1 2015-01-17 02:40:22 <jgarzik> hum
  2 2015-01-17 02:40:30 <jgarzik> where are those darned 0.10 release notes.
  3 2015-01-17 02:42:37 <sipa> in the 0.10 branch, under doc/release-notes.md
  4 2015-01-17 02:43:11 <sipa> https://github.com/bitcoin/bitcoin/blob/0.10/doc/release-notes.md
  5 2015-01-17 07:57:31 <jonasschnelli> coryfields, chinese/utf8 i think it's even more broken now. Check: https://github.com/bitcoin/bitcoin/pull/5671
  6 2015-01-17 07:58:03 <jonasschnelli> kanzure, conflictedbalance.sh is broken (on all branches). The test needs overhaul or remove
  7 2015-01-17 08:03:22 <coryfields> jonasschnelli: I saw the chinese problem, but that's something I was willing to live with for the sake of not changing the SDK at this stage
  8 2015-01-17 08:03:27 <coryfields> the DE I wasn't aware of, though
  9 2015-01-17 08:03:38 <coryfields> jonasschnelli: did you see that problem with lucida ?
 10 2015-01-17 08:03:59 <jonasschnelli> coryfields, no lucidia was better
 11 2015-01-17 08:04:39 <coryfields> jonasschnelli: i assumed that helvetica was a safer bet, since it's default now in 10.10. looks like I was wrong on that assumption
 12 2015-01-17 08:05:31 <jonasschnelli> coryfields, i think by turing back the default for non CN/JA languages to lucidia grand could be okay.
 13 2015-01-17 08:05:32 <coryfields> jonasschnelli: to be clear, I'm not worried about our code when changing sdks. I'm worried about the qt code (there are reported bugs caused by crashes due to missing symbols), and the cost of requiring that builders track down a new sdk before release
 14 2015-01-17 08:05:56 <coryfields> imo, it's similar to switching compilers just before release
 15 2015-01-17 08:05:59 <coryfields> jonasschnelli: ok, 1 min for a bin
 16 2015-01-17 08:06:09 <jonasschnelli> coryfields but 10.8 sdk is in the wild since 3. oct 2013
 17 2015-01-17 08:06:33 <coryfields> jonasschnelli: yes, but only with an apple account, and only when extracted on a mac
 18 2015-01-17 08:07:01 <coryfields> you and i are 2 of the very few mac builders around here
 19 2015-01-17 08:07:36 <jonasschnelli> coryfields i would say there are very limited amount of differences on Qt level to 10.6. I would also guess most differences will not affect us. But maybe im to pro-active on this. :)
 20 2015-01-17 08:08:06 <coryfields> jonasschnelli: it only takes referencing one weak symbol to cause a crash...
 21 2015-01-17 08:09:21 <jonasschnelli> coryfields Yes. Indeed.
 22 2015-01-17 08:09:23 <coryfields> and at the rc stage, introducing new bugs should not be an option, imo. At some point, you just have to ship with known bugs.
 23 2015-01-17 08:10:02 <jonasschnelli> coryfields Okay. Sounds also okay for me. Maybe we should try to upgrade to 10.8 in 0.11?
 24 2015-01-17 08:10:12 <coryfields> and if the bug is "values may look weird in chinese in osx in 10.10", i can live with that
 25 2015-01-17 08:10:16 <jonasschnelli> coryfields and i would also recommend to use "Hiragino Kaku Gothic Pro" as JA font substitue
 26 2015-01-17 08:10:26 <coryfields> jonasschnelli: sure, for master, we go to 10.9 or 10.10
 27 2015-01-17 08:10:44 <jonasschnelli> coryfields nice.
 28 2015-01-17 08:11:03 <jonasschnelli> coryfields the switch of the japanese font could change the ugly spaces (see https://cloud.githubusercontent.com/assets/178464/5788135/ef761b7a-9e25-11e4-9f4e-a971443002e9.png)
 29 2015-01-17 08:11:14 <coryfields> 10.9 works as-is, 10.10 will work once we figure out the clang situation. So I'll probably recommend bumping to 10.9 to begin with
 30 2015-01-17 08:11:19 <jonasschnelli> you have to look here: https://github.com/bitcoin/bitcoin/pull/5671
 31 2015-01-17 08:11:31 <jonasschnelli> coryfields agreed
 32 2015-01-17 08:11:54 <coryfields> jonasschnelli: ok, I'll try plugging in those two now. eta 5 min
 33 2015-01-17 08:12:10 <jonasschnelli> coryfields, perfect. Thanks!
