#instantbird log on 06 09 2012

All times are UTC.

00:02:15 --> Suiseiseki has joined #instantbird
00:11:07 <flo> Good night :)
00:12:40 * flo wonders if anybody uses the "/action" command instead of "/me"
00:12:55 <flo> it's in my way when I type /a<tab> to have /away ;)
02:59:06 <-- devfil has quit (Client exited)
03:00:03 <instant-buildbot> build #521 of linux-nightly-default is complete: Success [build successful]  Build details are at http://buildbot.instantbird.org/builders/linux-nightly-default/builds/521
05:03:55 <instant-buildbot> build #614 of win32-nightly-default is complete: Success [build successful]  Build details are at http://buildbot.instantbird.org/builders/win32-nightly-default/builds/614
05:53:03 <instant-buildbot> build #511 of macosx-nightly-default is complete: Success [build successful]  Build details are at http://buildbot.instantbird.org/builders/macosx-nightly-default/builds/511
06:34:46 --> TestFliege has joined #instantbird
06:35:06 <-- TestFliege has quit (Input/output error)
06:36:03 --> FireFly_TB has joined #instantbird
07:35:26 <-- Kaishi has quit (Quit: Kaishi)
07:35:30 --> Kaishi has joined #instantbird
07:35:45 <-- Kaishi has quit (Quit: Kaishi)
07:35:51 --> Kaishi has joined #instantbird
08:35:56 <-- FireFly_TB has quit (Ping timeout)
09:29:17 --> gerard-majax has joined #instantbird
09:37:02 --> Optimizer has joined #instantbird
10:24:21 <-- Optimizer has quit (Ping timeout)
10:27:47 --> Optimizer has joined #instantbird
11:02:54 <-- Optimizer has quit (Ping timeout)
11:18:15 --> Tomek has joined #instantbird
11:29:50 --> Mic has joined #instantbird
11:29:51 * ChanServ sets mode +h Mic 
11:31:01 --> Optimizer has joined #instantbird
11:32:00 <Mic> http://log.bezut.info/instantbird/today#m3 -> New Bug: Completion should learn which commands a user prefers and suggest these first? ;)
11:33:03 <Mic> (and I'm sooo going to ignore comments about being too clever today;)
11:47:38 <flo> Mic: these are the "active" commands, aren't they? :)
11:50:27 --> igorko has joined #instantbird
11:57:02 <-- gerard-majax has quit (Ping timeout)
11:59:38 <-- Even1 has quit (Quit: Instantbird 1.2a1pre -- http://www.instantbird.com)
11:59:42 --> Even1 has joined #instantbird
12:00:02 <-- Even1 has quit (Quit: Instantbird 1.2a1pre -- http://www.instantbird.com)
12:00:08 --> Even1 has joined #instantbird
12:00:20 <-- Even1 has quit (Quit: Instantbird 1.2a1pre -- http://www.instantbird.com)
12:00:25 --> Even1 has joined #instantbird
12:17:26 --> devfil has joined #instantbird
12:17:53 <-- Mic has quit (Ping timeout)
12:24:50 --> gerard-majax has joined #instantbird
12:27:54 <-- Optimizer has quit (Ping timeout)
12:28:46 --> FireFly_TB has joined #instantbird
12:31:16 --> Optimizer has joined #instantbird
12:47:52 --> Mic has joined #instantbird
12:47:52 * ChanServ sets mode +h Mic 
12:49:39 <-- igorko has quit (Connection reset by peer)
12:51:29 <Mic> "We need a BCI (brain computer interface) to make the computer do what we want." ->
12:51:40 <Mic> "Keep on dreaming" ;)
12:52:09 * Mic participated in a BCI study today and it's surprisingly hard to move a cursor around with just your brains ;)
12:54:30 <flo> why would you still need a cursor?
12:55:37 <Mic> Oh, it was not exactly a cursor. Rather a cross that should be moved in one direction or the other (no actual computer ui involved).
12:58:56 <Mic> EEG/NIRS/fMRI are so limited methods :(
12:59:44 <Mic> It's mostly like trying to figure out how a processor works by putting a thermometer to the die to tell which part is currently using more or less energy :D
13:21:58 --> Felix has joined #instantbird
13:22:19 <-- EionRobb has quit (Quit: Leaving.)
13:22:44 --> EionRobb has joined #instantbird
13:24:26 <-- EionRobb has quit (Ping timeout)
13:24:51 <-- Felix has quit (Ping timeout)
14:00:09 <-- Mic has quit (Quit: Instantbird 1.2a1pre -- http://www.instantbird.com)
14:53:31 <aleth> what, minitel is being shut down?
