[01:06] hey all [08:20] good morning [09:11] Cimi! Let me know when you're around, I'd like to start playing with font styles in the theme [09:11] hi ronoc [09:11] hey mpt [09:12] davidbarth, I have computer woes still, its either natty or my hardware, can't figure it out as it's very random, going to wipe my partition this morning and try a fresh install [09:24] didrocks, ronoc can't log into a session (upgraded to natty), gdm just hangs for him sometimes....any ideas? Maybe the capabilities check? [09:25] njpatel, working now, (needed to tick 'accept invalid ssl certificates') [09:25] ah, of course [09:25] hum? [09:26] njpatel, didrocks , I can sometimes login using desktop classic but standard desktop ain't working for me [09:26] ronoc: which version of compiz do you have? [09:27] didrocks, not on my distribution right now, give me one sec and I will fetch another machine so that we can continue this converstation with natty open in front of me [09:27] ronoc: sure [09:43] didrocks, the funny thing is desktop classic works but the straight desktop does not, the usual symptom is being to get to the desktop except no panel or launcher are running [09:44] didrocks, the compiz version is 1.0.9.2.1+glibmainloop3=0ubuntu5 [09:45] ronoc: are you sure about your settings? can you try unity --reset? [09:46] didrocks, sure [09:49] didrocks, what settings ? just did the unity reset, everything reset itself, will try to login to normal desktop [09:51] ronoc: ok, that should set the default settings for both your ubuntu desktop session and the classic session, let's see how it goes [09:51] didrocks, okay that seems to have fixed it [09:51] didrocks, thx :) [09:52] didrocks, I think yesterday I was playing with ccsm and i disabled the unity plugin and renabled , i think this is what did it in [09:52] also had disk issues [09:52] ... [09:52] ronoc: cool! yeah, probably, we will need further testing and making some tweak to avoid the user shooting on his feets [09:54] didrocks, yeah i was just trying to increase my number of work spaces [09:54] didrocks, thx for all the help, i thought i was going to have to wipe and reinstall [10:27] seb128: any possibility we can have a compiz-dbg package in main? We get a lot of crashers inside compiz and it would be most useful... [10:28] kamstrup, what is wrong with the dbgsym binaries? [10:33] seb128: in the maverick cycle i had the dbgsym repo enabled and it gave me unbearable amounts of headache ... [10:34] brb, session restart === akshat_ is now known as akshatj [10:43] kamstrup: should I include vala files in dist (ie .tar.gz). automake wants to do that, but I'm hesitent [10:44] *hesitant [10:44] smspillaz: hey dude, you were on the gconf bug i saw earlier [10:44] Does doing a "bzr branch lp:" crash for anybody here too? [10:45] kvalo: i'd do so... [10:45] MacSlow, python 2.7? [10:45] kamstrup: ok, I'll do that then [10:46] MacSlow, this error https://bugs.launchpad.net/bzr/+bug/693880? [10:47] API, python 2.7 yes... but not that error... hold on I'm pasting it... [10:49] API, http://pastebin.ubuntu.com/550574 [10:49] MacSlow, is the same error [10:49] the bug description is somewhat misleading [10:50] take a look to the output on the bug [10:50] MacSlow, the error is this: [10:50] "readline() takes exactly 1 argument (2 given)" [10:50] and your output also have that [10:50] API, ah... yeah [10:51] MacSlow, in my case, while the bug is not solved [10:51] I just edited /usr/bin/bzr [10:51] in order to use an older python [10:51] the first line of the script: [10:51] #!/usr/bin/python2.6 [10:52] API, I guess I do the same then [10:52] is a hack, but works for me [10:52] just odd that only on two of my three natty boxes this happens [10:52] ehm... ignore that [11:05] API, thanks btw [11:06] MacSlow, you are welcome [11:14] didrocks, unity doesn't like my dual monitor setup [11:14] either crashes compiz [11:14] or makes the display unusable [11:15] need a crash course in xrandr - cause --auto isn't fixing it :) [11:15] ronoc: there are multiple dual monitor setup issues in compiz + unity, right. But it shouldn't crash [11:15] ronoc: I think njpatel and smspillaz are the best person to discuss about it :) [11:16] ronoc, as there isn't auto adjustment yet, when you set the new display setup, just do compiz --replace & from a terminal [11:16] i'm using unity on a dual monitor right now :) [11:16] didrocks, well it hangs or something, very difficult to debug, i'll get by for now, but i'll sue if my neck hurts at the end of the day :) [11:16] ronoc: hehe [11:17] ronoc: That will surely be easier to debug during the Rally [11:17] njpatel, ah so now I have a desktop which is across both monitor and laptop screen, [11:17] njpatel, i can't see the launcher properly [11:17] its on the laptop screen which is too small for the screen projected on it [11:18] njpatel, whats the name of the binary to start the monitor settings gui ? [11:19] gnome-display-properties [11:19] ronoc, ^ [11:20] njpatel, seb128 : bingo, sorted thx! [11:21] great [11:29] ronoc, the panel will stretch, but the rest should still work (I'm using it in dual-display