 34 2015-01-17 08:13:27 <coryfields> jonasschnelli: fwiw, I used a few text editors in 10.10, and tried to see what they defaulted to. Though the japanese fonts were more suited, Heiti SC was always used as a fallback. I figured there may be less unicode gaps there.
 35 2015-01-17 08:14:48 <jonasschnelli> coryfields i didn't dive into the ja font issue... but i just saw in the "font app" (shipped with osx) that "Hiragino" looks very standard and wikipedia also said that this is one of the default ja-fonts. But i'm not sure on this.
 36 2015-01-17 08:14:55 <jonasschnelli> i think let's try and see.
 37 2015-01-17 08:14:59 <coryfields> not arguing one way or another, just saying that I didn't pick those fonts randomly :)
 38 2015-01-17 08:15:12 <jonasschnelli> currently the space-gaps are really big and the font looks somehow to script.
 39 2015-01-17 08:16:36 <coryfields> no luck with Hiragino Kaku Gothic Pro, I get blocks with that
 40 2015-01-17 08:17:22 <coryfields> jonasschnelli: the new OSX fonts use a system of fallbacks that QT doesn't understand. If it's starts with a ., qt ignores them. So we can't depend on the normal symbol fallbacks it usually uses.
 41 2015-01-17 08:18:59 <jonasschnelli> coryfields okay. Let me check again japanese with the font you choose. Once sec.
 42 2015-01-17 08:20:04 <jonasschnelli> coryfields, yes, it looks ugly. But for a rc3 bugfix it's okay IMO.
 43 2015-01-17 08:20:36 <coryfields> jonasschnelli: ok, I'll build a bin with just the lucida change. Maybe we can compromise there?
 44 2015-01-17 08:20:57 <jonasschnelli> coryfields fine for me
 45 2015-01-17 08:21:31 <jonasschnelli> I mean the 0.9.3 release even show only [] blocks in ja! So at least it's better then that!
 46 2015-01-17 08:22:25 <coryfields> jonasschnelli: i apologize if I seem stubborn here. I've handled many releases in other projects that have been sabotaged by innocent-looking last minute changes. I'd prefer to avoid that here :)
 47 2015-01-17 08:23:13 <coryfields> jonasschnelli: are you ok with the arial change?
 48 2015-01-17 08:23:35 <jonasschnelli> coryfields i totally trust your experience. I have no expertise in release management and probably would fall into these last-minut-bug-creation-changes.. :)
 49 2015-01-17 08:23:52 <jonasschnelli> coryfields the arial change makes no sense to me...
 50 2015-01-17 08:23:58 <coryfields> heh, thanks
 51 2015-01-17 08:24:00 <jonasschnelli> ah. wait.
 52 2015-01-17 08:24:09 <coryfields> ultimately wumpus has the final say. Maybe he'll slap me around :)
 53 2015-01-17 08:25:13 <jonasschnelli> i was talking about the switch from lucida to helvetica.
 54 2015-01-17 08:25:25 <jonasschnelli> whats about it with the arial thing?
 55 2015-01-17 08:25:28 <jonasschnelli> didn't check that
 56 2015-01-17 08:25:45 <coryfields> i'm changing back from helvetica to lucida, which should fix the DE issue, as i understand
 57 2015-01-17 08:25:47 <coryfields> correct?
 58 2015-01-17 08:25:53 <jonasschnelli> Yes.
 59 2015-01-17 08:25:59 <coryfields> ok
 60 2015-01-17 08:26:05 <jonasschnelli> Aha. The splash screen arial.