14:53:50 <aleth> I'm kind of impressed it still existed
14:54:43 <-- Optimizer has quit (Ping timeout)
14:58:52 --> Optimizer has joined #instantbird
15:16:09 <-- Optimizer has quit (Ping timeout)
15:40:27 <-- chrisccoulson has quit (Ping timeout)
15:56:59 --> Optimizer has joined #instantbird
16:06:48 <-- Even1 has quit (Quit: Instantbird 1.2a1pre -- http://www.instantbird.com)
16:11:53 <-- FireFly_TB has quit (Ping timeout)
16:16:53 <flo> it's sad it won't work any more
16:17:11 <flo> we have at least 2 of them here, and one is probably still connected
16:17:31 <flo> (and it's never been quite clear to me how the second one appeared in the garage a few years ago :-S)
16:17:59 <-- Optimizer has quit (Ping timeout)
16:23:03 <flo> it's on a minitel that I learned to type when I was 2 years old, I knew how to type "Florian" on a minitel keyboard way before I was able to write it with a pen.
16:23:33 <flo> Those small and very resistant keyboards were great to let children play with them
16:24:55 --> Optimizer has joined #instantbird
16:25:44 <aleth> It can't have cost much to keep running... maybe people had just stopped using it
16:26:28 <aleth> You learned to type before you learned to write? natural born coder :P
16:27:48 <flo> "natural born coder" possibly ;)
16:28:09 <aleth> Actually also before I went to school, I used to type on an ancient mechanical typewriter my parents still had... it had a satisfying sound, especially the bell when you started a new line
16:29:08 <aleth> and when you pressed more than one key at once you could make them jam together ;)
16:31:40 --> igorko has joined #instantbird
16:32:00 <-- Optimizer has quit (Ping timeout)
16:33:03 <flo> aleth: hey, my father also had (and likely still has) an old mechanical typewriter in the attic; I also played with it :)
16:36:13 --> Optimizer has joined #instantbird
16:41:41 --> Lalae has joined #instantbird
16:44:36 <-- Lalae has left #instantbird ()
16:53:22 <aleth> Aha, FF15 now updates silently in the background (though nothing changes until you restart)
17:05:45 --> mmkmou has joined #instantbird
17:43:53 <-- gerard-majax has quit (Ping timeout)
18:08:33 --> DGMurdockIII has joined #instantbird
18:11:18 --> sonny has joined #instantbird
18:40:28 --> FireFly_TB has joined #instantbird
18:42:14 <-- FireFly_TB has quit (Ping timeout)
19:56:36 <-- meh has quit (Ping timeout)
19:58:58 --> meh has joined #instantbird
20:02:44 <aleth> "/action is the refuge of people who have nothing whatsoever to do." Oscar Wilde (kind of)
20:10:22 <-- Optimizer has quit (Ping timeout)
20:14:19 --> Optimizer has joined #instantbird
20:16:53 <flo> aleth: :)
20:17:08 <flo> aleth: is profiling memory usage another refuge for these people? ;)
20:17:59 <flo> I'm wondering how come Instantbird takes 160MB of resident memory, for only 71MB of "explicit" in about:memory.
20:18:22 <flo> I've just emailed someone from memshrink to hopefully have clarifications on what these numbers actually mean
20:19:07 <-- igorko has quit (Connection reset by peer)
20:22:05 * flo ponders profiling a debug build with his default profile on massif
20:37:03 <-- Optimizer has quit (Ping timeout)
20:38:43 --> Optimizer has joined #instantbird
20:40:26 <aleth> Will be interesting if you get some answers :)
20:40:41 <-- sonny has quit (Client exited)
20:41:00 <aleth> Aren't there some guys writing a built-in JS profiler for FF? How hard would it be to turn that on for IB?
20:43:06 <flo> that's for finding performance issues / excessive CPU usage
20:43:12 <flo> I'm after excessive memory usage right now ;)
20:43:30 <flo> but... once it works in Tb, I think it will be trivial to adapt for Ib ;)
20:43:32 <aleth> Oh, it doesn't do memory? oh well...
20:43:40 <aleth> It would be nice to get the JS debugger too :)
20:44:11 * flo retries using massif on his debug linux build after installing the glib debug symbols
20:45:28 <instantbot> aletheia2@fastmail.fm cancelled review?(florian@instantbird .org) for attachment 1499 on bug 1450.
