[00:17] <RAOF> hicham: Yup.
[00:17] <hicham> RAOF: what is is it ?
[00:18] <hicham> RAOF: indicator-network if for conmann
[00:18] <hicham> *is
[00:18] <RAOF> I think we patch the nm-applet source to add an indicator.
[00:18] <hicham> oh
[00:18] <hicham> thanks
[00:28] <hicham> what about programs that use the notification area ?
[00:28] <hicham> they need to be patched as well ?
[00:32] <RAOF> There is a notification area in Unity, but it's protected by a whitelist.  Wine is on that whitelist, as I think is skype.
[00:32] <RAOF> Others get to use indicators instead :)
[00:32] <hicham> how bad
[00:33] <RAOF> I presume it'd be relatively easy to remove the whitelist; then you'd have a notification area and indicator area.
[00:34] <RAOF> Which wouldn't interact very well, which is why the notification area is restricted to a whitelist :)
[06:32] <DBO> RAOF, NOooooooooooooooooooooooooooooooooooooOOOOoooooooooooooooooOOOOOOooooooooooooooo
[06:33] <RAOF> DBO: We got *one* set of bugs, at least!
[06:34] <DBO> RAOF, I was thinking about it
[06:34] <DBO> and the only one that doesn't make sense is the dri2 freeze
[06:34] <DBO> thats on the same display as the grabs
[06:34] <DBO> it doesn't make sense for that to be an XGrabServer issue then
[06:34] <DBO> I have asked Chris to see if it still happens for him with vsync disabled
[06:35] <RAOF> I expect it won't.
[06:35] <DBO> right
[06:36] <DBO> so now we get to figure out why the hell vsync fails so hard sometimes
[06:36] <RAOF> I'm parsing his kernel logs to check that the kernel hasn't simply missed the vblank event.
[06:37] <RAOF> 250MiB of kernel log is *fun* to wade through!
[06:42] <RAOF> Wow.  We really waited the *hell* out of from 3788 - we waited for it no fewer than 7 times!
[06:42] <RAOF> s/from/frame/
[07:15] <RAOF> DBO: Suspicions!
[07:15] <DBO> they abound
[07:16] <RAOF> It looks like drm_queue_vblank_event might be broken if you call it at a particularly inopportune time.
[07:17] <DBO> is that our fault?
[07:17] <RAOF> No.
[07:17] <RAOF> Looks like it may be a kernel problem.
[07:19] <RAOF> Oh, maybe not :/
[07:20] <RAOF> Well, suspicions remain, but it's not what I thought.
[07:55] <didrocks> good morning
[08:04] <oSoMoN> good morning
[08:05] <MacSlow> greetings everybody
[09:01] <hicham> didrocks : morning
[09:02] <hicham> didrocks: session, datetime, and sound indicators run ok here
[09:02] <didrocks> hey hicham
[09:02] <didrocks> nice :)
[11:43] <om26er> kamstrup, ping!
[11:44] <om26er> kamstrup, seems like bug 764352 reporter is telling the truth, so does the issue go into unity-places or is it a unity one?
[11:46] <kamstrup> om26er: I have a hunch as to why that could happen - i'll add a unity-place-files task for it. Thank's for the heads up
[11:47] <om26er> yw
[13:43] <om26er_> didrocks, Hi! any words on bug 754327 ?
[13:44] <didrocks> om26er_: not a priority anyway, wishlist for O I would say
[13:44] <om26er_> ok
[13:57] <davidcalle> Hi all. njpatel or didrocks, is there any way to kill a lens service? It keeps respawning.
[14:01] <davidcalle> Nervermind, I did it.
[14:04] <didrocks> davidcalle: you can probably chmod it temporarly
[14:04] <davidcalle> didrocks, I just commented out the exec line in the .service.
[14:04] <didrocks> works too :)
[15:37] <jcastro> kenvandine:  njpatel: so did we ever find the problem in the python lens race login mystery?
[15:38] <kenvandine> jcastro, not sure, but the AU lens is working for me now
[15:41] <davidcalle> jcastro, kenvandine, for me, it's when I add the AU lens that all others fail to load at times.
[15:42] <kenvandine> davidcalle, that sounds scary
[15:42] <kenvandine> davidcalle, but i can kind of confirm that
[15:43] <kenvandine> last week, the books lens wouldn't work either, at the same time the AU lens didn't
[15:43] <kenvandine> the gwibber lens kept working though, which makes it more confusing
[15:45] <jcastro> kenvandine: right, so depending on my luck either one, both, or none of them would work
[15:46] <didrocks> om26er: there is is no opened bug on the first icon being displayed as if we were in a keynav state, isn't it?
[15:47] <davidcalle> kenvandine, it seems to be related to loading things via web APIs. That would explain why the gwibber lens keeps working.
[15:49] <om26er> didrocks, hm I dont think so
[15:49] <om26er> on the other note unity just crashed on clicking empty trash :p
[15:51] <didrocks> om26er: on trunk?
