#instantbird log on 10 21 2012

All times are UTC.

00:06:36 --> rosonline has joined #instantbird
00:08:55 <-- SM0TVI has quit (Quit: I view things as they are, without regard to place or person; my country is the world, and my religion is to do good. -- Thomas Paine (*1737 †1809))
00:16:51 --> SM0TVI has joined #instantbird
00:23:34 <-- SM0TVI has quit (Ping timeout)
00:24:35 --> SM0TVI has joined #instantbird
00:24:40 <-- rosonline has quit (Quit: Experimentem Instantbird 1.2 -- http://www.instantbird.com)
00:26:38 <-- SM0TVI has quit (Ping timeout)
00:27:41 --> SM0TVI has joined #instantbird
00:45:18 --> wesj1 has joined #instantbird
00:53:46 <-- wesj1 has quit (Quit: wesj1)
01:10:00 <-- meh has quit (Quit: I don't want to live on this planet anymore.)
01:27:24 <instant-buildbot> build #317 of win32-onCommit is complete: Failure [failed shell_1]  Build details are at http://buildbot.instantbird.org/builders/win32-onCommit/builds/317  blamelist: Florian Qu?ze <florian@instantbird.org>
01:34:45 <-- mconley has quit (Input/output error)
02:03:50 --> deltafalcon has joined #instantbird
02:07:44 --> rosonline has joined #instantbird
02:11:22 <-- DGMurdockIII has quit (Quit: ChatZilla 0.9.89 [Firefox 17.0/20121010150351])
02:37:34 <instant-buildbot> build #661 of linux-nightly-default is complete: Success [build successful]  Build details are at http://buildbot.instantbird.org/builders/linux-nightly-default/builds/661
03:06:31 <-- rosonline has quit (Client exited)
03:15:26 <instant-buildbot> build #299 of macosx-onCommit is complete: Success [build successful]  Build details are at http://buildbot.instantbird.org/builders/macosx-onCommit/builds/299
03:15:28 <-- Optimizer has quit (Ping timeout)
03:31:38 <-- wnayes has quit (Quit: Instantbird 1.3a1pre -- http://www.instantbird.com)
03:47:31 --> mconley has joined #instantbird
04:09:23 <instant-buildbot> build #749 of win32-nightly-default is complete: Failure [failed shell_3]  Build details are at http://buildbot.instantbird.org/builders/win32-nightly-default/builds/749
04:20:23 <-- mconley has quit (Input/output error)
04:39:35 --> clokep has joined #instantbird
04:39:35 * ChanServ sets mode +o clokep 
04:48:42 <-- clokep has quit (Quit: Instantbird 1.3a1pre -- http://www.instantbird.com)
05:18:54 <-- skeledrew has quit (Ping timeout)
05:30:05 --> skeledrew has joined #instantbird
05:49:14 <-- Gizmokid2005 has quit (Ping timeout)
05:51:05 --> Gizmokid2005 has joined #instantbird
05:55:47 <-- Gizmokid2005 has quit (Ping timeout)
05:57:05 --> Gizmokid2005 has joined #instantbird
06:16:53 <-- Gizmokid2005 has quit (Ping timeout)
06:17:05 --> Gizmokid2005 has joined #instantbird
06:24:49 <-- Gizmokid2005 has quit (Ping timeout)
06:26:05 --> Gizmokid2005 has joined #instantbird
06:45:36 --> jeffgman has joined #instantbird
06:57:36 <-- jeffgman has quit (Quit: Instantbird 1.2 -- http://www.instantbird.com)
07:19:30 <-- Gizmokid2005 has quit (Ping timeout)
07:21:07 --> Gizmokid2005 has joined #instantbird
08:07:34 <-- deltafalcon has quit (Ping timeout)
08:09:55 --> deltafalcon has joined #instantbird
08:12:15 --> mikk_s has joined #instantbird
08:12:38 <-- mikk_s has quit (Client exited)
08:17:24 --> FeuerFliege has joined #instantbird
08:22:16 <-- deltafalcon has quit (Ping timeout)
08:23:29 --> deltafalcon has joined #instantbird
08:33:43 --> FeuerFliege_VM has joined #instantbird
08:34:20 --> jb has joined #instantbird
08:48:40 <-- jb has quit (Ping timeout)
08:53:39 <-- Mook has quit (Quit: Mook)
09:02:22 --> Optimizer has joined #instantbird
09:25:38 <-- FeuerFliege has quit (Quit: Instantbird 1.3a1pre -- http://www.instantbird.com)
09:25:44 --> FeuerFliege has joined #instantbird
09:29:12 <FeuerFliege> I just noticed, that the buddy icons don't work when you're not online. The icons are saved in the profile folder but it seems that they are not used until the account is/was connected.
09:29:12 <FeuerFliege> Is this know bug (or a feature)?
09:29:21 <flo> what would you all think of releasing 1.3 soonish (before Mozilla 17 is out), and have it our last release compatible with OS X 10.5 ?
09:30:17 <flo> FeuerFliege: I'm not surprised by that behavior, but if you are you can file a bug. I would be surprised if it got fixed soon though, as getting that right is a lot of work for a very limited benefit.
09:43:36 <FeuerFliege> flo: right, I only noticed it, becaus 
09:44:14 <FeuerFliege> e i tried some styles on the contact list and didn't want to reconnect every time.
09:44:39 <FeuerFliege> how old is OSX 10.5?
09:50:47 <flo> FeuerFliege: old enough that Mozilla 17 won't support it any more
09:51:18 <flo> FeuerFliege: released in october 2007
09:51:46 <flo> and the next version was released in august 2009
09:52:17 <flo> so 10.5 was still shipping on new machines in august 2009.