 61 2015-01-17 08:26:16 <coryfields> the arial change is here: https://github.com/theuni/bitcoin/commit/437de97c3ceab36189bf37b789d4c37e4386e166
 62 2015-01-17 08:26:21 <coryfields> yes
 63 2015-01-17 08:26:21 <jonasschnelli> I once did this and for no reason. So feel free to change it. :)
 64 2015-01-17 08:26:52 <coryfields> ok great. Chinese breaks down here because the unicode fallbacks aren't working. When using the default, which we set with these other changes, it works fine
 65 2015-01-17 08:28:17 <coryfields> uploading a test bin
 66 2015-01-17 08:28:54 <coryfields> (also, I'd like to change master so that osx builds use the pre-defined SDK for depends, rather than the system one)
 67 2015-01-17 08:29:36 <coryfields> i used the system one to make it easier for builds, but ultimately I'd prefer that you are able to build as releases are built, even if it's a bit more hassle
 68 2015-01-17 08:30:30 <coryfields> jonasschnelli: https://bitcoincore.org/cfields/bitcoin-qt-10.6_quickfix4-73cd4edb
 69 2015-01-17 08:31:02 <coryfields> only change from the current PR is helvetica -> lucida
 70 2015-01-17 08:31:37 <coryfields> so zh/ja will stay as-is, but de should be fixed
 71 2015-01-17 08:32:58 <jonasschnelli> coryfields Testes. Looks good. I mean we still have issues with whitecaps in CN/JA, but as said, better then it was before...
 72 2015-01-17 08:33:07 <jonasschnelli> coryfields ACK for me.
 73 2015-01-17 08:33:36 <jonasschnelli> you can push the change to 5671 and i will upload some screens
 74 2015-01-17 08:33:54 <coryfields> jonasschnelli: i'm willing to live with that
 75 2015-01-17 08:34:33 <coryfields> jonasschnelli: can we agree to go with whatever wumpus decides? If he'd rather bump the sdk, I'll do that instead.
 76 2015-01-17 08:35:12 <jonasschnelli> coryfields also good for me. I was not aware of that the 0.9.3 release also had this issue. First i thought we introduce new chars issue with 0.10. So 5671 is totally fine for me!
 77 2015-01-17 08:35:29 <coryfields> pushed
 78 2015-01-17 08:35:42 <jonasschnelli> coryfields yes. but now i also strongly recommend to keep the 10.6 sdk
 79 2015-01-17 08:36:12 <coryfields> jonasschnelli: thanks very much for all your help on this one! I thought I had a good enough system for testing, but I didn't realize that DE was an issue.
 80 2015-01-17 08:36:57 <coryfields> your patience is very appreciated :)
 81 2015-01-17 08:37:23 <jonasschnelli> coryfields i would say thank you for fixing this. I'm just the testing guy. :)
 82 2015-01-17 08:41:41 <coryfields> heh, hardly
 83 2015-01-17 08:41:47 <coryfields> ok, time for bed. Thanks again
 84 2015-01-17 08:41:51 <coryfields> nnite
 85 2015-01-17 08:43:16 <jonasschnelli> nite!
 86 2015-01-17 11:21:16 <jtimon> coryfields coryfields_ around? I just answered to your last comment on #5669
 87 2015-01-17 11:22:38 <jtimon> Basically I like your approach, just noting that it subtly changes the logs, replacing the error string with the reason string (which is perfectly fine IMO)
 88 2015-01-17 11:28:21 <LightRider> Is there a platform specific package of core for elementary os?
 89 2015-01-17 11:29:44 <wumpus> unless your distro has its own package, only for linux windows and mac
 90 2015-01-17 11:30:01 <LightRider> any ppa?
 91 2015-01-17 12:41:10 <fanquake> wumpus should we just open trivial pulls on theuni’s trival branch, rather than opening them on the main repo?
 92 2015-01-17 13:37:02 <wumpus> fanquake: indeed, could be a good idea
 93 2015-01-17 13:37:39 <fanquake> Be good to keep them out of the main repo/
 94 2015-01-17 14:14:14 <wumpus> coryfields: looks good to me for 10.0; for master I'd like to avoid all the workarounds and language-specific font matching
 95 2015-01-17 14:14:37 <wumpus> coryfields: so we'll just bump the sdk there
 96 2015-01-17 14:32:07 <kanzure> jonasschnelli: kthx
 97 2015-01-17 14:37:46 <wumpus> jonasschnelli: I think your httpbasics test has a mistake; I'm testing bitcoind with a different http server, and it passes on all the other tests except for  httpbasics. The reason for that is that you use GET to submit he RPC commands instead of POST.
 98 2015-01-17 14:38:36 <wumpus> GET is not normally used with a payload
 99 2015-01-17 15:09:56 <jonasschnelli> wumpus, hmm...
100 2015-01-17 15:14:12 <jonasschnelli> wumpus, did you try replacing the conn.request('GET with conn.request('POST?
101 2015-01-17 15:19:14 <Affirmed> Hee guys, i made some cash with bitcoins last year and bought a double bass, but then i gave the double bass away, how stupid am i? Is there any kind of bitcoin algo against this behaviour, can i zap myself? (i dont have a gun)...
