#instantbird log on 01 19 2014

All times are UTC.

00:11:07 --> mconley has joined #instantbird
00:19:01 <-- EionRobb has quit (Quit: Leaving.)
00:26:57 <-- mconley has quit (Input/output error)
00:32:54 --> BWMerlin has joined #instantbird
00:34:34 --> Mook has joined #instantbird
00:38:37 <-- nathanr has quit (Ping timeout)
00:39:31 --> nathanr has joined #instantbird
03:25:32 --> mconley has joined #instantbird
04:21:07 <-- mconley has quit (Input/output error)
05:53:50 --> iamjayakumars has joined #instantbird
05:54:30 <-- Mook has quit (Quit: Mook)
06:15:11 <-- iamjayakumars has quit (Quit: )
07:28:59 <instant-buildbot> build #1716 of macosx-nightly-default is complete: Failure [4failed compile]  Build details are at http://buildbot.instantbird.org/builders/macosx-nightly-default/builds/1716
07:40:58 <instant-buildbot> build #1236 of win32-nightly-default is complete: Failure [4failed shell_3]  Build details are at http://buildbot.instantbird.org/builders/win32-nightly-default/builds/1236
08:10:30 --> jb has joined #instantbird
08:22:25 --> GeKo has joined #instantbird
08:51:59 <-- jb has quit (Ping timeout)
09:13:56 --> jb has joined #instantbird
10:28:10 --> gerard-majax_ has joined #instantbird
10:43:58 <-- mrphs has quit (Ping timeout)
10:46:53 --> mrphs has joined #instantbird
10:54:07 <-- gerard-majax_ has quit (Ping timeout)
11:24:46 <-- jb has quit (Ping timeout)
11:58:25 --> mconley has joined #instantbird
12:39:57 <-- BWMerlin has quit (Quit: BWMerlin)
12:53:48 --> jb has joined #instantbird
12:58:22 <-- jb has quit (Ping timeout)
13:30:54 --> aleth has joined #instantbird
13:30:55 * ChanServ sets mode +h aleth 
13:32:44 <aleth> Re the website font issue - how about switching it to Open Sans?
13:33:19 <aleth> Similar enough to probably not cause any issues, and open.
13:33:28 --> gerard-majax_ has joined #instantbird
13:35:03 <aleth> Would it make sense to make "show offline contacts" the default setting?
13:35:19 <aleth> Seems like that would solve the problem.
13:38:20 --> mconley_ has joined #instantbird
13:38:33 <-- mconley has quit (Ping timeout)
13:56:10 --> hadi has joined #instantbird
14:02:14 <-- mconley_ has quit (Input/output error)
14:07:53 --> FireFly_TB has joined #instantbird
14:12:30 <-- FireFly_TB has quit (Ping timeout)
14:32:37 --> mconley has joined #instantbird
14:38:43 <-- mconley has quit (Ping timeout)
14:54:06 <Tonnes> aleth: please fix it accordingly :) or propose it in the bug
14:55:27 <Tonnes> (the font issue)
15:08:59 <-- aleth has quit (Ping timeout)
15:12:42 --> aleth has joined #instantbird
15:12:42 * ChanServ sets mode +h aleth 
15:19:38 <-- aleth has quit (Quit: exit stage left)
15:19:48 --> aleth has joined #instantbird
15:19:48 * ChanServ sets mode +h aleth 
15:26:02 <flo-retina> aleth: "make "show offline contacts" the default setting?" no
15:26:36 <flo-retina> at least not without a way to make online/available (or even better frecent) contacts bubble up
15:27:19 <aleth> Yes, it would be an ugly "solution"...
15:32:03 <flo-retina> aleth: have you seen my about:memory pastebin?
15:32:31 <aleth> Yes
15:33:22 <flo-retina> when I see:
15:33:22 <flo-retina> "0.25 MB (00.05%) ── string(length=65729, ":adams.freenode.net 322 fqueze ...")