09:52:22 <flo> machines running it are 3+ years old
09:56:20 --> ueffel has joined #instantbird
09:56:23 <FeuerFliege> Well Gecko 17 Fx and TB will launch on 20.11.2012 ... not much time t finish 1.3
10:00:22 <flo> FeuerFliege: why would we need much time?
10:05:44 --> jb has joined #instantbird
10:20:09 <-- FeuerFliege_VM has quit (Input/output error)
10:27:45 <-- FeuerFliege has quit (Input/output error)
10:28:32 --> FeuerFliege has joined #instantbird
10:30:39 --> flo-retina has joined #instantbird
10:30:39 * ChanServ sets mode +qo flo-retina flo-retina 
10:33:52 --> DGMurdockIII has joined #instantbird
10:54:47 <-- FeuerFliege has quit (Input/output error)
10:58:56 --> meh has joined #instantbird
11:07:15 <-- jb has quit (Ping timeout)
11:13:46 --> FeuerFliege has joined #instantbird
11:14:47 --> gerard-majax_ has joined #instantbird
11:15:00 <-- gerard-majax has quit (Connection reset by peer)
11:21:48 <instantbot> New Instantbird (UI) bug 1735 filed by florian@instantbird.org.
11:21:49 <instantbot> florian@instantbird.org requested review from aletheia2@fastmail. fm for attachment 2000 on bug 1735.
11:21:50 <instantbot> Bug https://bugzilla.instantbird.org/show_bug.cgi?id=1735 nor, --, ---, florian, NEW, Shutdown leak if there are opened conversations at shutdown
11:22:11 <flo-retina> that shutdown leak turned out to be easier to debug than it seemed :)
11:23:16 <flo-retina> although it shows my review of bug 1683 sucked :-S
11:23:25 <instantbot> Bug https://bugzilla.instantbird.org/show_bug.cgi?id=1683 enh, --, 1.3, aletheia2, RESO FIXED, Restore participants' active status if still appropriate after a reconnect
11:23:49 <flo-retina> I had a bad feeling about that line, but didn't mention it because I couldn't come up with an explanation of why it felt wrong.
11:26:38 <-- flo-retina has quit (Ping timeout)
11:54:46 --> clokep has joined #instantbird
11:54:46 * ChanServ sets mode +o clokep 
11:57:43 <clokep> flo: Probably makes sense to do a 1.3 before end-of-lifing Mac OS X 10.5.
11:57:47 <clokep> Aka I'd be for tat plan.
12:01:49 <clokep> I guess there's not really any huge new features, but oh well...
12:05:06 <clokep> Windows build failed to make the installer? Was that looked at at all?
12:09:38 --> jb has joined #instantbird
12:10:26 --> jb1 has joined #instantbird
12:11:27 <-- jb has quit (Ping timeout)
12:11:35 --> jb has joined #instantbird
12:12:16 --> jb2 has joined #instantbird
12:12:19 <-- jb1 has quit (Ping timeout)
12:12:22 --> jb1 has joined #instantbird
12:12:23 <-- jb has quit (Connection reset by peer)
12:12:47 <-- jb2 has quit (Input/output error)
12:12:58 --> jb has joined #instantbird
12:14:36 <-- jb1 has quit (Ping timeout)
12:16:58 <instant-buildbot> build #657 of macosx-nightly-default is complete: Success [build successful]  Build details are at http://buildbot.instantbird.org/builders/macosx-nightly-default/builds/657
12:19:07 --> aleth has joined #instantbird
12:19:07 * ChanServ sets mode +h aleth 
12:19:29 <aleth> Thanks flo for tracking down the Linux startup bug! :)
12:19:33 <aleth> Sounds like that was quite the hunt...
12:19:40 <-- jb has quit (Quit: jb)
12:20:10 --> jb has joined #instantbird
12:20:25 <aleth> moz16 starting up fine now :)
12:20:33 <aleth> Just a couple new warnings...
12:23:50 --> igorko has joined #instantbird
12:32:34 <clokep> Can we close bug 1490?
12:32:37 <instantbot> Bug https://bugzilla.instantbird.org/show_bug.cgi?id=1490 nor, --, ---, nobody, NEW, Remove usage of nsILocalFile for Mozilla 16
12:33:04 <instantbot> aletheia2@fastmail.fm granted review for attachment 2000 on bug 1735.
12:33:06 <instantbot> Bug https://bugzilla.instantbird.org/show_bug.cgi?id=1735 nor, --, ---, florian, NEW, Shutdown leak if there are opened conversations at shutdown
12:34:06 <aleth> clokep: right - "Please note that a lot of servers actually set you to +i automatically when you first connect."
12:34:28 <clokep> aleth: Doesn't surprise me.
12:35:13 <aleth> I thought it might have something to do with nickserv settings.
12:35:30 <clokep> You might be able to disabl eit.
12:35:46 <aleth> It doesn't bother me- I was just wondering where it came from
12:35:58 * aleth is fixing that usermode bug just to get rid of it
12:36:42 <clokep> Hahah.
12:36:48 <clokep> What're you doing w/ the information?
12:37:04 <aleth> Thought about it, and am just going to write it to the server tab
12:37:16 <aleth> We have no good use for it and only power users will be interested in the first place
12:38:13 <aleth> Unless you have a better idea ;)
12:39:13 <clokep> Save it to the account object and write it to the server tab?
12:39:49 <aleth> Yes, the main thing is just to handle the associated messages
12:41:40 <clokep> right
12:50:23 <aleth> I also think it's a good idea to release before moz17
12:57:38 <clokep> We need to get Win builds working again first. ;)
12:57:55 <aleth> along with some other things ;)
12:58:07 <clokep> What other things?
12:58:23 <clokep> Well I think I just fixed the installer bustage...
12:58:23 <aleth> The checkin queue is getting long...
12:58:30 <clokep> Ah...yes.