20:45:29 <instantbot> aletheia2@fastmail.fm requested review from florian@instantbird .org for attachment 1581 on bug 1450.
20:45:30 <instantbot> Bug https://bugzilla.instantbird.org/show_bug.cgi?id=1450 nor, --, ---, aletheia2, ASSI, Only remove unread ruler when switching away from a tab if tab has been visible for a certain time
20:50:07 <flo> I don't understand why I don't see any libpurple allocation in my massif profile :(
20:50:22 <flo> I can't believe libpurple is using less than the threshold of 0.1% of the memory the application uses
20:55:41 <aleth> Strange. Probably listed as something misleading?
20:58:38 <-- mmkmou has quit (Ping timeout)
20:59:02 <flo> if I remove the --pages-as-heap=yes command line parameter, I do have some libpurple stuff in there :)
20:59:32 <flo> I probably won't have accurate measurements of what Mozilla does for JS arenas, but I don't think that matters much :)
21:00:35 <aleth> not when you're trying to track down 90M... ;)
21:01:53 --> mmkmou has joined #instantbird
21:03:38 <flo> now I've got 23% of the memory used by the spellchecker, some JS stuff, 18% ignored because in 3000+ different allocation stacks all of them under the 0.5% threshold, 5% in xpt, lots of things I don't understand
21:03:48 <flo> g_slice_alloc is at 2.97%
21:03:53 --> Lalae has joined #instantbird
21:04:12 <flo> purple_status_new takes 0.77% of the memory of the whole application
21:04:17 <-- Lalae has left #instantbird ()
21:04:40 <flo> not a surprising discovery, it's been known for a while that libpurple allocates status objects for all buddies, even when they are offline
21:04:53 <flo> not very interesting either; it's a waste, but doesn't grow over time
21:05:17 <flo> ~221kB of wasted memory
21:05:27 <aleth> not that relevant
21:05:38 <aleth> The spellchecker uses more memory than all of my IB? Hmm... I'm going to turn it on and restart to see if I can reproduce
21:06:20 <flo> hmm, no it's 2.05% and 590kB actually
21:06:57 <flo> aleth: mesuring the memory usage almost immediately after startup is almost completely uninteresting in the results it gives ;)
21:07:08 <flo> I'm just checking that the results I see at this point are plausible
21:07:24 <aleth> I didn't restart, just turned it on, so we'll see what happens.
21:07:25 <flo> no point in running with my default profile for a few hours if it's to produce completely meaningless results
21:08:19 --> Lalae has joined #instantbird
21:08:33 <aleth> I havent restarted IB in over 48 hrs I think and its using 52M. 3 channels and a lot of past activity, now closed
21:08:56 <flo> it's possible that my high memory usage comes from a leaky prpl that you don't use :)
21:09:03 <flo> netsoul maybe? :)
21:09:08 <aleth> Wouldn't surprise me.
21:09:46 <flo> or msn? :)
21:09:56 <aleth> I have that too.
21:10:14 <flo> I have 3 connected accounts of it
21:10:17 <aleth> Though I suspect turning it off would make IB downright sleek...
21:10:40 <-- Lalae has left #instantbird ()
21:15:21 <Optimizer> hey am I still connected ?
21:15:49 <aleth> Optimizer: looks like it
21:16:13 <Optimizer> oh yes, sorry to disturb, but sometimes I get disconnected, and I come to know only when I try to ping something
21:16:34 <flo> Optimizer: you can see that in instantbot's logs ;)
21:19:30 --> Mic has joined #instantbird
21:19:30 * ChanServ sets mode +h Mic 
21:19:56 <flo> hmm, I've created a new profile on my linux machine and copied there my prefs.js file, hoping that would let me connect all my accounts
21:20:10 <flo> but they all fail with "Entering a password is required to connect this account."
21:20:25 <flo> is there any obvious reason for that?
21:20:59 <flo> afaik even though we now write the passwords in the password manager, they haven't been removed from the preferences files, and the migration should still work
21:21:13 <Mic> flo: if your answer about the terms on about:memory isn't answered yet: areweslimyet has an FAQ that explains some if I recall correctly.
21:22:16 <Optimizer> flo: yes I can see, but this kind of thing should not happen, right ?
