[00:04] RAOF: You're probably just used to autotools. [00:04] I'm also used to being able to do an in-tree build :P [00:04] heh [00:05] But make distcheck does out of tree builds. :) [00:06] For convenience, yeah. But it's generally happy to do in-tree builds! [00:09] yeah [00:11] Well, unity's building and the espresso machine has warmed up. Time to make coffee! [00:12] heh === jono_ is now known as jono [00:59] What do you guys think about the recent removal of the Maximize and Minimize buttons in GNOME Shell? [01:00] It's good for them to experiment as long as they put it back. ;) [01:00] Personally, I think they might be on to something. [02:10] Ok. I'm now freaked out. [02:26] RAOF, me too, but that's probably because it's nearly 2.30 am ;) [02:29] Heh. [02:29] I'm freaked out by something suddenly being null when I want to relayout the shell. [02:30] Null is nothing, so no problem, right? [02:33] It's getting set up correctly at init time, and I can't see where it's getting set anywhere else, so I'm mystigofied. [03:32] Damn you unity folks and your flagrant disregard for method signatures! [03:46] Ok. Well, that now kinda works. [03:52] Hey, two quick design questions: is the name “Me menu” strictly a development code name, or can I use that name for users, too? [03:53] Also, do we have a rule for capitalizing these, like “Messaging Menu?” [07:17] Oh well, I have to turn in for the night. I took a guess that it's how I wrote it already (“Me Menu” and “messaging menu”). Hopefully it's right. Will try to clarify tomorrow. Good night / day / whichever! :) [07:24] I loaded Ubuntu 10.4 Netbook edition and hated the Unity interface. [07:24] The reason is that: [07:24] - a lot of space is wasted by permanent icons you can't conceal [07:25] - every panel is maximised, to take up the entire space on the screen [07:25] - there are no minimise, maximise and close buttons on each panel [07:26] - the minimise, maximise and close functions are moved up to a panel at the top of the screen, that is muddy and difficult to see. [07:27] - the "place markers" for each panel don't have pop-out legends, individual icons or explanatory texts [07:27] So I wiped it and installed 10.4 Desktop instead. [07:30] I hope that the Unity interface is only offered as an option - the xfce interface works perfectly on my Acer Aspire One. [07:31] Of course, I would like to express my admiration and gratitude for all the work that Canonical and others have put into Ubuntu. [07:31] End of message = [07:37] good morning [08:05] DBO: By the way, I hate you and your broken unity/nux multihead code :). It kinda sorta ish works here, though. [08:06] RAOF, what multihead code? [08:06] we never did any multihead code [08:06] Ok. The lack of multihead code, then. [08:07] RAOF, expect to get double renders of nux on the left monitor [08:07] dont try to fix that [08:07] I know what the problem is [08:07] What would double-renders look like? [08:07] transparency is suddenly less transparent [08:07] Ah. [08:07] why? [08:07] Well, I don't have anything transparent on my left monitor. [08:08] RAOF: the version in natty is "broken" in the way we forcec double transparency :) [08:08] force* [08:08] wherever the launcher is [08:08] like 0.8 instead of 0.9 for the dash [08:08] Apart from the launcher's tooltips, which are where they would be if the launcher was on the left monitor ;) [08:08] (it's reverted in trunk, blame mr Patel :)) [08:08] wait he really did that? [08:08] what a tart... [08:08] DBO: yeah, he was thinking the value was inaccurate :) [08:08] RAOF, oh well then [08:09] DBO: so it was a last minute commit before alpha3 :) [08:09] didrocks, yeah I can fix that as soon as I get this event processing shit done [08:09] Nux almost handles resizing crtcs, too. [08:09] then saw it was because of the double painting in multihead [08:09] it's reverted in trunk :) [08:09] DBO: heh, you have time, next release is Monday [08:09] no I mean the double painting thing [08:09] right, I konw [08:09] know* [08:09] i think you overestimate how quickly I can rip out an entire event system of a toolkit [08:09] I mean I am good [08:09] DBO: just no need to hurry, you still have until Monday [08:09] but I am not jesus [08:10] heh :) [08:10] you are dude! [08:10] I've got the hair, he's got the beard :P [08:10] I imagine nux as a medium sized mammal [08:10] and what I am doing right now is akin to cutting it open with a chainsaw [08:10] DBO: if you can't get it rewrote for this release, aren't you afraid of ETOOMANYCONFLICTS when merging? [08:10] ripping out its spine [08:10] grabbing some GTK mammal spine, and sloshing it in there [08:11] good morning [08:11] didrocks, i am seriously considering bzr pushing to lp:nux and then cutting it off half way through [08:11] didrocks, jay wont figure that out for weeks [08:11] and then nux trunk is frozen! [08:11] DBO: please, don't break nux trunk ;) [08:11] its not broken [08:11] DBO: there are crashes to fix! [08:11] a lot of them :) [08:11] I'll just be holding the lock file [08:12] * didrocks fires bzr break-locks [08:12] or whatever it is :) [08:12] didrocks, so my plan to avoid too many merge issues is to insert the new spinal system for nux alongside its existing mechanisms [08:12] then just switch it over in WindowCompositor [08:13] if I only add code I can A) make your life a living hell [08:13] and B) avoid merge conflicts [08:13] DBO: waow, sounds more work though [08:13] So, I also discovered today that unity kinda-sorta has dynamic multihead code already hooked up, in that it watches the monitors-changed signal. [08:13] to ensure both system are working [08:13] RAOF, are you saying between the two of us we are Jesus? [08:14] RAOF, because I always thought of you as the romance novel anti-hero [08:14] RAOF: yeah, just came with that release (thanks cyphermox) [08:14] but the magic is really elementary though :/ [08:14] RAOF, i told you that [08:14] and introduce design issues [08:14] like, if you primary monitor is the right one [08:14] didrocks, i'll make it work [08:14] the launcher is there :) [08:14] thats crap [08:14] so, quite funny to not have any "resistance" :) [08:14] right [08:14] Yeah, but the monitors-changed signal triggers the setting of the needsRelayout variable, which then isn't used anywhere. [08:15] MacSlow, keynav is my problem from now on [08:15] that okay with you? [08:15] 09:14:03 didrocks | RAOF: yeah, just came with that release (thanks cyphermox) [08:15] s/thanks/blames [08:15] :) [08:15] DBO, hey there [08:15] I am ripping out all of nux's spinal system (ProcessEvents is dying a terrible death) and replacing it with something sane [08:15] DBO, yeah [08:15] Also, the signal is hooked up backwards, so if you actually tried to use the UnityScreen that it gets passed in you'll immediately SIGSEGV :) [08:16] RAOF, SIGSEGV is for manly men [08:16] manly men with chest hair [08:16] Actually… it's a bit worse. [08:16] * DBO tries to think of whats worse than sigsegv [08:16] deadlock? [08:16] DBO, I'm atm doing desperate things like starting timers upon opening ql with calling SetInputFocus() at the end in the hope to get it back working :) [08:16] You *won't* necessarily SIGSEGV, but you *will* set some byte of the GdkScreen to TRUE. [08:16] Who knows what that'll do later ;) [08:17] MacSlow, dont bother, I already implemented the global grab thing I talked about [08:17] MacSlow, your grabs were racing, now they dont [08:17] MacSlow, I am further implementing input grabs with nux [08:17] SetInputFocus is a worthless method for us now [08:17] :) [08:17] I hate to kill all your fun [08:17] * RAOF can push a branch with these fixes if you like. [08:17] DBO, but why only on the ql and not on the launcher... I still don't see that... the difference between launcher and ql [08:18] MacSlow, ql isn't mapped when you call SetInputFocus [08:18] you just *requested* it be mapped [08:18] but its not yet mapped [08:18] hence why the timer sometimes fixes the issue [08:18] again, leave