12:58:44 <aleth> And there's wnayes patch
12:58:52 <clokep> We can push that to 1.4 if need be.
13:12:45 <flo> wnayes' patch won't be ready before the moz17 update
13:14:33 <flo> if we go with a 3-month cycle between 1.2 and 1.3 (ie release around November 10, and string freeze end of October), what I can commit to doing is:
13:14:34 <flo> - deal with the checkin-needed queue
13:14:35 <flo> - clean up most of my review queue (exclusing wnayes' patch that needs several iterations I think)
13:14:35 <flo> - finish the work on memory reporters for purplexpcom
13:14:35 <flo> - integrate Show Nick
13:14:36 <flo> - sync the build system with the current comm-release build system.
13:14:41 <flo> I think trying to do more would be unrealistic in that time frame
13:16:19 <flo> aleth: "Sorry - should have tested what actually happens on shutdown, rather than just assuming it's like closing a conversation." It's almost impossible to test without a debug build that dumps all error messages and shows shutdown leaks in the terminal. With a regular build, the error console gets closed at shutdown...
13:16:57 <aleth> flo: Yes, but dump() would have worked I think.
13:17:16 <flo> aleth: if you knew exactly what you wanted to test, possibly, yes :)
13:17:28 <aleth> Right :)
13:17:56 <flo> that's more work than just exiting a debug build and see if its messed up
13:18:20 <flo> also, I was quite surprised, because we have had a shutdown leak when a conversation was opened since before we released 1.2
13:18:41 <flo> I decided today to go ahead and spend time to debug it and fix it (if possible).
13:18:57 <flo> and I found that trivial recent mistake... and once I fixed that, there was no shutdown leak any more
13:19:15 <aleth> That's odd, since that only landed recently
13:19:15 <flo> so I have to guess that the issue I actually wanted to debug has been fixed by one of the moz updates...
13:19:39 <flo> I won't complain about that though :)
13:19:57 <aleth> Especially since the update wasn't trivial ;)
13:21:12 <flo> well, moz15 was painful, but moz16 more or less just worked (especially on PPC!!)
13:22:00 <flo> (well, on PPC it throws a crash reporter dialog at startup, but the application works fine after dismissing the dialog, so I'm assuming it's a background plugin-container process that crashes (when it shouldn't be started at all in the first place on PPC), and I don't really care about it)
13:23:01 <-- ueffel has quit (Ping timeout)
13:23:16 <-- aleth has quit (Quit: Au revoir)
13:26:04 --> aleth has joined #instantbird
13:26:04 * ChanServ sets mode +h aleth 
13:26:40 --> ueffel has joined #instantbird
13:27:57 <-- ueffel has quit (Quit: ueffel)
13:34:09 <-- gerard-majax_ has quit (Ping timeout)
13:39:09 --> gerard-majax_ has joined #instantbird
13:40:54 <-- gerard-majax_ has quit (Ping timeout)
13:41:46 <flo> clokep: are you looking at the windows build ? How are the xpcshell tests for you on Moz16 ?
13:42:56 <clokep> flo: I'm rebuilding right now...
13:43:33 <flo> I wonder how it was possible to notice that installer issue without having a finished build :-S
13:43:43 <flo> oh, you saw it on buildbot
13:43:44 <flo> nevermind :-S
13:44:27 <clokep> Yup! :)
13:44:55 <clokep> I started rebulding since it will take a while...and started seeing what was up w/ the installer...
13:45:02 <clokep> Parellelization! :)
13:46:04 <clokep> Something I would like to do (soonish?) is silent upgrades on Windows.
13:46:13 <clokep> I think that's Moz 17 though?
13:47:00 <flo> isn't the patch you ported related to that?
13:48:16 <instantbot> aletheia2@fastmail.fm granted review for attachment 2002 on bug 1727.
13:48:18 <instantbot> Bug https://bugzilla.instantbird.org/show_bug.cgi?id=1727 min, --, ---, aletheia2, ASSI, Separate strings for channel and user mode system messages
13:49:50 <clokep> Yeah, but I vaguely feel like it didn't work really well until 17?
13:50:46 <clokep> Well I'd like to do it ASAP, how does taht sound? :-D
13:52:41 <flo> that sounds like volunteering ;)
13:53:10 --> gerard-majax_ has joined #instantbird
13:54:22 <clokep> I even volunteered myself. :p
13:55:19 <-- DGMurdockIII has quit (Ping timeout)
13:56:14 <flo> "I vaguely feel like it didn't work really well until 17?" isn't it that if it shipped in 17, it can only work for the first time when updating to 18?
13:56:33 <-- jb has quit (Ping timeout)
13:56:54 --> jb has joined #instantbird
13:57:51 <clokep> Oh hmmm...maybe that's what I'm thinking. :)
14:00:42 * aleth has messed up something with git
14:01:46 <-- GeekShadow has quit (Ping timeout)
14:02:35 <aleth> Thankfully not a problem, but "broken HEAD" is not something I've seen before.
14:03:12 <clokep> aleth: I far too often hvae to deal w/ co-workers broken git repos. :(
14:03:52 <aleth> clokep: That sounds like painful merges :-(
14:04:39 <flo> clokep: on my local windows build, the purplexpcom xpcshell test crashes at shutdown
14:04:44 <flo> no idea of why :(
14:04:50 <flo> nor of how I could get a stack
14:05:16 --> GeekShadow has joined #instantbird
14:08:15 <clokep> aleth: Well yeah, we're bad at merging anyway. :-/
14:08:18 <clokep> flo: That's...weird. :(
14:08:48 <clokep> Doesn't that test just initialize purplexpcom and that's it?