[15:52] <om26er> official versions didrocks
[15:52] <didrocks> om26er: ok, I changed that part of code since Friday, should be safer now :)
[15:55] <marcobiscaro2112> om26er: this is bug #761643
[15:55] <om26er> marcobiscaro2112, thx :)
[15:56] <didrocks> hey marcobiscaro2112 :)
[15:56] <marcobiscaro2112> hi didrocks !
[15:57]  * davidcalle is happy, has found a way to deal with slow API in lenses.
[15:59] <didrocks> om26er: I've just reproduce bug #763791 once, couldn't at all then, do you have any clue?
[16:00] <jcastro> kenvandine: ok so as you login from now on can you keep an eye on the AU lens?
[16:00] <om26er> didrocks, its duplicate
[16:01] <didrocks> om26er: oh really? didn't notice
[16:01] <om26er> didrocks, assigned to sam ;)
[16:01] <kenvandine> jcastro, will do
[16:01]  * om26er finds the duplicate
[16:02] <om26er> didrocks, bug 760436
[16:04] <didrocks> om26er: no, it's no a duplicate
[16:04] <didrocks> om26er: here, the mouse/keyboard input are correct
[16:04] <didrocks> bug #760436 is an old bug I could reproduce at will with chromium some time ago
[16:05] <om26er> both aren't the same
[16:05] <om26er> the screenshot seems same
[16:05] <didrocks> yeah, but to have experienced it once, I'm sure the input window was matching what we see
[16:05] <didrocks> (I see it's a race in our undecorate function)
[16:06] <om26er> didrocks, bug 727060 was also the old bug you reported
[16:06] <om26er> wasn't that also related?
[16:06] <didrocks> om26er: so, this one is indeed related to bug 760436
[16:06] <didrocks> where the input is shifter
[16:06] <didrocks> shifted*
[16:07] <didrocks> on the other bug (I unduped it), the focus/input is fine
[16:07] <didrocks> it's just that we undecorated on maximize after compiz set the size of the window
[16:07] <didrocks> (but the input corresponds to what is drawn)
[16:08] <om26er> hmm hmm
[16:10] <om26er> comment#9 in bug 760436 sounds similar to bug 763791 description
[16:11] <om26er> I guess once one of them is fixed we'll know if the other is the same or not ;)
[16:20] <jcastro> DBO: I don't know how to quite reproduce this
[16:21] <jcastro> but every once in a while a launcher tooltip will just hang around
[16:21] <DBO> yeah I have heard about this
[16:25] <didrocks> jcastro: do you wrong keynav color appearing as well? (or sometimes, quicklist appearing)
[16:26] <jcastro> sometimes I get a white glow on my top icon for seemingly no reason
[16:26] <jcastro> haven't seen a quicklist yet though
[16:27] <didrocks> jcastro: I got the white glow (and quicklist sometimes) just down, but the tooltip (I saw that too), doesn't seem to be related :/
[16:27] <jcastro> right, they don't seem to be
[16:27] <jcastro> I've been getting the tooltip off and on for a long time
[16:27] <jcastro> the glow thing is only recently
[16:27] <didrocks> right
[16:28] <didrocks> it's due to a compiz change
[16:28] <didrocks> the tooltip, let me have a look if I can see anything obvious in the code…
[16:39] <davidcalle> kenvandine, I got some news on the lenses. When a 1 or 2 sec delay is introduced before the daemon loop (to preload some stuff), it fails to register at unity startup. That's what Stefano is doing and if I try the same thing, my lens fails at statup too. That's quite logical, but here is a confirmation.
[16:43] <kenvandine> davidcalle, ah... interesting
[16:58] <jcastro> didrocks: from the apps lens, try to drag an application to the desktop, do you get some dumb error dialog?
[16:59] <didrocks> jcastro: there was a crash which has been fixed, by not dumb error dialog though
[16:59] <didrocks> jcastro: is it displayed by nautilus?
[16:59] <didrocks> like, you can click through the launcher and so the dnd happened in nautilus?
[17:01] <jcastro> There was an error getting information about "/".
[17:01] <jcastro> The specified location is not supported
[17:01] <jcastro> looks like a nautilus window
[17:03] <didrocks> jcastro: yeah, so it seems that you can click through the launcher
[17:03] <didrocks> DBO: in case you didn't notice, we still have some cases of that issue ^^
[17:03] <didrocks> (particularly on edge reveal)
[17:03] <DBO> yeah I saw that
[17:03] <DBO> its annoying me deeply
[17:03] <didrocks> so, the reveal happens, you put the mouse over the launcher and nothing happens, so it hides again
[17:03] <DBO> its not click through actually
[17:04] <didrocks> I thought we killed it by not spamming too much X…
[17:04] <DBO> you *can* click on something if you click before it hides
[17:04] <DBO> its actually a hide state issue
[17:04] <didrocks> DBO: hum really? I have no mouse over tooltip appeareance for instance
[17:04] <DBO> yep
[17:04] <didrocks> DBO: I tried to keep super pressed, hovering with the mouse
[17:04] <DBO> thats part of the weirdness
[17:05] <DBO> i am not sure what causes it yet
[17:05] <didrocks> no tooltip, nothing… click was ending up in nautilus
[17:05] <DBO> had looked at it a bit this weekend...