this one to me [08:18] I got it from here on [08:18] morning MacSlow :) [08:19] RAOF, you sexy devil, lets run away together and live in a pine studded bungalow (thats what australians live in right?) [08:19] DBO, but I added the ability to hook up to MapNotify in GraphicsDisplay::GetSystemEvent() and ::ProcessXEvent() [08:19] yeah I saw that, it looked nasty... [08:19] DBO, and that didn't help much [08:19] salut didrocks btw :) [08:19] hey RAOF [08:19] Hey MacSlow :) [08:19] MacSlow, dont worry, we dont need it anymore, really :) [08:20] Going to complain about mesa at me? :) [08:20] RAOF, no... but do you want me to? :) [08:20] RAOF, I mean I can think of something I bet ;) [08:20] Not particularly :) [08:20] RAOF, if it wasn't clear, my last message to you roughly translated to "please upload your branch" [08:20] DBO: I'll clean it up and push. [08:20] RAOF: if MacSlow doesn't complain, can I complaing about my nvidia magnetic issue again? [08:21] RAOF, mesa is a terrible name, we should change its name in a distro patch [08:21] RAOF: I tried some vodoo without any chance [08:21] Obviously, it also needs nux to not suck. [08:21] But that's a one-liner. [08:21] didrocks, "magnetic issue" ?? [08:21] I was obviously really helpful :) [08:21] hopeful* [08:21] oh and hi there everyone [08:21] MacSlow: yeah, you don't want to see that [08:21] it's ugly :) [08:21] imagine that you monitor was still a lcd, with electrons [08:22] and you bring a magnetic next to it [08:22] didrocks, you mean CRT? [08:22] then, some part of the display is distored [08:22] yeah [08:22] that's what I have there :) [08:22] loicm has some too [08:22] didrocks, wow... you got that with and LCD? [08:22] only with unity (think that nux is guilty) [08:22] MacSlow: right… [08:22] compiz only don't get it [08:22] didrocks, spooky! [08:22] and the more "interesting" [08:23] is that I get it when I only have one display [08:23] if I use the twinview mode, no issue at all [08:23] (even on the screen having it) [08:23] didrocks, turn of lighting in compiz [08:23] off* [08:23] didrocks, uff... when did that happen first? [08:23] didrocks, i am 99% sure this is a compiz setting needing tweaking [08:23] I saw it once... [08:24] DBO, you mean GL-lighting so the cube looks nicer upon rotatino? [08:24] rotation [08:24] DBO: I think it's already turned off, let me check [08:24] MacSlow: mid-november IIRC [08:24] MacSlow, yeah that [08:24] it causes some issues with other shiz right now [08:25] seems nux messes up the lighting for compiz somehow [08:25] let me check lightning again [08:25] im not sure why to be honest [08:25] didrocks, if its not that start messing around with other settings [08:25] its something in there I am like 99% sure [08:25] I turned it off already [08:25] DBO: ok, will try some config change [08:26] 80% sure [08:40] didrocks, do you still get this issue LP: #727890 [08:44] MacSlow: when I added the ql items, I checked for that [08:44] it didn't happen [08:44] (tried 5 times to ensure) [08:44] has I was afraid of it [08:45] I think it's a race [08:45] didrocks, hm [08:46] as* [08:48] Man, bzr is awesome. “bzr lp-open” is love. [08:51] DBO: Eat hot lp:~raof/unity/multihead-fixes [08:58] RAOF, you know about bzr lp-submit? [09:06] (I have to use it everytime with -R unity-team though) [09:11] Hm. While I'm at it, I should probably see if it's trivial to fix the layout with DPI≠96 [09:11] RAOF: DBO is changing a lot in the Layout code [09:11] RAOF: maybe worth to wait for a week? [09:31] DBO: Do you actually want the a nux branch? It contains exactly the one-line change you suggested. [10:12] hi, this is still annoying me. Is there anything known about libappindicator failing to pick up changes for the underlying menu in case the menu has been modified by appending/inserting the entries at runtime? [10:12] * rye meant that it is again him, "annoying rye" === ogra_ is now known as ogra [11:04] is unity supposed to work on intel graphic cards? [11:05] rodrigo_, if there's