14:10:27 <-- FeuerFliege has quit (Quit: Instantbird 1.3a1pre -- http://www.instantbird.com)
14:10:28 --> FeuerFliege has joined #instantbird
14:15:25 <flo> clokep: it also tests that libpurple prpls are accessible
14:15:36 <flo> clokep: the strange thing is that it crashes after the test has fully completed
14:18:23 <clokep> Oh, weird. :-/
14:18:30 <clokep> Does the build start and close successfully?
14:19:41 <flo> apparently calling core.getProtocols() or core.getProtocolById() is enough to make it crash at shutdown
14:19:50 <flo> the build works correctly with an AIM account
14:20:51 * flo is curious to know if the crash happens for you too
14:21:56 <clokep> I'll let you know when my build finishes!
14:35:26 --> mconley has joined #instantbird
14:36:36 <-- FeuerFliege has quit (Quit: Instantbird 1.3a1pre -- http://www.instantbird.com)
14:36:37 --> FeuerFliege has joined #instantbird
14:40:24 * aleth tries to use new FF16 JS feature, discovers it is not implemented yet after all
14:44:48 <-- jb has quit (Ping timeout)
14:48:55 <-- mconley has quit (Input/output error)
14:50:25 <clokep> aleth: What feature? :(
14:50:49 <aleth> The spread operator is only partially implemented
14:50:56 <aleth> I just updated the mdn page
14:54:06 <clokep> Those pages have been...not so great of late. :(
14:56:19 <flo> I won't troll about MDN, I won't troll about MDN, I won't...
14:56:37 <flo> wait, has the drop in qualify started with the introduction of persona as the only possible way to login? :-P
14:57:14 <aleth> Ha! That would be interesting :P
14:58:35 <instantbot> clokep@gmail.com set the Resolution field on bug 1490 to FIXED.
14:58:37 <instantbot> Bug https://bugzilla.instantbird.org/show_bug.cgi?id=1490 nor, --, ---, florian, RESO FIXED, Remove usage of nsILocalFile for Mozilla 16
15:05:44 <clokep> flo: The installer completed successfully with that pathc.
15:06:09 <clokep> And the tests run w/ no further patching.
15:06:12 <clokep> But one did fail...
15:08:14 <clokep> http://pastebin.instantbird.com/86851 is the results...
15:08:18 <clokep> Not sure if they match yours?
15:09:26 --> rosonline has joined #instantbird
15:11:06 <instantbot> aletheia2@fastmail.fm requested review from clokep@gmail.com for attachment 2003 on bug 1419.
15:11:09 <instantbot> Bug https://bugzilla.instantbird.org/show_bug.cgi?id=1419 nor, --, ---, aletheia2, ASSI, Handle the remaining MODE responses (user mode in particular)
15:14:37 <instantbot> aletheia2@fastmail.fm denied review for attachment 2003 on bug 1419.
15:14:40 <instantbot> Bug https://bugzilla.instantbird.org/show_bug.cgi?id=1419 nor, --, ---, aletheia2, ASSI, Handle the remaining MODE responses (user mode in particular)
15:14:54 * clokep likes when people r- their own reviews...
15:15:06 <clokep> aleth: When you upload a new one, let me know what's different.
15:17:54 <-- rosonline has quit (Client exited)
15:17:56 --> rosonline has joined #instantbird
15:20:03 <aleth> clokep: IRC is good at making simple things just a little more complicated.
15:20:47 <-- igorko has quit (Quit: Instantbird 1.2 -- http://www.instantbird.com)
15:23:07 <clokep> aleth: I find myself feeling lucky when it's only a *little* more complicated.
16:02:56 --> wesj1 has joined #instantbird
16:03:52 <-- wesj1 has quit (Quit: wesj1)
16:06:31 <-- deltafalcon has quit (Connection reset by peer)
16:06:47 <-- FeuerFliege has quit (Input/output error)
16:06:52 --> FeuerFliege has joined #instantbird
16:08:28 --> DGMurdockIII has joined #instantbird
16:08:39 <-- FeuerFliege has quit (Input/output error)
16:08:47 --> FeuerFliege has joined #instantbird
16:10:01 --> Even2 has joined #instantbird
16:10:47 <clokep> flo: It seems like it has something to do w/ memory reporters?
16:13:49 <instantbot> aletheia2@fastmail.fm requested review from clokep@gmail.com for attachment 2004 on bug 1419.
16:13:52 <instantbot> Bug https://bugzilla.instantbird.org/show_bug.cgi?id=1419 nor, --, ---, aletheia2, ASSI, Handle the remaining MODE responses (user mode in particular)
16:15:06 <-- FeuerFliege has quit (Input/output error)
16:21:27 <instantbot> clokep@gmail.com denied review for attachment 2004 on bug 1419.
16:21:29 <instantbot> Bug https://bugzilla.instantbird.org/show_bug.cgi?id=1419 nor, --, ---, aletheia2, ASSI, Handle the remaining MODE responses (user mode in particular)
16:29:49 --> jb has joined #instantbird
16:41:33 <-- jb has quit (Ping timeout)
16:44:06 <aleth> clokep: Ah, you're right of course, I forgot MODE doesn't always send the full user mode, so we need the whole apparatus.
16:44:53 <aleth> For some reason I thought the user's own mode was always returned in full and not as a diff.
16:46:14 <clokep> aleth: I don't think it is, even if it is...I'd like it stored in the same way.
16:47:24 <aleth> It's certainly better that way (just at first seemed like overkill to me)
16:47:42 <clokep> I'm a big fan of symmetry though. :)
16:49:20 <aleth> Me too... just that without your initial comment I probably wouldn't have stored the user mode at all, since we make no use of it.
16:49:44 <aleth> But we might as well...
16:50:25 <-- gerard-majax_ has quit (Client exited)
16:51:16 --> gerard-majax has joined #instantbird
16:51:31 <clokep> Yeah...I wasn't positive we should, but I think we shuld since we do for the conversations and participants.