102 2015-01-17 15:19:36 <wumpus> jonasschnelli: yes, that worked
103 2015-01-17 15:20:51 <jonasschnelli> wumpus, POST seems correct for JSON RPC. I assume bitcoinds RPC is more flexible with this. But it looks like a mistaken in the httpbasics.py test
104 2015-01-17 15:21:23 <jonasschnelli> wumpus, do you work on it and will change it or should i? It also needs update for gmaxwells keep-alive pr.
105 2015-01-17 15:23:04 <wumpus> jonasschnelli: well bitcond's RPC doesn't look at the http method at all, I think you could use 'WOOF' and it'd still work :)
106 2015-01-17 15:23:30 <jonasschnelli> wumpus: indeed. :)
107 2015-01-17 15:23:35 <wumpus> jonasschnelli: no need to change it right now
108 2015-01-17 15:24:15 <jonasschnelli> wumpus travis fails on https://github.com/bitcoin/bitcoin/pull/5674 because of the test. So i assume we need a fix for the test before merging (or master will fail on travis)
109 2015-01-17 15:25:50 <wumpus> jonasschnelli: you mean for the keepalive test; suer
110 2015-01-17 15:28:26 <sipa> Affirmed: no clue whatba double bass is, but no, bitcoin does not determije what you can do with it, and isbgenerally designed to be irreversible (if you need a refund, ask for it). please continue in #bitcoin though, this has nothing to do withbdevelopment
111 2015-01-17 15:28:52 <sipa> *determin, is generally, with development
112 2015-01-17 15:30:22 <jonasschnelli> wumpus, httpbasic.py is the keep-alive test, so i'll change the behavior there to represent #5674 and will also update from GET to POST
113 2015-01-17 15:32:23 <Affirmed> sipa: my privkey was a bit unsigned but it is working now thanks to constant development, tnx
114 2015-01-17 15:33:03 <sipa> private keys are not signed...
115 2015-01-17 15:34:08 <Affirmed> sipa: You got me, dammed...
116 2015-01-17 16:47:00 <andytoshi> good news everyone! the rust bindings for libsecp256k1 are working again
117 2015-01-17 17:24:23 <wumpus> woohoo
118 2015-01-17 17:29:06 <msvb-lab> wumpus: hoowoo
119 2015-01-17 18:35:04 <andytoshi> (i tried to bisect but it didn't fail at all during the bisection, so i'm checking a few commits near the tip as a sanity check. obvs i messed up the bisect command somehow..)
120 2015-01-17 18:35:23 <andytoshi> oops, that entire line was a typo (up-up-enter in the wrong screen)
121 2015-01-17 18:36:34 <michagogo> Erm, I was just reading impulse.pdf, noticed what seems like a mistake.
122 2015-01-17 18:36:42 <michagogo> 4. Party A builds but does not sign funding transaction TF1.
123 2015-01-17 18:37:06 <michagogo> Shouldn't that be "builds, signs, but does not transmit"?
124 2015-01-17 18:37:27 <michagogo> Or "builds, signs, and keeps secret"?
125 2015-01-17 18:38:14 <michagogo> Later it says 7. Party A signs TF1 and publishes to B and blockchain. Funds are now committed, once confirmed. Payment channel is open and funded.
126 2015-01-17 18:39:08 <michagogo> But AIUI, step "5. Party A builds refund transaction TR2, and transmits to party B." is impossible without signing TF1, because signing changes the txid
127 2015-01-17 18:39:54 <michagogo> Am I missing something? I wouldn't think BitPay would get something like that wrong
128 2015-01-17 18:44:16 <michagogo> That same mistake (assuming I'm not missing anything) is repeated over and over again throughout the document
129 2015-01-17 19:13:58 <michagogo> 21:02:10 <gdm85> midnightmagic: how do I exactly run something in the LXC container? I am not familiar with lxc-attach etc | 21:11:43 <midnightmagic> gdm85: in an lxc container, even if it's not running you can run a bash inside it with a single line that looks something like this: lxc-start -n gitian -f var/lxc.config -- sudo -u root LXC_ARCH=amd64 -i --
130 2015-01-17 19:13:58 <michagogo> netstat -rn
131 2015-01-17 19:14:11 <michagogo> If you look at gitian's readme, it's pretty easy
132 2015-01-17 19:14:26 <michagogo> The on-target script does it for both kvm and LXC
133 2015-01-17 19:14:54 <michagogo> Something like PATH=$PWD/libexec:$PATH
134 2015-01-17 19:15:23 <michagogo> Then, LXC_ARCH=amd64 LXC_SUITE=precise on-target bash
135 2015-01-17 21:56:43 <jgarzik> word
136 2015-01-17 22:01:04 <sipa> excel
137 2015-01-17 22:02:57 <justanotheruser> bitcoin enterprise edition 2014
138 2015-01-17 22:03:58 <jgarzik> rofl -- we should change 0.10 to 1.0 and call it Bitcoin Enterprise Edition ;p
139 2015-01-17 22:04:57 <justanotheruser> make sure there is Ribbon and Clippy
140 2015-01-17 22:05:47 <justanotheruser> "It looks like you're trying to make a transaction. Would you like help?"
