help.
 28 2015-05-28 13:58:28 <StormDev> Hi
 29 2015-05-28 14:02:09 <kuchbhi_llele> dhill, thank you, it cleared my doubt.
 30 2015-05-28 14:02:12 <kuchbhi_llele> :)
 31 2015-05-28 15:24:36 <maaku> petertodd: there's a demo application given in the email: fast channel closing. relative lock-time is a commonly talked about feature for exaxtly this transaction replacement purpose.
 32 2015-05-28 15:54:04 <aliakbar> hi! Question regarding newest release: in [main.cpp] was "bool static ProcessMessage(..)" changed to "bool static Processmessages(..)" since bitcoin-0.10.0????
 33 2015-05-28 16:03:37 <aliakbar> hi! Question regarding newest release: in [main.cpp] was "bool static ProcessMessage(..)" changed to "bool static Processmessages(..)" since bitcoin-0.10.0????
 34 2015-05-28 16:13:42 <wumpus> aliakbar: no, those are different functions
 35 2015-05-28 16:14:45 <petertodd> maaku: that's not a demo... there's no-where near enough details to figure out what's actually going on there
 36 2015-05-28 16:28:07 <maaku> petertodd: that's what the bip's for
 37 2015-05-28 16:28:48 <maaku> petertodd: but i'm a bit confused over why you're confused over relative lock-time
 38 2015-05-28 16:29:04 <maaku> how did you imagine fast channel close-outs being done?
 39 2015-05-28 16:29:34 <aliakbar> wumpus: thanks!
 40 2015-05-28 16:29:51 <aliakbar> Is it possible to test single functions of bitcoin source code quickly? without having built the entire code?
 41 2015-05-28 16:30:49 <StormDev> build your test case
 42 2015-05-28 16:32:31 <aliakbar> what is a test case?
 43 2015-05-28 16:32:36 <aliakbar> sorry I'm a newbie
 44 2015-05-28 16:32:54 <StormDev> a programming newbie?
 45 2015-05-28 16:33:04 <helo> you write code that only includes and executes the parts you want to test, and just compile and execute that
 46 2015-05-28 16:33:23 <StormDev> basically when you are building a software you want to test part of your code
 47 2015-05-28 16:33:32 <aliakbar> helo: is that what StormDev also means?
 48 2015-05-28 16:33:41 <StormDev> yes
 49 2015-05-28 16:33:51 <StormDev> usually I use a python script
 50 2015-05-28 16:34:13 <StormDev> I launch my C++ function with it and I look if the output is correct
 51 2015-05-28 16:35:18 <aliakbar> ok so I need to name my relevant modules, classes, libs, namespaces etc. in a test code's head?
 52 2015-05-28 16:35:55 <StormDev> Usually I use a wrapper to do it
 53 2015-05-28 16:36:03 <StormDev> it's  a kinda complex subject
 54 2015-05-28 16:36:17 <aliakbar> StormDev: ok, i'll research on wrapper
 55 2015-05-28 16:37:00 <aliakbar> StormDev: so with python, do you kinda "rephrase" your function from c++ to python or what?
 56 2015-05-28 16:37:05 <petertodd> maaku: it's "obvious" what the intent is, but I'd like to nail down specifics before I criticise it further :)
 57 2015-05-28 16:38:47 <StormDev> In python I compile my test
 58 2015-05-28 16:38:54 <StormDev> I check if the compilation is ok
 59 2015-05-28 16:39:26 <StormDev> then I execute my code with redirection of the imput/ouputs
 60 2015-05-28 16:39:39 <aliakbar> do you have a good aproach for self-learning, a source of tutorial or something
 61 2015-05-28 16:39:39 <StormDev> if the output == referenceoutput
 62 2015-05-28 16:39:44 <StormDev> my code si still valid
 63 2015-05-28 16:40:10 <StormDev> well, it really depends of what you want to do
 64 2015-05-28 16:40:17 <StormDev> unit test are hard to build
 65 2015-05-28 16:40:47 <aliakbar> let's say i want to test "ProcessMessages" from main.cpp of bitcoin-0.10.0
 66 2015-05-28 16:40:48 <StormDev> I learned it because i spent a lot of time working on my algorithms skills and co
 67 2015-05-28 16:41:05 <aliakbar> no problem i'll invest time
 68 2015-05-28 16:41:05 <StormDev> just write your output in a file
 69 2015-05-28 16:41:10 <StormDev> a check that file
 70 2015-05-28 17:37:05 <ahmed_> petertodd: you around?