a class of graphics card where it should work, that's intel's [11:05] unless it's poulsbo [11:05] aruiz, any idea what package do I need to install for the intel drivers? it seems to not work at all here [11:07] xserver-xorg-video-intel [11:07] robtaylor, try removing /etc/X11/xorg.conf if you happen to have one [11:07] oops [11:08] that's not rodrigo_ [11:08] :-) [11:08] ok, I think I already have that one, checking === MacSlow is now known as MacSlow|lunch [12:47] seb128, didrocks, kenvandine, I just pushed a branch here; lp:~karl-qdh/indicator-datetime/recurring-and-gsettings-fixes which fixes a whole bunch of issues - we get correct recurrence times and a few small gsettings fixes, I'm battling through to get colours added to the menu items too which seems to at least be feasible now [12:47] would you like me to file a merge request? [13:00] seb128: didrocks https://code.launchpad.net/~karl-qdh/indicator-datetime/recurring-and-gsettings-fixes/+merge/52059 [13:00] klattimer: hey, will get over it if I find some time today, otherwise, kenvandine and seb128 knows more about it than I :) [13:01] ok [13:01] hey [13:02] klattimer, hi, I meant to talk to you today about your way to work, nice that you bring it on [13:02] klattimer, could you try to get merge requests happening on a regular basis for fixes and incremental changes [13:03] klattimer, like ideally you would have your work of the day landing to trunk every day in small diffs [13:03] seb128: I'd love to, but there are a lot of bits of evolution which are taking time to get the hang of === MacSlow|lunch is now known as MacSlow [13:03] stacking one week of work makes review not as easy and means that if that doesn't land we have nothing to land in natty [13:04] klattimer, well it seems some of the gsettings watching and update issues could have landed as separated fixes during the week [13:04] well you can also see there's a lot of testing that's needed to go back and forth to try and tame eds into this indicator [13:04] you seem to stack refactoring and simple bug fixes in the same vcs [13:04] seb128: agreed those smaller changes should have been in a separate branch [13:05] we want at least the few liners fixed to land [13:05] even if you still fight eds for other things [13:05] but in any case today is tarball day so yeah please do a merge proposal with what you have [13:05] then we can talk to ted and others about helping you to fix remaining issues so we can roll a tarball [13:06] seb128: it's already done, link above [13:06] great [13:07] kenvandine and ted should review that when they will be online [13:20] hey andreasn [13:23] hey all [13:50] seb128: I have a new branch which fixes colours [13:50] seb128: lp:~karl-qdh/indicator-datetime/fixed-appointment-icon-colours [13:50] how would you like me to merge these [13:51] would you like a merge request to the previous branch or to trunk? [13:51] actually hang on, I need to do a small fix in there still [13:56] ronoc, hey, do you have a list of all the players that have integrated with the sound menu so far? [13:57] hmm lets see, banshee, rb, xnoise, spotify, symphony, amarok, mpd, apparently xmms, clementine [13:57] mpt ^ [13:57] did I leave any out ? [13:57] wow, that's a lot [13:57] that's cool :-) [13:57] yup [13:58] mpt, moving over to the mpris only registration in December has removed unnecessary hurdles [13:59] ah, of course [13:59] the weather is beautiful in Ireland [13:59] * ronoc is surprised [14:00] ronoc: in Frebruary? Waow! The only time I really saw in my 6 months in dubin was April [14:00] February* [14:01] ok, March in fact :) [14:01] didrocks, clear skies, really really bright. really calm. great skies at night (i'm down the country on the south coast) [14:01] the air is so clean [14:01] ronoc: where are you btw? [14:01] (I visited a lot the country) [14:01] didrocks, makes me realise what the hell am i doing in smelly london [14:01] heh ;) [14:02] didrocks, a small town called Dungarvan in Co. Waterford [14:02] ok, I