141 2015-01-17 22:05:47 <michagogo> jgarzik: so was I right, or is there something I'm missing?
142 2015-01-17 22:06:43 <michagogo> justanotheruser: don't think the two go together
143 2015-01-17 22:07:20 <michagogo> Pretty sure the assistant was only until 2003, and then 2007 dropped it and added the ribbon
144 2015-01-17 22:07:57 <jgarzik> michagogo, Twitter royally sucks for an ongoing conversation
145 2015-01-17 22:08:10 <michagogo> jgarzik: indeed it does.
146 2015-01-17 22:08:21 <jgarzik> michagogo, Yes, you need to sign but not reveal
147 2015-01-17 22:08:36 <michagogo> That's what I thought
148 2015-01-17 22:08:56 <justanotheruser> sorry, I'll try to make my Office references more accurate
149 2015-01-17 22:08:57 <michagogo> I'm honestly surprised BitPay would make that kind of mistake
150 2015-01-17 22:09:49 <jgarzik> Rather disappointing.  I wonder if satoshi would have used sighash rather than raw hash in more places.
151 2015-01-17 22:10:04 <jgarzik> michagogo, <shrug>  it amounts to a typo, not a protocol problem
152 2015-01-17 22:10:41 <michagogo> Well, it's pervasive through the document, iirc
153 2015-01-17 22:19:55 <sipa> jgarzik: Bitcoin for Workgroups
154 2015-01-17 22:20:04 <jgarzik> lol
155 2015-01-17 22:21:07 <michagogo> Bitcoin Me.
156 2015-01-17 22:22:38 <sipa> Bitcoin Really Good(tm) Edition
157 2015-01-17 22:24:19 <michagogo> Bitcoin Bob?
158 2015-01-17 22:26:34 <jgarzik> michagogo, I can email the updated doc if you'd like
159 2015-01-17 22:26:45 <jgarzik> Just three places needed fixing AFAICS
160 2015-01-17 22:27:01 <jgarzik> thanks for the review
161 2015-01-17 22:29:29 <michagogo> jgarzik: sure
162 2015-01-17 22:30:42 <michagogo> No problem
163 2015-01-17 22:33:09 <jgarzik> heh
164 2015-01-17 22:33:21 <jgarzik> now that we've dropped the beta label, changing to v1.0 might not be a bad idea
165 2015-01-17 22:33:31 <jgarzik> we are certainly production software
166 2015-01-17 22:33:43 <jgarzik> OTOH, there is always the emacs model:  0.22.44
167 2015-01-17 22:36:20 <sipa> jgarzik: i think we should be converging towards 0.11
168 2015-01-17 22:36:33 <sipa> gavin would like that
169 2015-01-17 22:36:43 <jgarzik> hehehe
170 2015-01-17 22:36:53 <sipa> tex style
171 2015-01-17 22:39:11 <PRab> IMO, bitcoin core should hit 1.0 once pruning and some sort of UTXO commitment are integrated and tested.
172 2015-01-17 22:40:27 <PRab> I realize that the version number doesn't actually mean anything, but many people interpret 1.0 as "this is ready for the real world".
173 2015-01-17 22:41:09 <sipa> i think there is a significant difference between "the protocol is ready" and "the software is ready"
174 2015-01-17 22:41:12 <gmaxwell> PRab: I don't think non-contributors to the project musing about their favorte pet features as being the necessary and sufficent requirements  is really a productive use of time.
175 2015-01-17 22:41:17 <sipa> they ought to be mostly independent
176 2015-01-17 22:41:55 <PRab> fair enough on both accounts.