16:52:08 <aleth> Maybe someone will use it for an add-on ;)
17:04:56 <clokep> aleth: Thanks for doing some of these silly patches for IRC that don't have a huge UX improvement.
17:07:24 --> wnayes has joined #instantbird
17:19:34 --> Kaishi has joined #instantbird
17:31:33 <flo> clokep: oh, you did that with a debug build and it printed a stack of the crash, awesome! :)
17:33:43 <flo> ah, or is that stack just for the ": 'Error', file c:/Users/Patrick/instantbird/objdir-debug/mozilla/xpcom/build/BlockingResourceBase.cpp, line 128" and not for the crash? :
17:33:45 <flo> (
17:36:13 <-- dewmugg has quit (Quit: Instantbird 1.2 -- http://www.instantbird.com)
17:36:28 <-- Optimizer has quit (Ping timeout)
17:39:43 --> Optimizer has joined #instantbird
17:40:23 <clokep> flo: It looks like a stack of the error, not the crash. :-/
17:40:50 <clokep> Is there a way to get a stack of the crash?
17:41:20 <flo> if you can find a way to reproduce the crash, attach visual studio to the crashed process? :)
17:41:43 <flo> or find how one is supposed to use that MINIDUMP_STACKWALK environment variable
17:42:22 <clokep> Hmm....OK... There's no crash when I actually run Instantbird. :-S
17:42:33 <flo> same for me
17:42:44 <flo> (otherwise I would have already attached visual studio ;))
17:45:11 <clokep> Ah, right. :)
17:48:31 --> mconley has joined #instantbird
17:49:08 <flo> it's just too bad there's no valgrind for windows :(
17:52:12 * flo is about to valgrind the xpcshell tests on Linux, just in case
17:53:06 <clokep> flo: Would trying to run the test in xpcshell manually allow us to attach to it?
17:53:26 <clokep> (Core.init is giving me an error though about nsIJSCID...)
17:53:27 <flo> you can try
17:53:46 <flo> I got stuck with an error in imContacts.js when getting the profile dir
17:53:57 <flo> clokep: that error just means you should read more of the test file ;)
17:54:15 <flo> especially http://lxr.instantbird.org/instantbird/source/purple/purplexpcom/src/test/test_purplexpcom.js#45
17:54:29 <flo> oh, and *I* should have noticed the do_get_profile file
17:54:33 <flo> s/file/line/
17:55:32 <clokep> ah-ha.
17:57:56 <flo> I only get valgrind errors from the python binary :-/
17:58:29 <flo> so I'm afraid the linux machine won't be of any help to debug this :(
18:01:15 * clokep wonders why that file defines Ci and doesn't use it everywhere...
18:15:59 <-- Optimizer has quit (Ping timeout)
18:19:23 --> Optimizer has joined #instantbird
18:23:42 <-- gerard-majax has quit (Ping timeout)
18:33:20 <aleth> "The MODE command is a dual-purpose command in IRC ... The rationale for this choice is that one day nicknames will be obsolete and the equivalent property will be the channel."
18:50:14 <-- mconley has quit (Input/output error)
18:51:42 <clokep> Hahaha.
18:54:53 <aleth> Thanks for clearing that up, RFC ;)
18:56:25 --> gerard-majax has joined #instantbird
18:59:48 --> FeuerFliege has joined #instantbird
19:05:02 <-- FeuerFliege has quit (Ping timeout)
19:11:24 --> FeuerFliege has joined #instantbird
19:18:52 <-- DGMurdockIII has quit (Quit: ChatZilla 0.9.89 [Firefox 17.0/20121010150351])
19:20:59 <instantbot> aletheia2@fastmail.fm requested review from clokep@gmail.com for attachment 2005 on bug 1419.
19:21:01 <instantbot> Bug https://bugzilla.instantbird.org/show_bug.cgi?id=1419 nor, --, ---, aletheia2, ASSI, Handle the remaining MODE responses (user mode in particular)
19:29:41 <clokep> aleth: I'm confused.
19:29:45 <clokep> Why does this patch look so different.
19:30:38 <clokep> Bah I'll reply in the bug.
19:38:53 <instantbot> clokep@gmail.com denied review for attachment 2005 on bug 1419.
19:38:56 <instantbot> Bug https://bugzilla.instantbird.org/show_bug.cgi?id=1419 nor, --, ---, aletheia2, ASSI, Handle the remaining MODE responses (user mode in particular)
19:41:57 <aleth> Huh, we do already handle 329 :P Just nobody updated that bug at the time ;)
19:55:08 <-- FeuerFliege has quit (Quit: Instantbird 1.3a1pre -- http://www.instantbird.com)
19:56:05 <instantbot> aletheia2@fastmail.fm requested review from clokep@gmail.com for attachment 2006 on bug 1419.
19:56:07 <instantbot> Bug https://bugzilla.instantbird.org/show_bug.cgi?id=1419 nor, --, ---, aletheia2, ASSI, Handle the remaining MODE responses (user mode in particular)
19:57:04 <aleth> typical MODE patch... gets longer every time
20:00:13 <clokep> aleth: A new bug had been filed for it at some point.
20:00:46 <aleth> I probably even reviewed the patch ;)
20:01:04 <aleth> Easy to forget stuff that does nothing.
20:01:38 <clokep> aleth: "That would be nice, but unfortunately when we are informed of our mode on
20:01:38 <clokep> connection, we get a MODE and not a 221. So it doesn't map as neatly as one
20:01:39 <clokep> would like." What?
20:02:02 <clokep> setUserMode is only called twice as far as I can tell, is there a third place?
20:02:09 <aleth> No
20:02:42 <aleth> On connection, the server tells us our mode, but it uses MODE to do so
20:02:59 <clokep> OK...