[17:05] <didrocks> no reproducible test case though :/
[17:06] <DBO> yeah I get it sometimes
[17:06] <DBO> and then debugging is like impossible
[17:06] <didrocks> DBO: right. Hum, I'm in a weird state again, the leave event isn't received in the Launcher
[17:06] <DBO> i hate that
[17:07] <didrocks> the weird thing is that the LauncherIcon received it
[17:07] <didrocks> (no more tooltip)
[17:07] <DBO> I am afraid to fix that in nux for fear of triggering other bugs
[17:07] <didrocks> DBO: agreed
[17:07] <didrocks> DBO: anyway, the day we can fix this leave event stuff, we will remove a lot of hidemachine workaround
[17:08] <DBO> yeah
[18:11] <stefano-palazzo> kenvandine, are you around?
[18:13] <kenvandine> stefano-palazzo, hey
[18:13] <stefano-palazzo> hi :)
[18:13] <stefano-palazzo> maybe you can help me, I'm pretty flummoxed
[18:13] <stefano-palazzo> My lens is still not working on login, so, I replaced every print statement with a logging function,
[18:14] <stefano-palazzo> after logging in, it wrote 400 megabytes of logs
[18:14] <stefano-palazzo> what in the name of all that's holy could cause that do you know?
[18:14] <kenvandine> ugh!
[18:14] <kenvandine> what the hell!
[18:14] <kenvandine> what kind of stuff?
[18:14] <kenvandine> it is working at login for me
[18:15] <stefano-palazzo> it looks like the first few hundred megabytes (yeah..) are just telling me the program has been started, and that the imports have finished
[18:15] <kenvandine> stefano-palazzo, like is it looping and printing the same thing?
[18:16] <stefano-palazzo> kenvandine, towards the end, it looks like it's starting properly, but by that time my computer is completely overloaded anyway
[18:16] <stefano-palazzo> it does work for you? maybe I should just dist-upgrade, maybe it wasn't my fault at all? David said something about a long startup time before the gobject main loop
[18:16] <kenvandine> davidcalle noted this morning that if he adds a 1 or 2s delay before the daemon class it fails to load
[18:17] <kenvandine> stefano-palazzo, well it worked for me on this login :)
[18:17] <stefano-palazzo> right, that's what I was trying to tackle, I have this sort of delay just because I'm downloading stuff
[18:17] <kenvandine> but unity should be able to handle that
[18:17] <kenvandine> so if that is the case, it is a bug
[18:18] <kenvandine> however you should make sure you have the latest unity
[18:18] <stefano-palazzo> oh good,
[18:18] <kenvandine> i think there was some lazy loading stuff
[18:18] <stefano-palazzo> I'll take a break (trying to fix it with unity crashing left and right wasn't much fun), create a completely fresh usb disk and check again
[18:19] <kenvandine> ok
[18:19] <stefano-palazzo> I might bug you later (:
[18:19] <stefano-palazzo> thanks for your help
[18:19] <kenvandine> i had assumed the lazy loading in unity fixed it, which i just updated to this morning
[18:19] <kenvandine> anytime!
[18:19] <stefano-palazzo> oh yes that sounds very encouraging
[20:01] <stefano-palazzo> kenvandine, I've gotten rid of the startup-delay, but the problem is still there
[20:05] <kenvandine> stefano-palazzo, bummer
[20:05] <kenvandine> it is still working here
[20:05] <kenvandine> rebooted
[20:05] <stefano-palazzo> that is strange
[20:06] <stefano-palazzo> I've just been logging out, I'm going to reboot for good measure
[20:07] <nhaines> Are we getting another Unity release today?
[20:10] <stefano-palazzo> kenvandine, nope, still nothing. Do you have any idea how I might investigate this?
[20:10] <kenvandine> stefano-palazzo, dunno... i think we need kamstrup to help debug
[20:11] <kenvandine> stefano-palazzo, try this, start the daemon in a terminal
[20:11] <kenvandine> then "setsid unity"
[20:11] <kenvandine> and look at the console output
[20:11] <kenvandine> see if you see assertions, like DEE_IS_MODEL
[20:11] <kenvandine> on the sections model
[20:12] <kenvandine> stefano-palazzo, when i was debugging this the other day, i was consistently getting those assertions that the sections model wasn't actually a dee model
[20:12] <kenvandine> which made me suspect an offset problem
[20:12] <kenvandine> but i couldn't see where
[20:13] <stefano-palazzo> kenvandine, mh. I don't get any of those - but neither do I get any search results
[20:13] <stefano-palazzo> it was working a minute ago
[20:14] <kenvandine> i was getting that when it wouldn't work at all
[20:14] <stefano-palazzo> then, when I try to ^C it, it takes five interrupts to quit -
[20:14] <stefano-palazzo> (which sometimes happens with threaded code, in my experience)
[20:16] <stefano-palazzo> mh, the traceback is all socket stuff, it's probably the only the server not responding
[20:36] <ronoc> kenvandine, I think you can go ahead and patch that
[20:36] <kenvandine> ok
[20:36]  * kenvandine will do
[21:31] <Omega> Compiz is using 1.8 gigs of RAM.