15:33:22 <flo-retina> 0.25 MB (00.05%) ── string(length=73176, ":hitchcock.freenode.net 322 fqu...")"
15:33:23 <flo-retina> Am I right in thinking that adams and hitchcock are the name of the servers I was connected to at the time we sent LIST; which means the old results have been kept around?
15:33:54 <aleth> I'm puzzled because those strings are so very long
15:34:14 <flo-retina> isn't that the size we would expect when the server sends us LIST results?
15:34:35 <aleth> 65K single strings?
15:34:53 <flo-retina> that may be the size in the socket's buffer
15:35:01 <aleth> Hmm, could be.
15:35:23 <aleth> It does looks like a leak, the question is where :-/
15:36:40 <flo-retina> it seems the only code which clears the list is: http://lxr.instantbird.org/instantbird/source/chat/protocols/irc/irc.js#945
15:36:54 <flo-retina> so maybe if (this.isRoomInfoStale && !this._pendingList) { is never true for some reason?
15:37:07 <flo-retina> hmm, except we wouldn't send LIST again in that case :-S
15:37:51 <aleth> I should connect to freenode and see if I can reproduce...
15:38:15 <flo-retina> so I guess you would need to connect/disconnect a bunch of times
15:39:19 <flo-retina> about:memory now says "623.74 MB (100.0%) -- explicit"
15:39:25 <flo-retina> that's large for an IM client ;)
15:39:35 <aleth> just a bit ;)
15:39:48 <flo-retina> there are at least 4k unread messages in convs on hold
15:42:37 <aleth> What are your debug log settings?
15:42:48 <aleth> Do you keep infinitely long debug logs?
15:43:20 <flo-retina> messenger.accounts.maxDebugMessages;2000
15:43:32 <flo-retina> so no
15:43:59 <aleth> Asking because those strings seem to be unparsed IRC messages (so LIST may be incidental to the bug)
15:44:00 <flo-retina> and I think even with infinite debug logs, we clear them when an account is cleanly disconnected
15:45:10 <flo-retina> aleth: one thing to know about JS strings is that if you have a string saying "blahfoobar" and you create new sub strings containing "blah" and "bar", they will share the same string buffer, so "blahfoobar" will stay in memory even if you don't keep a reference to it.
15:45:31 <flo-retina> I think that's what happens here
15:46:02 <aleth> So we have to ensure the only thing we keep are /copies/ of substrings?
15:46:13 <flo-retina> so the leak we are looking for could be either at the socket/debug log level, or the list of channels in the irc prpl, or the stats service or awesometab keeping references to possible conversation objects
15:46:56 <flo-retina> aleth: not really. The garbage collector can reorder/cleanup. If you keep references to only tiny fractions of a large string, it can create new string buffers for these tiny strings, and free the large buffer
15:47:15 <aleth> But it looks like it didn't do that in your case?
15:48:41 <flo-retina> we don't know which strings we are keeping references to
15:48:52 <flo-retina> but if it's the topics, it's likely large enough that keeping the whole buffer makes sense
15:49:30 <flo-retina> (initially I suspected a bug of the JS engine not cleaning up large string buffers when it should; it's only after noticing that the server names for freenode LIST results isn't always the same that I started thinking _we_ are leaking stuff)
15:53:19 <aleth> Am I right in thinking that running xpcshell tests via mach fails because c-c and not m-c?
15:53:38 <flo-retina> pastebin ?
15:55:18 <aleth> http://pastebin.instantbird.com/544807
15:55:51 <flo-retina> try |mozilla/mach xpcshell-test chat|
15:56:25 <aleth> No change.