20:03:08 <aleth> Unfortunately.
20:03:21 <clokep> I'm still really confused.
20:03:22 <aleth> It's incorrect as it's the server setting the mode, not the user.
20:03:35 <aleth> Confused me too...
20:03:54 <clokep> Do you ever set aDisplayFullMode or is it hard coded as true in one of the places?
20:04:15 <aleth> It's hard coded as true for 221 (where we explicitly get the full mode)
20:04:49 <aleth> For MODE, we set it when previously no mode was set at all
20:04:57 <clokep> So my comment was...why do we pass that into setUserMode instead of just handling it as part of MODE and as part of 221?
20:05:29 <aleth> You want the MODE handler to mess with this._modes of the account?
20:05:40 <aleth> I thought that went against the encapsulation
20:05:55 <clokep> Why would it need to touch this._modes?
20:06:03 <aleth> To check its length
20:06:07 <clokep> Why can't it just print out the message and then call setUserMode?
20:06:19 <clokep> Oh...hmm...
20:06:29 <clokep> You're setting the parameter of a function internally, I didn't see that.
20:08:02 <clokep> I'm also a little confused about why we have two strings.
20:08:32 <clokep> My memory says we can ONLY ever have our own mode, so why do we need to have a version of the string that gives the nick? Isn't it always "Your mode is %S."?
20:09:02 <aleth> That's what I thought and that's what I did in the first patch
20:09:12 <aleth> But it's wrong when the mode changes.
20:09:33 <aleth> e.g. if you change your own mode "/mode -x"
20:09:42 <aleth> (or the server does it for some reason)
20:10:01 <clokep> What response to get in that situation then?
20:10:09 <aleth> The usual mode change string
20:10:23 <clokep> Please be more specific.
20:10:39 <aleth> Mode %1$S for %2$S set by %3$S.
20:11:49 <aleth> It's the most consistent thing to do I thought.
20:12:27 <clokep> Hmm....I need to think about it.
20:13:24 <aleth> It would be neat if there could be one handler to just display the mode, another for changes. But that's not possible due to the way the server behaves
20:14:47 <clokep> Alright, I understand that the server is sending back slightly different information, but I'm still confused at why we're checking the length of the modes to decide how to display the string.
20:15:16 <aleth> It's a proxy for "has the mode ever been set and displayed before"
20:16:00 <aleth> Alternatively one could remove _modes it from the prototype, but that's not good in case anything ever does use it.
20:16:35 <clokep> Alright.
20:17:10 <aleth> It's not ideal but I couldn't think of a better way to check without introducing a boolean, which seemed a bit much too
20:17:42 <clokep> I'll need to look again at what the server does.
20:18:25 <aleth> Basically on connect you get a message as if the user were setting the mode
20:19:57 <clokep> Alright.
20:20:21 <clokep> So you're trying to differentiate that time we receive a MODE message from other times?
20:20:28 <aleth> Exactly, yes
20:20:55 <clokep> Yeah, so that definitely needs a comment...
20:21:09 <clokep> But I wouldn't attach a new patch yet.
20:21:47 <aleth> Maybe you'll come up with a better way :)
20:21:49 <aleth> I wasn't sure where to put the shared function either
20:22:54 * clokep was hoping to not have to move the shared function...
20:37:33 --> DGMurdockIII has joined #instantbird
20:45:48 <-- Optimizer has quit (Ping timeout)
20:48:43 <-- meh has quit (Quit: I don't want to live on this planet anymore.)
20:53:16 --> Optimizer has joined #instantbird
20:59:20 <-- gerard-majax has quit (Ping timeout)
21:07:45 <instantbot> aletheia2@fastmail.fm cancelled review?(clokep@gmail.com) for attachment 2006 on bug 1419.
21:07:46 <instantbot> aletheia2@fastmail.fm requested review from clokep@gmail.com for attachment 2007 on bug 1419.
21:07:50 <instantbot> Bug https://bugzilla.instantbird.org/show_bug.cgi?id=1419 nor, --, ---, aletheia2, ASSI, Handle the remaining MODE responses (user mode in particular)
21:11:46 <-- wesj has quit (Ping timeout)
21:15:11 --> wesj has joined #instantbird
21:15:15 <-- DGMurdockIII has quit (Quit: ChatZilla 0.9.89 [Firefox 17.0/20121010150351])
21:16:27 --> DGMurdockIII has joined #instantbird
21:53:29 <flo> clokep: so what's the plan for the Windows crash?
21:53:55 <clokep> flo: I don't know what to try next. :-/ We can disable that test for now to at least get a nightly back.
21:53:59 <flo> I'm going to push the installer patch, as it can't hurt to fix that part
21:54:01 <clokep> But that's not a good long term plan obviously. :)
21:54:16 <flo> hmm
21:54:36 <flo> can we do a runtime OS check in xpcshell, or is that busted by the fake AppInfo ?
21:55:21 <clokep> Hmm...we can check if the registry editing service is available instead of using AppInfo.
21:57:09 <clokep> flo: Do you need to "make" at all again after changing a test or do you just rerun the test?
21:57:30 <clokep> (Also is there a way to run just the purplexpcom tests? I couldn't find the right incantation besdies just objdir/mozilla xpcshell-tests
21:57:34 <flo> I need to make in the objdir/purple/purplexpcom/src/ folder after changing that test
21:58:19 <flo> I think it's make SOLO_TEST=<dir name or test name> check-one
21:58:38 <flo> but our xpcshell tests are fast enough that I don't mind running them all
21:58:57 <clokep> Oh OK. I was just curious since I run the irc ones by building objdir/chat/protocols/irc
21:59:29 <clokep> I didn't realize there was like more complicated syntax.