 71 2015-05-28 21:58:48 <temujin> any update to #3656 or should i just build a 0.8 bitcoind and use that?
 72 2015-05-28 22:00:04 <gmaxwell> temujin: you shouldn't be using that.
 73 2015-05-28 22:00:18 <gmaxwell> (also you shouldn't be deploying exploitably vulerable software. 0_o)
 74 2015-05-28 22:00:58 <gmaxwell> temujin: you haven't said what you're trying to accomplish (see topic) but it's very likely that PR doesn't do what you think it dos.
 75 2015-05-28 22:01:03 <gmaxwell> s/dos/does/
 76 2015-05-28 22:01:47 <gmaxwell> (and what it does turned out to be more or less usless for any application)
 77 2015-05-28 22:02:51 <temujin> ah. well, i've been scouring the internet for a while, trying to find a way to recognize my own transaction in the block chain after i've sent it to the mempool.
 78 2015-05-28 22:03:19 <temujin> should i simply implement my own input/output hashing scheme and use that to identify my transaction?
 79 2015-05-28 22:03:32 <temujin> if there is available code, i'd like to use it
 80 2015-05-28 22:03:55 <temujin> (FWIW, using the 0.8 node ONLY for getnormalizedtxid RPC calls wouldn't be that bad, right?)
 81 2015-05-28 22:05:11 <gmaxwell> temujin: you shouldn't be identifying your transactions that way. Doing so will cause things to break. Identify your transactions by the fact that they spent your coins.
 82 2015-05-28 22:09:31 <temujin> gmaxwell: so, instead of trying to identify transactions by some scheme of input hashing, i should wait for 6 confirms and then check if each transaction contains an input spending one of my previously unspent outputs?
 83 2015-05-28 22:10:04 <temujin> thinking about it that way, i'd have to create a list of unspent outputs, and wait for each output to be spent in a fully-confirmed transaction...
 84 2015-05-28 22:11:33 <gmaxwell> It's unclear to me where your "wait" it coming from.  A transaction is on of yours if it pays you (obviously) or spends an output that previously paid you.
 85 2015-05-28 22:12:11 <temujin> well, the only reason i can't use the hash of inputs in a transaction to identify it is because those inputs can be mutated, right?
 86 2015-05-28 22:12:25 <temujin> so i must wait for 6 confirms for that to not be the case
 87 2015-05-28 22:14:57 <gmaxwell> Use for what?  If you're talking to payments to you, there is no hashing scheme that can prevent the hash of the final payment from changing.
 88 2015-05-28 22:19:54 <temujin> i want to somehow identify transactions as my own as they go from the mempool to a block, and finally reaching 6 confirms. we can think of these each as a step, and a transaction can go back or forward steps depending on what happens on the network. let's also assume every transaction uses "regular" pay to pubkeyhash, nothing fancy
 89 2015-05-28 22:24:01 <Luke-Jr> gmaxwell: H(scriptPubKey)
 90 2015-05-28 22:28:53 <temujin> i'm guessing the best algorithm to accomplish this task is to make a list of my unspent outputs, wait for them to be spent in a confirmed transaction, then mark that transaction as my own and mark the unspent outputs as spent
 91 2015-05-28 22:36:34 <gmaxwell> temujin: right.
 92 2015-05-28 22:37:19 <temujin> gmaxwell: ... is listtransactions trustworthy enough?
 93 2015-05-28 22:40:56 <gmaxwell> temujin: thats what its for; if its sutiable for any particular application depends on the details of the application.
 94 2015-05-28 22:41:40 <gmaxwell> Previous research showed Bitcoin Core as the only wallet which correctly handled numerous corner cases the researchers threw at it.
 95 2015-05-28 22:44:54 <temujin> gmaxwell: i guess my only remaining question is this: would listtransactions "know" about a mutated transaction? would it show the new txid, or the old one? or would it have two entries for the same transaction, each with a separate txid?
 96 2015-05-28 22:45:14 <temujin> (i can go read the code and figure this out... i suppose...)