15:57:56 <flo-retina> I have the same error
15:59:17 <-- aleth has quit (Quit: exit stage left)
15:59:26 --> aleth has joined #instantbird
15:59:27 * ChanServ sets mode +h aleth 
15:59:50 <-- aleth has quit (Quit: exit stage left)
15:59:58 --> aleth has joined #instantbird
15:59:58 * ChanServ sets mode +h aleth 
16:00:04 <flo-retina> make xpcshell-tests -C obj-instantbird/ doesn't work either (different error)
16:39:05 <aleth> flo-retina: If you turn off wifi so you have no connection to the internet, and start IB, do you also see lots of connection attempts (each one failing and blaming the server)?
16:39:18 <aleth> This is a bug imho, but probably not an IB bug.
16:39:46 * flo-retina isn't very interested in turning off the wifi right now
16:39:57 <aleth> :D
16:48:08 --> tux has joined #instantbird
16:48:48 <-- tux has quit (Quit: ChatZilla [Firefox 26.0/20131206152142])
17:02:28 <flo-retina> aleth: why the statusNames -> statusAttributes change?
17:02:55 <aleth> Because it confused me that a toAttribute method returned a statusName.
17:03:21 <aleth> I don't think the term statusName is used anywhere else.
17:05:41 <flo-retina> ok
17:06:11 <-- Tonnes has quit (Connection reset by peer)
17:07:17 --> Tonnes has joined #instantbird
17:08:12 <aleth> Trying to understand a bit the WIPs in bug 95509, are the remaining parts of the makefiles supposed to go away in the end too (and be generated from moz.build files)?
17:10:59 <flo-retina> depends if "in the end" in your sentence means "once the bug is resolved" or "once the mozilla build peer are done recreating a faster build system"
17:12:02 <aleth> OK
17:12:21 * aleth wonders which makefile is run first - the moz.build-generated one or the explicit one
17:12:41 <flo-retina> the moz.build generated one seems to run when we include rules.mk
17:24:39 --> FireFly_TB has joined #instantbird
17:26:51 <-- FireFly_TB has quit (Input/output error)
17:26:53 --> FireFly_TB has joined #instantbird
17:47:04 <-- aleth has quit (Quit: exit stage left)
18:26:59 <-- Suiseiseki has quit (Ping timeout)
18:30:39 --> Suiseiseki has joined #instantbird
18:44:41 --> rosonline has joined #instantbird
19:05:56 <-- nathanr has quit (Ping timeout)
19:06:38 --> nathanr has joined #instantbird
19:08:23 <-- nathanr has quit (Connection reset by peer)
19:10:53 --> nathanr has joined #instantbird
19:33:15 --> EionRobb has joined #instantbird
19:35:06 <-- FireFly_TB has quit (Ping timeout)
19:42:44 <flo-retina> blogging software are such a pain :(
19:43:59 <-- flo-retina has quit (Quit: Instantbird 1.6a1pre -- http://www.instantbird.com)
19:52:00 --> jb has joined #instantbird
20:05:23 --> Guido has joined #instantbird
20:06:28 <Guido> Just a comment: I liked the old status notification more. (the design and that they are not on top of each other)
20:41:22 --> mrphs_ has joined #instantbird
20:41:47 <-- mrphs has quit (Ping timeout)
21:11:52 --> wnayes has joined #instantbird
21:46:50 <-- Guido has quit (Quit: Instantbird 1.5 -- http://www.instantbird.com)
21:49:57 <-- gerard-majax_ has quit (Ping timeout)
21:59:08 --> mconley has joined #instantbird
22:06:51 --> Mook has joined #instantbird
23:04:06 <-- jb has quit (Ping timeout)
23:12:17 <-- mconley has quit (Input/output error)
23:21:15 --> mconley has joined #instantbird
23:28:04 <-- hadi has quit (Quit: Instantbird 1.5 -- http://www.instantbird.com)
23:28:21 <-- rosonline has quit (Quit: Instantbird 1.5 -- http://www.instantbird.com)
23:38:08 <-- dew has quit (Quit: Instantbird 1.5 -- http://www.instantbird.com)
23:38:09 <-- mconley has quit (Input/output error)