21:59:57 --> wesj1 has joined #instantbird
22:07:02 <flo> why does lufthansa's online check-in require so much flash :(
22:08:02 <clokep> Oooo good thing you're not trying to use your phone.
22:13:07 <flo> how can they ask for "name, as on your passport" and not accept accentuated characters in the form? :-S
22:18:46 <-- aleth has quit (Input/output error)
22:20:07 <-- wesj1 has quit (Quit: wesj1)
22:25:25 --> Mook has joined #instantbird
22:33:25 <clokep> flo: Can we just attach to xpcshell during the test and catch the crash or does it not work that way?
22:33:46 <-- Optimizer has quit (Ping timeout)
22:34:06 <flo> I haven't used visual studio for debugging in years, so I won't be of much help, sorry :(
22:34:50 <clokep> Hmm....OK.
22:35:23 <clokep> Yeah I don't see an xpcshell in the list. :-/
22:35:39 <clokep> Was why I was kind of hoping to be able to reproduce it via a running xpcshell.
22:37:02 --> Optimizer has joined #instantbird
22:38:43 <flo> have you tried to reproduce it in xpcshell?
22:39:26 <clokep> A little bit.
22:39:55 <clokep> Some of the lines were giving me cryptic errors / I don't really have any idea how to use xpcshell...
22:40:49 <flo> ./xpcshell <name of the script to execute>
22:41:12 <flo> (you don't have to execute the commands one by one at the interactive prompt)
22:41:21 <clokep> Oh, that's easy.
22:41:31 <clokep> Yeah that didn't seem to work anyway...:-/
22:41:37 <clokep> (The individual ones...)
22:42:49 <Mook> if that works, you can open xpcshell.exe as a solution in visual studio (...), right click on the thing in the project tree, and change startup params there, and run it
22:42:54 <clokep> flo: http://pastebin.instantbird.com/86952
22:43:02 <clokep> No crash. :(
22:43:18 <clokep> http://pastebin.instantbird.com/86953 is what I ran FWIW.
22:50:08 <clokep> Any other ideas of how to debug this Mook? :)
22:50:49 <Mook> gimme a minute, need to read logs :)
22:52:02 <Mook> we're looking at http://pastebin.instantbird.com/86851 ?
22:52:32 <clokep> Yes.
22:53:03 <Mook> can you run _only_ that one test, instead of the whole suite, and have it crash?
22:53:10 <clokep> Hmm...I haven't tried.
22:54:06 <Mook> That's your first step, I think :) The second step (regardless of results of the first step) is https://developer.mozilla.org/en-US/docs/Writing_xpcshell-based_unit_tests#Via_check-interactive I think
22:54:37 <clokep> First step is finding how to run a single test. ;)
22:54:42 <Mook> (the difference would be how many tests you need to run in step 2 - attach the debugger before running check-interactive)
22:55:35 <Mook> note that I'm assuming here the docs actually match what you're running :D
22:57:06 <clokep> I'm running $ mozmake SOLO_TEST=purple/purplexpcom/src/test/test_purplexpcom.js -C objdir-debug check-one
22:57:09 <clokep> Does that seem right?
22:57:14 <clokep> It's saying nothing to be done for check-one.
22:57:58 <Mook> SOLO_FILE?
22:58:53 <Mook> or possibly, SOLO_FILE=test_purplexpcom.js make -C objdir-debug/purple/purplexpcom/src/test/ check-one
22:58:58 <clokep> Mook: Arg. I'm an idiot...
22:59:36 <clokep> Hmm....no module named mozinfo...
23:01:09 <flo> so clokep, in your log your test wasn't run
23:01:26 <flo> "imContacts.js:14: NS_ERROR_FAILURE: Component returned failure code: 0x80004005 (NS_ERROR_FAILURE) [nsIProperties.get]"
23:01:33 <flo> it fails to get to the profile directory
23:01:38 <flo> which is exactly where I failed when I tried
23:01:39 <clokep> Oh. Interesting...
23:01:47 <flo> http://log.bezut.info/instantbird/today#m304
23:02:07 <flo> if you are willing to spend one minute on that error though, you can just reimplement do_get_profile
23:02:26 <flo> you will need to hard code the path to a temporary profile file
23:02:32 <clokep> OK, how do I do that?
23:02:49 <flo> you take mxr, you look at how it's implemented, you copy paste, etc...
23:02:56 --> jb has joined #instantbird
23:03:14 <clokep> http://mxr.mozilla.org/mozilla-central/source/testing/xpcshell/head.js#730
23:03:57 <flo> sure, see http://mxr.mozilla.org/mozilla-central/source/testing/xpcshell/head.js#744
23:04:01 <flo> you can hard code something instead of that
23:04:07 <flo> or give a path in that environment variable
23:04:17 <clokep> Wait. I'm kind of confused.
23:04:36 <clokep> Is this in the xpcshell test itself or when running it as a script in xpcshell?
23:04:46 <flo> Mook: the test reliably crashes, if and only if we have done an action that forces libpurple to initialize itself.
23:05:56 <flo> Mook: "an action that forces libpurple to initialize itself" that is, core.getProtocols() (listing all protocol plugins forces the libpurple init) or core.getProtocolById() (with an id of a prpl implemented by libpurple)
23:07:14 <flo> Mook, clokep: what's the name of the interface I should check to see if we are on Windows?
23:07:33 <flo> ah,  isWindows = '@mozilla.org/windows-registry-key;1' in Components.classes;
23:07:42 <-- SM0TVI has quit (Ping timeout)
23:07:50 <flo> (so why did I look for it for ~5 minutes, and find it within 10s after asking for it?)
