1 2015-04-09 00:11:29 <phantomcircuit> sipa, ah right
2 2015-04-09 01:04:57 <fanquake> ;;blocks
3 2015-04-09 01:04:57 <gribble> 351314
4 2015-04-09 01:36:01 <phantomcircuit> sipa, is there are reason not to sprinkle AssertLockHeld in various places?
5 2015-04-09 01:40:43 <phantomcircuit> 2015-04-09 01:09:39 ERROR: CScriptCheck(): e65dfdead7adaada52185d99f8e206b80df65fcb8028e79d66392625718f61b7:1 VerifySignature failed: Public key is neither compressed or uncompressed
6 2015-04-09 01:40:44 <phantomcircuit> wat
7 2015-04-09 01:44:20 <phantomcircuit> oh i see
8 2015-04-09 06:15:49 <wumpus> phantomcircuit: no reason; that's exactly what AssertLockheld is for. It only does the check when DEBUG_LOCKORDER is enabled, so in the normal case it is zero-overhead.
9 2015-04-09 06:52:26 <Anggga> hy
10 2015-04-09 06:52:40 <Anggga> help me please ?
11 2015-04-09 06:56:53 <Anggga> hadware bitcoin ?
12 2015-04-09 06:58:18 <wumpus> Anggga: #bitcoin
13 2015-04-09 07:20:30 <wumpus> ... going to tag 0.10.1rc2 in a jiffy, if anything needs to go in let me know ASAP
14 2015-04-09 07:22:07 <lewellyn> wumpus: btw, are there pre-built binaries (osx and/or windows) of master?
15 2015-04-09 07:22:19 <wumpus> no
16 2015-04-09 07:22:34 <wumpus> there were some initiatives at some point, but I think all of them dwindled
17 2015-04-09 07:22:55 <lewellyn> figures. romonster wanted to see master's icons in context, but doesn't have any build environments (as expected of an artsy person).
18 2015-04-09 07:23:04 <lewellyn> i'll have to try to find time to build one or both for her later.
19 2015-04-09 07:23:37 <wumpus> in general, master is quite risky to use, so providing ready-built executables may be dangerous, on the other hand it would result in more testing
20 2015-04-09 07:24:04 <wumpus> so I'm ok with "nightly builds" but don't want to advertise them too much
21 2015-04-09 07:24:13 <lewellyn> yeah. i wasn't expecting a positive answer for that reason. but at the same time, it seemed silly to build it myself for her to see the ui if someone else was offering binaries.
22 2015-04-09 07:25:16 <wumpus> right; I think jonasschnelli has some older builds with the new icons, if she isn't going to use the software that's enough to see the UI
23 2015-04-09 07:32:35 <wumpus> lewellyn: https://builds.jonasschnelli.ch/nightlybuilds/ it is still active!
24 2015-04-09 07:36:51 <lewellyn> awesome. i'll message it to her.
25 2015-04-09 08:16:06 <sipa> phantomcircuit: nope, go ahead and sprinkle
26 2015-04-09 08:21:13 <wumpus> * [new tag] 0.10.1rc2 -> 0.10.1rc2
27 2015-04-09 08:21:32 <wumpus> eh, that needs a v before it
28 2015-04-09 09:31:39 <rubensayshi> hmm, bitcoin core asumes OP_RETURN is followed by chunks of varint+string when parsing the script -> ASM? but that's just an asumption to make 'something' out of it I suppose (considering this TX does not follow that: https://gist.github.com/rubensayshi/2f8a2b50100345f278c2 )
29 2015-04-09 09:32:48 <Luke-Jr> rubensayshi: what? Bitcoin Core does not assume that
30 2015-04-09 09:32:52 <Luke-Jr> OP_RETURN is just a single opcode
31 2015-04-09 09:33:30 <rubensayshi> the gist I linked it output from `bitcoin-cli -testnet getrawtransaction 66385bf1c129d5659a93e1a00aeb576a5410306a29adc22f38fbeaaf9d8c9dbd 1`
32 2015-04-09 09:33:35 <Luke-Jr> also, it would be a pushdata opcode, not a varint
33 2015-04-09 09:33:43 <rubensayshi> sorry eya
34 2015-04-09 09:34:07 <Luke-Jr> *that transaction* has an OP_RETURN OP_PUSHDATA(â¦), but it isn't required
35 2015-04-09 09:36:36 <rubensayshi> okay, not sure I follow, why does bitcoin core give "[error]" at the end?