went to Wateford :) [14:02] Waterford* [14:02] didrocks, http://mjcurran.com/home/store_finder [14:02] * didrocks can't type today [14:02] didrocks, i'm like that every day ;) [14:03] ronoc: heh :) I probably went not near from Dungarvan as I drove from Waterford to Cork [14:04] didrocks, around the bypass I would imagine [14:04] ronoc: exactly :) [14:05] didrocks, its where i grew up [14:05] ronoc, are you in Dublin atm? [14:05] aruiz, nope -> http://mjcurran.com/home/store_finder [14:05] ah yeah [14:05] weather bust me better there most def [14:05] ronoc: oh nice! the south coast is beautiful :) [14:06] * aruiz never went to the south coast [14:06] didrocks, yeah its nice alright. [14:06] * ronoc is wondering if he could move back here ... [14:07] * ronoc back to bug fixing [14:15] klattimer, in case you missed my comment yesterday, your indicator-datetime branch is MUCH better [14:16] klattimer, it still doesn't get the timezone's right for events [14:16] but it does show me the next 5 reliably :) [14:16] kenvandine: have you used todays? [14:16] the timezone stuff still needs playing with a bit [14:16] but I'm getting there [14:16] I've just added a new branch which even has colours for the events [14:17] woot [14:17] same branch? [14:17] well, the one which works well is [14:17] i want to get your current fixes at least uploaded today [14:17] the colours are in a separate branch I'm working on [14:17] ok [14:18] i had to remove a g_debug line to get it to build yesterday, casting to an int or something [14:18] oh yeah, sorry [14:18] I have a delay sometimes when I save :/ [14:18] no worries :) [14:18] it causes a few minor glitches when I push [14:19] ok, didn't look at your changes yet... but after pulling i am getting mostly dupes in the menu [14:19] which i didn't get before [14:20] oh... but it got the timezone right :) [14:20] hmm [14:20] it's showing my 1:1 with jasoncwarner 4 times, but displaying it in my timezone not Australia like it had been [14:20] for me I get recurring appointments with the right date/time and no repeats [14:21] kenvandine: can you run the indicator on its own with no watchers and send me a copy of the output [14:22] you mean in the indicator loader? [14:22] and without the service running? [14:22] no just the datetime service [14:23] ok, what do you mean with no watchers? === m_conley_away is now known as m_conley [14:24] INDICATOR_ALLOW_NO_WATCHERS=1 [14:24] ah, ok [14:25] I think I might have a buffer overflow somewhere but not sure where [14:26] klattimer, http://pastebin.ubuntu.com/574992/ === API is now known as Guest80757 [14:29] kenvandine: wait 10 minutes [14:29] I want to see if it increases in size [14:29] ok [14:30] looks to me like old items haven't been removed [14:33] kenvandine: http://dl.dropbox.com/u/889043/Screenshot-ecalindicator.png - this is what I see on my testing machine [14:35] klattimer, yeah, your's looks sane [14:36] not sure if other events would show multiple times like that... i am only seeing the dupe on the one event [14:36] but since i see it 4 times, no room to see if i get dupes for others [14:38] kenvandine: as far as I can tell they all have different comps [14:38] so maybe your duplicate calendar problem is interfering with it [14:38] not this one, this event is on one of my google calendars [14:38] hmm [14:38] and it is only on one of them [14:39] klattimer, note before i pulled your changes a few minutes ago, it was displaying it only once [14:39] kenvandine: what I see is 3 completely different ecal components [14:39] all with the same data [14:39] so there's got to be something up in eds [14:40] weirder is that there are 4 of them displayed :) [14:40] kenvandine: try evolution --force-shutdown then start evolution and restart the indicaotr [14:41] klattimer, no difference [14:43] kenvandine: I'm adding in a toggle to prevent an overlap occurring [14:44] if that's happening then it *might* cause your problem [14:44] but I need to make sure things don't get overwritten [14:45] this should also fix some quirks I'm seeing in the colours [14:45] kenvandine, i'll have a release for you in a bit [14:45] ronoc, cool [14:46] ronoc, bug fix only right? [14:46] kenvandine: lp:~karl-qdh/indicator-datetime/fixed-appointment-icon-colours [14:46] kenvandine, yup [14:46] kenvandine: if you can try the above branch [14:46] I think it might fix your trouble and some of mine [14:46] if so we can work out the merge path from here ;) [14:48] ah ha! [14:48] so having the colors helped a lot [14:49] it is on other calendars which i don't have enabled [14:49] :) [14:50] right [14:50] so hadn't you fixed that before so it didn't display the disabled calendars? [14:50] now checking a calendar is enabled in evo is a different challenge [14:50] because evo stores this in gconf or gkey files [14:50] didn't you do that with gconf? [14:50] kenvandine: nope not yet [14:50] humm [14:50] because I'm not sure which method evo is using atm [14:50] so why wasn't this happening earlier? [14:51] kenvandine: because it was borked in many other ways [14:51] there was no generate instances in the earlier version [14:51] and getting objects is not the same as generating instances [14:51] DBO: https://code.launchpad.net/~cando/unity/fix_692444/+merge/48767 [14:51] even though the manual says its a combination of that and generate recurring [14:51] hook that up yo [14:52] I'll dig into evolution enabled/disabled calendars [14:52] see if I can get a handle on doing it [14:56] cool [14:56] klattimer, good progress... :) [14:56] klattimer, can you propose this for merge so we can get the current bug fixes in today? [14:56] kenvandine: in a little while I'm just trying to fix a small bug [14:57] ok, just want to make sure we do it in time for tedg to review and get released today :) [14:59] hi, people. May I talk to somebody from the libappindicator team? [14:59] in particular I am interested in why the following test - http://paste.ubuntu.com/575010/ - does not actually work for dynamic gtk.MenuItem()s [15:01] rye, let me look [15:02] kenvandine: https://code.launchpad.net/~karl-qdh/indicator-datetime/fixed-appointment-icon-colours/+merge/52080 [15:03] rye, so the problem is that the dynamic item doesn't change label or get removed right? [15:04] kenvandine, after dynamic item is added, everything gets broken, so no changes to menu are seen for all the entries. [15:05] kenvandine, in that test after I add new gtk.MenuItem() i change its label (not working), change the static item label (also not working) and then hide and remove the dynamic item from the menu (which is not working as well) [15:06] seb128, is there a policy for bugs filed by apport with a failed trace? [15:06] it's pretty hard to figure out what went wrong without a stack trace :( [15:07] lamalex, no, there is stock replies on the wiki you can use to ask the submitter to use gdb and get a stacktrace [15:08] ah ok [15:08] or close the bug saying that apport didn't work and that he should submit a new report using apport again if that's still an issue [15:08] when I do that I tend to add a comment saying that the versions changed and it might be fixed so we close the bug but that he should open a new bug if he gets a crash again with the current version [15:20] rye, your code looks like it should work... i am betting there is either a appindicator or dbusmenu bug here [15:20] cyphermox: you still looking after multimonitor? [15:21] andreasn: Man, this is a thing of beauty: https://mozillalabs.com/messaging/messaging-menu [15:22] kenvandine, well, i did dbus-monitor and found no requests after i added the dynamic item, don't know where to direct the bug report in this case. Is that a responsibility of libappindicator to call dbusmenu methods? [15:23] libappindicator [15:23] kenvandine, ok, thanks! [15:24] rye, i had it drop to a debugger for me and i tried setting the label then used get_label to test it [15:24] it got the new label [15:24] but the indicator wasn't getting updated [15:25]