23:08:26 <-- jb has quit (Ping timeout)
23:08:32 <clokep> flo: Yes, that's whe one. :)
23:08:58 --> SM0TVI has joined #instantbird
23:09:02 <Mook> so, reading the pastebin I linked (I have no clue what's going on):
23:09:18 <Mook> things might work if you use NS_THREADSAFE_MEMORY_REPORTER_IMPLEMENT instead at http://lxr.instantbird.org/instantbird/source/purple/purplexpcom/src/purpleMemory.cpp#73
23:11:03 * clokep is trying that...
23:11:11 <flo> Mook: heh, we debug in the same way ;). I've already tried that. It didn't help. (and I was quite disappointed)
23:11:37 <Mook> aww
23:12:03 <Mook> (I'm assuming http://dxr.allizom.org/mozilla-central/xpcom/glue/nsISupportsImpl.h#l428 is failing at this point)
23:12:07 <clokep> Yeah. Didn't work...
23:13:04 <Mook> same stack?
23:13:36 <Mook> (actually, just pastebin it again in case I'm reading the wrong one to start with, please)
23:13:37 <flo> Mook: we don't have any stack for the crash
23:13:45 <flo> Mook: the stack in the log is just for an error in a debug build
23:14:01 <flo> Mook: although we obviously all believe that the error and the crash are *likely* related :)
23:14:07 <Mook> ah, okay. (yeah, it's just a deadlock assertion)
23:14:28 <Mook> oh, minidump found! can you load that dmp in visual studio?
23:14:38 <clokep> minidump found?
23:14:49 <Mook> http://pastebin.instantbird.com/86851 line 96/97
23:15:23 <Mook> (also, you still want to fix the deadlock anyway, but that's not important at the moment, it'll just hang later :p )
23:15:23 <clokep> There's no dmp in that folder.
23:15:27 <Mook> sigh :(
23:15:37 <clokep> I don't actually have minidump AFAIK.
23:15:53 <flo> Mook: aren't these dumps in the breakpad format, rather than MS format?
23:16:20 <flo> Mook: I wonder if it would be possible to reproduce the deadlock assertions on a linux debug build
23:18:22 <Mook> IIRC, the breakpad format is the MS format, at least on Windows
23:18:39 <flo> clokep: http://pastebin.instantbird.com/86970 r?
23:18:50 <Mook> (it calls into platform things to generate them, I think. breakpad has its own symbol format, though, that's different from *.pdb)
23:19:56 <clokep> flo: r+
23:19:58 <flo> (at this point I think it's that patch, or no Windows nightlies for a few more days)
23:20:01 <flo> clokep: ok
23:20:16 <clokep> (Although that makes me nervous because there's no pressure to fix it....)
23:21:05 <Mook> this crash is windows-only? (because building the whole thing on Windows is a pain on my side :p)
23:21:07 <clokep> Is this minidump made as part of a build?
23:21:24 <Mook> no, it's a dump of the process state on a crash
23:21:29 <flo> Mook: yes, Windows only
23:21:59 <flo> Mook: I have some (limited) hope that the assertions about dead locks would also exist on more useful platforms.
23:22:53 <flo> clokep: yeah, it makes me nervous too (also the fact that tests don't run on Mac because I couldn't figure out how to make them work on PPC is ugly), but at this point the length of the checkin-needed queue is also a concern for me
23:23:15 <clokep> flo: I agree. Let's do this. Get the checkin queue down.
23:23:26 <flo> but please fix it anyway :)
23:23:34 <clokep> I'll try to, of course!
23:23:38 * flo doesn't want to feel responsible for fixing Windows crap.
23:23:39 <-- DGMurdockIII has quit (Quit: ChatZilla 0.9.89 [Firefox 17.0/20121010150351])
23:23:43 <clokep> But I' mkind of out ideas...
23:23:46 <flo> PPC is already enough of a hassle
23:25:14 <Mook> this is trunk ib?
23:25:47 <Mook> ugh, nevermind, I don't have a working windows VM around.. and I need to leave now :(
23:26:03 <clokep> Mook: Yes. Trunk.
23:26:32 <Mook> I'll try to set up a dev env when I get back, and see if I can repro there. but for now, I need to leave.
23:26:41 <-- Mook has quit (Quit: gone)
23:26:44 <clokep> Thanks a lot Mook!
23:27:19 <instant-buildbot> build #300 of macosx-onCommit is complete: Exception [exception interrupted]  Build details are at http://buildbot.instantbird.org/builders/macosx-onCommit/builds/300  blamelist: Florian Qu?ze <florian@instantbird.org>, Patrick Cloke <clokep@gmail.com>
23:27:51 <clokep> Did you cancel the build?
23:27:56 <flo> yes
23:28:01 <clokep> OK. :)
23:29:18 <flo> Mac oncommit builds started after 1am, cause the nightly to fail, because the machine powers itself off before the oncommit + nightly are both finished
23:29:18 <flo> Good night
23:29:18 <flo> I really needed to go 2 hours ago :-/.
23:29:25 <flo> i'll be offline most of tomorrow (flights...)
23:29:29 <-- flo has quit (Quit: Instantbird 1.3a1pre -- http://www.instantbird.com)
23:30:26 <instantbot> Check-in: http://hg.instantbird.org/instantbird/rev/cd8445156002 - Patrick Cloke - Follow-up to bug 1489 - Port Bug 765596 - Error in macro InstallOnInitCommon on macroline 152.
23:30:27 <instantbot> Check-in: http://hg.instantbird.org/instantbird/rev/3e6bdc5f695c - Florian Quèze - Follow-up to bug 1489 - avoid initializing libpurple from the test_purplexpcom.js xpcshell test on Windows, as it causes the test to crash, r=clokep over IRC.
23:49:23 <instant-buildbot> build #331 of linux-onCommit is complete: Success [build successful]  Build details are at http://buildbot.instantbird.org/builders/linux-onCommit/builds/331