36 2015-04-09 09:43:05 <Luke-Jr> rubensayshi: the last pushdata is invalid
37 2015-04-09 09:43:33 <rubensayshi> yea and the script is 262 bytes, how is this valid?
38 2015-04-09 09:43:42 <Luke-Jr> OP_RETURN, push-4-bytes, push-46-bytes, push-46-bytes, push-39-bytes, push-34-bytes, push-38-bytes, push-44-bytes
39 2015-04-09 09:43:50 <Luke-Jr> but the last one is cut off before 44 bytes of data
40 2015-04-09 09:44:03 <Luke-Jr> rubensayshi: the invalid opcode is never reached in the interpreter
41 2015-04-09 09:51:03 <rubensayshi> Luke-Jr, why is that? and why if the script is invalid is this tx still valid? (I understand it's non standard for default configured bitcoin core because of op_return with > 80 bytes)
42 2015-04-09 10:03:53 <Luke-Jr> rubensayshi: OP_RETURN terminates the script immediately. Furthermore, pubkey scripts are never executed until they are spent.
43 2015-04-09 10:04:06 <Luke-Jr> So an unspent pubkey script can be basically anything (IIRC)
44 2015-04-09 10:05:35 <rubensayshi> so the OP_RETURN can be (probably is) follow by some data that is not at all a valid script, the fact that it looks like "push-4-bytes, push-46-bytes, push-46-bytes, etc" is probably a coincidence?
45 2015-04-09 10:06:02 <Luke-Jr> you could probably omit the OP_RETURN too (but then nodes couldn'
46 2015-04-09 10:06:10 <Luke-Jr> you could probably omit the OP_RETURN too (but then nodes couldn't prove it was unspendable/prunable)
47 2015-04-09 10:09:28 <rubensayshi> I feel a little stupid that up until now I've always sorta known that OP_RETURN terminates the script and now that I tried to dig a little deeper I started to confuse myself :/
48 2015-04-09 10:09:33 <rubensayshi> but thanks for the help!
49 2015-04-09 10:11:05 <Luke-Jr> np
50 2015-04-09 11:14:01 <michagogo> Hm, what's the "Upgrading system, may take a while" in gitian?
51 2015-04-09 11:21:31 <michagogo> It seems to be failing
52 2015-04-09 11:23:29 <michagogo> also, wtf? It worked for 0.10.1rc1 last night
53 2015-04-09 11:27:30 <wumpus> took a while here too, but eventually it passed, it can' tbe due to any difference between rc1 and rc2 (there's just one, not build system related)
54 2015-04-09 11:28:57 <Luke-Jr> wumpus: btw, another possible 0.10 backport: fixing TNIAB; if that's desired, let me know and I can finish the patch (but it requires the new chain param)
55 2015-04-09 11:29:26 <wumpus> Luke-Jr: ok maybe for a later 0.10 version, let's try to fix that for master first
56 2015-04-09 11:29:37 <Luke-Jr> k
57 2015-04-09 11:29:44 <Luke-Jr> should I go ahead and PR it with the chain params?
58 2015-04-09 11:30:17 <wumpus> I guess it will result in some discussion, maybe people will suggest better approaches, but it can't hurt to PR it
59 2015-04-09 11:32:17 <michagogo> wumpus: took a long time and then failed here
60 2015-04-09 11:32:19 <michagogo> twice
61 2015-04-09 11:32:37 <michagogo> I just edited the upgrade-system.sh to remove the redirects to /dev/null
62 2015-04-09 11:32:43 <wumpus> michagogo: strange - must be something with debian, then
63 2015-04-09 11:32:51 <wumpus> I mean, ubuntu
64 2015-04-09 11:32:53 <michagogo> (also, one line seems to redirect both to /dev/null and to a file?!?)