 97 2015-05-28 22:45:18 <kanzure> well it definitely wouldn't know about it if your node hasn't seen it
 98 2015-05-28 22:45:54 <temujin> kanzure: good point, but for my case, let's just say i only care about transactions that have 6 confirms
 99 2015-05-28 22:48:46 <gmaxwell> temujin: if a conflict successfully gets into the blockchain it will show both, the old one with confirmations: -1   and both showing a txconflicts array.
100 2015-05-28 22:50:27 <temujin> awesome, good to know :)
101 2015-05-28 22:50:32 <temujin> i can just use that, then
102 2015-05-28 23:14:24 <drfoo> if i currently create a tx with 80 bytes of OP_RETURN do current nodes reject it?
103 2015-05-28 23:18:18 <s7r> drfoo isn't that the max. allowed limit?
104 2015-05-28 23:19:25 <drfoo> s7r: my understanding was that the limit is currently 40 but the change to 80 was just merged into master
105 2015-05-28 23:20:08 <drfoo> but it wont be live until 0.11 gets released
106 2015-05-28 23:21:21 <gmaxwell> thats a node specific setting, some won't relay with any opreturn data at all, some will relay larger. Defaults are different in master, things running that will apply that default unless overridden.
107 2015-05-28 23:22:53 <s7r> drfoo I don't know if there has recently been a change in master for OP_RETURN.
108 2015-05-28 23:23:03 <s7r> but I know this fugre for a very long time
109 2015-05-28 23:26:36 <gmaxwell> drfoo: It's really unlikely that you actually should need that much data there in any case, it's mostly intended for carrying hashes.
110 2015-05-28 23:30:27 <drfoo> gmaxwell: i know, but a project i'm working on has need for more than 40 bytes. i may have to just change what the data is or find a way to hash it
111 2015-05-28 23:30:47 <drfoo> or maybe encode it as receiving addresses
112 2015-05-28 23:31:40 <gmaxwell> drfoo: if you start shoving arbritary data in as a communications channel, esp in addresses you should expect nodes to start blocking your transactions at some point.
113 2015-05-28 23:32:23 <drfoo> even if its only 1 or 2 receiving addresses worth?
114 2015-05-28 23:33:10 <BlueMatt> cfields: /usr/include/qt/QtCore/qglobal.h:1050:4: error: #error "You must build your code with position independent code if Qt was built with -reduce-relocations. " "Compile your code with -fPIC (-fPIE is not enough)."
115 2015-05-28 23:33:32 <aschildbach> Does anyone know technical details about the recent blockchain.info RNG issue?
116 2015-05-28 23:34:12 <cfields> BlueMatt: shouldn't be an issue, how are you hitting that?
117 2015-05-28 23:34:26 <BlueMatt> bitcoin is defaulting to fPIE
118 2015-05-28 23:34:36 <BlueMatt> and my qt is #error'ing
119 2015-05-28 23:36:37 <BlueMatt> cfields: can we just switch to PIC and move on?
120 2015-05-28 23:36:38 <cfields> erm
121 2015-05-28 23:36:54 <cfields> what qt version?
122 2015-05-28 23:37:30 <gmaxwell> I think current Fedora, Ubuntu, and hardened gentoo just compile all things with pic all the time.
123 2015-05-28 23:38:11 <BlueMatt> cfields: https://www.archlinux.org/packages/extra/x86_64/qt5-base/
124 2015-05-28 23:38:31 <BlueMatt> or maybe qt4
125 2015-05-28 23:38:36 <BlueMatt> one of those
126 2015-05-28 23:39:50 <cfields> BlueMatt: iirc, fPIE has some advantages over fPIC. Since we're not building shared libs, I'd like to understand what qt's issue is first
127 2015-05-28 23:56:14 <cfields> BlueMatt: i don't see that in the upstream qglobal.h. can you grab the QT_VERSION_STR from that file ?
128 2015-05-28 23:56:34 <cfields> (upstream says: "Compile your code with -fPIC or -fPIE")
129 2015-05-28 23:59:31 <cfields> ok, there we go
130 2015-05-28 23:59:32 <cfields> https://qt.gitorious.org/qt/qtbase/commit/36d6eb721e7d5997ade75e289d4088dc48678d0d