21:22:38 <Optimizer> if I am disconnected, it should tell me directly, and not when I try to ping someone
21:23:06 <Optimizer> sometimes this difference increases to 1 minute , and them I miss all the conversations
21:23:23 <Mic> https://areweslimyet.com/faq.htm#whats-the-difference-between-resident-memory-and-explicit
21:25:49 <flo> Mic: the faq answered some of my questions
21:25:52 <flo> thanks
21:25:57 <Mic> :)
21:26:22 <flo> I still don't know if libpurple's allocations are supposed to be counted in explicit or not though
21:26:55 <flo> if they are in explicit, then they are in heap-unclassified, and we have a terrible memory fragmentation issue
21:27:18 <flo> if they are not, then a leak in a prpl I use could explain why I see such a large difference between explicit and resident
21:28:31 <flo> ah, I need to flip a pref so that we convert passwords again
21:29:03 <aleth> Why wouldn't it be in explicit? Because it uses it's own malloc? Seems unlikely...
21:29:36 <flo> I suspect explicit would be whatever's linked to mozilla's jemalloc, and things allocated through the system allocator wouldn't be counted
21:30:19 --> flo1 has joined #instantbird
21:32:38 <-- flo1 has quit (Quit: Instantbird 1.2a1pre -- http://www.instantbird.com)
21:32:56 <-- flo has quit (Quit: Instantbird 1.2a1pre -- http://www.instantbird.com)
21:35:15 --> flo has joined #instantbird
21:35:15 * ChanServ sets mode +qo flo flo 
21:36:27 <-- mmkmou has quit (Quit: Instantbird 1.2a1pre -- http://www.instantbird.com)
21:37:46 <-- flo has quit (Ping timeout)
21:38:11 <-- Mic has quit (Quit: Instantbird 1.2a1pre -- http://www.instantbird.com)
21:39:28 <aleth> Optimizer: It should indeed tell you when you are disconnected.
21:39:56 <Optimizer> but sometimes it detects that it is disconnected only when i try to type in something
21:40:07 <Optimizer> but in reality it was disconnected a minute ago
21:40:17 <Optimizer> this minute lag leads to loss of conversation
21:40:28 <aleth> Do you have a really bad connection?
21:41:24 <aleth> Sounds like timeouts may be involved
21:41:45 --> flo has joined #instantbird
21:41:46 <aleth> It's possible some patches in the pipeline may help a bit, but I'm not sure. It depends on what's causing it
21:41:46 * ChanServ sets mode +qo flo flo 
21:43:00 <Optimizer> I have 1mbps connection
21:43:14 <Optimizer> though it is average i my area, its bad for the world
21:43:15 <Optimizer> :|
21:44:14 <flo> it's too easy to ping timeout when running with valgrind and connecting 12 accounts at once
21:44:28 <aleth> Optimizer: You could turn on logging and maybe you could find out what is going on
21:44:42 <Optimizer> how ?
21:45:10 <aleth> and bug 1499 may help a little soon
21:45:13 <instantbot> Bug https://bugzilla.instantbird.org/show_bug.cgi?id=1499 nor, --, ---, nobody, ASSI, Avoid attempting to connect accounts while offline, and disconnect connecting accounts when going of
21:46:28 <flo> AIM, MSN, Netsoul, Gtalk, Facebook, IRC :)
21:46:34 <flo> I think they are all connected, finally )
21:46:35 <aleth> Optimizer: purple.debug.loglevel for starters
21:46:44 <aleth> :)
21:47:19 <aleth> Optimizer: For more you would need a debug build I think
21:47:49 <Optimizer> why pruple ?
21:47:51 <Optimizer> :D
21:48:11 <flo> for historical reasons :)
21:48:14 <aleth> It should be prpl, but it's not, for historic reason ;)
21:48:46 <Optimizer> what should be teh log level ?
21:52:25 <aleth> 3 is normal, 2 is noiser, 1 is noisiest I think
21:54:54 <flo> it currently takes 800MB, but I believe most of it is the overhead caused by massif, so that value isn't really relelvant
21:56:27 <-- Optimizer has quit (Quit: Instantbird 1.2a1pre -- http://www.instantbird.com)
22:05:11 <flo> the waste of memory I was mentioning around libpurple's status is covered by http://developer.pidgin.im/ticket/14290
22:06:04 <flo> I hope I'll find some interesting massif output tomorrow. Good night :)
23:00:10 <-- skeledrew has quit (Ping timeout)
23:00:11 --> skeledrew has joined #instantbird
23:14:24 --> mmkmou has joined #instantbird
23:34:28 <-- DGMurdockIII has quit (Quit: ChatZilla [Firefox 14.0/20120605113340])