65 2015-04-09 11:33:11 <michagogo> Anyway, rerunning again, hopefully there'll be useful output this time
66 2015-04-09 11:34:42 <Luke-Jr> wumpus: is there already a PR to take checkpoints out of consensus?
67 2015-04-09 11:35:10 <wumpus> Luke-Jr: there's one by sipa that weakens the dependence, at least
68 2015-04-09 11:38:19 <Luke-Jr> hmm
69 2015-04-09 11:51:25 <michagogo> Okay, here's the relevant part of the log from the failure. No idea what changed here since last night... https://www.irccloud.com/pastebin/xLjCr9YR
70 2015-04-09 12:02:59 <michagogo> Anyone have any ideas?
71 2015-04-09 12:03:08 <michagogo> ACTION tries to tabcomplete devrandom and fails
72 2015-04-09 12:23:20 <solatis> is this channel meant for developers of Bitcoin Core, of developers that want to develop applications that make use of bitcoin?
73 2015-04-09 12:23:25 <solatis> *or
74 2015-04-09 12:26:26 <waxwing> solatis: the former.
75 2015-04-09 12:26:32 <waxwing> see "reference software" in title.
76 2015-04-09 12:28:47 <Luke-Jr> neither, actually. it's for people developing Bitcoin itself (whether that be BCCore, BitcoinJ, btcd, etc)
77 2015-04-09 12:28:50 <problumz> my client is stuck at 5 weeks behind, just says no block source available. it's not a network problem as everything else works fine
78 2015-04-09 12:29:08 <Luke-Jr> problumz: #bitcoin for support
79 2015-04-09 12:29:21 <problumz> ok
80 2015-04-09 12:34:50 <michagogo> Luke-Jr: any clue what might be going on?
81 2015-04-09 13:06:52 <Luke-Jr> michagogo: LXC?
82 2015-04-09 13:09:34 <michagogo> Luke-Jr: yes
83 2015-04-09 13:09:47 <michagogo> What's weird is that it worked last night
84 2015-04-09 13:09:51 <michagogo> And now it's not working
85 2015-04-09 13:10:12 <michagogo> I don't know what changed since then...
86 2015-04-09 13:14:05 <Luke-Jr> no surprise it can't update grub then.. not sure how to fix tho
87 2015-04-09 13:14:19 <Luke-Jr> what changed is probably Ubuntu putting out a grub update
88 2015-04-09 13:15:01 <wumpus> which breaks LXC :-/ that's why it worked for me, I use KVM
89 2015-04-09 13:15:33 <michagogo> Okay, I just nuked the base VM and and make-base-vm ing again
90 2015-04-09 13:16:10 <michagogo> I think there was some recent change to gitian involving grub or something
91 2015-04-09 13:16:17 <michagogo> Maybe that fixed one thing but broke another...
92 2015-04-09 14:24:06 <michagogo> I checked, and didn't see a grub update in the past 24 hours
93 2015-04-09 14:24:55 <michagogo> Anyway, as I expected/hoped, a freshly created base VM doesn't have any packages to upgrade and so now it's working without a problem
94 2015-04-09 15:03:49 <michagogo> wumpus: got the OS X sig yet?
95 2015-04-09 15:09:20 <michagogo> Also, a nice gitian feature (though only for KVM/vbox :-/) would be the ability to maintain multiple forked bases
96 2015-04-09 15:09:47 <michagogo> e.g. have a clean precise base, and then have 3 bases as forks off of that one
97 2015-04-09 15:10:12 <michagogo> With all the packages for Linux, Windows, and OS X installed
98 2015-04-09 15:10:59 <michagogo> Windows especially -- installing the compiler is a significant part of the time it takes
99 2015-04-09 15:32:55 <wumpus> michagogo: that would be a possiblity, though with the depends system the dependencies have been reduced enormously and we could do with one base image that has all the necessary packages installed
100 2015-04-09 15:33:44 <michagogo> wumpus: or in the other direction, wasn't there talk of making our own compilers?
101 2015-04-09 15:33:46 <wumpus> eg a possibility that someone could try is to just execute the gitian scripts on a Ubuntu 12.04 VM with the union of the packages installed
102 2015-04-09 15:34:05 <wumpus> yes, that's the eventual solution, it could do away with inner VMs entirely
103 2015-04-09 16:13:24 <gdistasi> Lightning network got the interest of Rusty Russel: http://rusty.ozlabs.org/?p=477
104 2015-04-09 16:35:31 <sdaftuar> wumpus: i think travis builds may fail until 5988 is merged
105 2015-04-09 16:42:37 <sdaftuar> oh, i just noticed you already fixed... never mind!
106 2015-04-09 16:45:46 <wumpus> thanks! but yes, I already noticed
107 2015-04-09 18:38:11 <phantomcircuit> gmaxwell, it seems like there's conflicting objectives with transactions propagation
108 2015-04-09 18:38:32 <phantomcircuit> on the one hand i believe rapid dissemination to the network is useful
109 2015-04-09 18:38:59 <phantomcircuit> but that seems to be in direct conflict with the goal of obscuring our connections to other nodes and obscuring which transactions are ours
110 2015-04-09 18:39:27 <phantomcircuit> clearly there's some middle ground here but i dont have an intuition as to what that should be
111 2015-04-09 18:42:58 <phantomcircuit> i'll just use a best guess but it should be scrutinized later
112 2015-04-09 19:04:08 <jonasschnelli> romonster, lewellyn: some of the nightly builds are broken. Had to flush/clear all of them. Just started to build a fresh one: https://builds.jonasschnelli.ch/nightlybuilds/2015-04-09/
113 2015-04-09 19:04:24 <lewellyn> ok. thanks. :)
114 2015-04-09 19:05:22 <jonasschnelli> lewellyn: if you need binaries, icons, screenshost: just ask.
115 2015-04-09 19:06:04 <lewellyn> it'd be romonster who'd need them. she's afk atm though.
116 2015-04-09 19:24:42 <phantomcircuit> https://github.com/bitcoin/bitcoin/pull/5989
117 2015-04-09 19:25:14 <phantomcircuit> improvimprovements to addr/inv trickle logic, critical analysis/review needed
118 2015-04-09 19:52:28 <sdaftuar> hm, looks like 5957 slowed regtest-mining down enough that some travis jobs are failing
119 2015-04-09 19:54:47 <sipa> i noticed that slowdown and was fixing it, but it git merged beforehand
120 2015-04-09 19:55:06 <sipa> i didn't realize it was so bad, though
121 2015-04-09 19:55:57 <sdaftuar> ok, if you've got a fix coming i won't worry about fixes/workarounds for the time being... just noticed that one of my travis jobs failed and was trying to figure out what to do with it
122 2015-04-09 21:25:00 <cfields> 0.10.1rc2 osx signature is up: https://bitcoincore.org/cfields/bitcoin-0.10.1rc2/signature.tar.gz
123 2015-04-09 21:26:06 <cfields> jonasschnelli / michagogo ^^
124 2015-04-09 21:35:31 <TheButterZone> any interesting ideas in "What if YOU could put 1 feature, change, whatever, into the Bitcoin protocol?" https://bitcointalk.org/index.php?topic=1017246.0;all
125 2015-04-09 21:43:15 <helo> see pullreqs?
126 2015-04-09 22:04:47 <paulo_> why Secp256k1?
127 2015-04-09 22:05:42 <sipa> ask satoshi
128 2015-04-09 23:33:43 <ldelarosa> I created a transaction and broadcasted without the network fee by mistake.
129 2015-04-09 23:34:03 <ldelarosa> Can I recreate the transaction with the network fee and broadcast it again?
130 2015-04-09 23:35:45 <moa> ldelarosa: you can, but do you want to?
131 2015-04-09 23:38:35 <cfields> gavinandresen: around?
132 2015-04-09 23:39:46 <ldelarosa> I would like the network to process the txs quicker .
133 2015-04-09 23:40:07 <ldelarosa> is it not safe right?!
134 2015-04-09 23:40:52 <cfields> gavinandresen: it doesn't look like the scheduler works correctly with older boost