now) [11:30] njpatel, all good now, I just needed to use your compiz --replace & trick from above and it auto adjusted things properly, nice one [11:31] yeah, hopefully the autoadjustment will land soon [11:34] good stuff [11:34] new forwind podcast -> forwind.net/podcasts/1 [11:34] previous -> /0 [11:43] dbarth: yeah I'm working on it === oubiwann is now known as oubiwann_ === oubiwann is now known as oubiwann_ === oubiwann_ is now known as oubiwann [12:15] smspillaz: after reinstalling compiz and unity packages I now get a different trace... Attached to bug #697587 [12:15] Launchpad bug 697587 in Unity "compiz crashed with SIGSEGV in CompRegion::CompRegion" [Undecided,New] https://launchpad.net/bugs/697587 [12:19] kamstrup: ok [12:20] kamstrup: those are really weird crashes [12:21] kamstrup: maybe try purging everything first, you might have some leftover stuff [12:21] were the packages recently updated? [12:21] i check the progress of unity on compiz about once a week using a daily build and yet to have a crash. :) [12:21] that's the thing with compiz, there seem to be very few who get crashes, and those few get lots of them [12:29] Hmmm, I can't figure out how to get the dbgsym packages... My /etc/apt/sources.list.d/ddebs.list contains the line "deb http://ddebs.ubuntu.com natty main restricted universe multiverse" which I thought should be enough [12:29] (keys imported and apt-get updated of course) [12:30] meh, now I have them... odd === MacSlow is now known as MacSlow|lunch [12:30] move along nothing to see here [12:38] smspillaz: i now have the original crash back... I added a trace with dbgsyms to bug #697587 [12:39] Launchpad bug 697587 in Unity "compiz crashed with SIGSEGV in CompRegion::CompRegion" [Undecided,New] https://launchpad.net/bugs/697587 [12:41] smspillaz: hehe... so there's a class called "PrivateRegion"... ;-) [12:42] And I see lots of stuff which could potentially segfault in CompRegion::CompRegion (const CompRegion &c) line 48 === njpatel is now known as njpatel|lunch === njpatel|lunch is now known as njpatel_ [12:55] is git+ssh://git.compiz.org/git/users/dbo/compiz-with-glib-mainloop still out main compiz branch? It doesn't look to be compatible with the dbgsym packages I have even though the last change in that branch is back from Dec. 5th [12:56] s/out/our/ === allan_ is now known as wers [13:05] hey all [13:06] kamstrup: I just did a change to that branch [13:06] kamstrup: a rather big one actually [13:06] could be enough to break the abi [13:07] smspillaz: that last comit I see is 78a5ebd15ac4b34b14ab07e95941364a09c187f3 [13:08] the latest is aa8edad38222309aee7aaf486d9f0881e37dcdb1 [13:08] wtf is going on with my git repo here then!?! :-) [13:09] kamstrup: which branch are you using? [13:09] kamstrup: glibmm-experimental or master [13:09] you should be using the former [13:09] smspillaz: git+ssh://git.compiz.org/git/users/dbo/compiz-with-glib-mainloop [13:09] kamstrup: git checkout glibmm-experimental [13:09] smspillaz: ah... so we have swicthed branches? I missed the memo... [13:09] yeah [13:10] well I was going to merge it into master [13:10] but I want to maintain both the C and C++ version for now [13:10] but I think glibmm-experimental is going to be the one that goes into core master [13:10] kamstrup: in any case, it won't be compatible with the system dbgsym as we have distro-patch… === oubiwann is now known as oubiwann_ === oubiwann_ is now known as oubiwann === dbarth_ is now known as dbarth [13:21] seb128, playlists have landed in rb -> http://git.gnome.org/browse/rhythmbox/commit/?id=d06dcc1 [13:21] back shortly lunch [13:21] ronoc, great === dbarth__ is now known as dbarth [13:49] kamstrup: what's the end of the story wrt to the dbgsym packages btw? [13:49] kamstrup: i'm having similar issues with x packages === MacSlow|lunch is now known as MacSlow [13:59] dbarth: i have them installed now. And they do seem to work, but I am crossing my fingers once I start upgrading... [14:06] so the dbgsym packages, right? not the -dbg ones [14:09] dbarth, what with dbgsyms? [14:15] hey iainfarrell [14:17] MacSlow: hello there [14:41] seb128, just testing amarok with the new registration process, I was under the impression that looking up a desktop file should only involve the name of the desktop file and not the full path [14:42] you mean? [14:42] seb128, i am attempting to lookup amarok.desktop but it can't find anything although there is an amarok.desktop in /usr/share/applications/kde4/ [14:42] kamstrup: sorry, but a really long one again: https://code.launchpad.net/~kvalo/indicator-network/settings-vala-convert/+merge/45244 [14:43] hyperair, all good with the name change in the desktop file ? [14:43] seb128, did you understand what I was asking there ? [14:43] ronoc, that directory is not a standard xdg one [14:43] ah [14:43] hmm didrocks, smspillaz: is this one really fix-released in ubuntu? [14:43] or is that more of that gconf bug? [14:44] dbarth_: "this" ? [14:44] i'm looking for the gconf bug id btw [14:44] https://bugs.launchpad.net/unity/+bug/691561 [14:44] seb128, right so how do I handle amarok then? [14:44] sorry, forgot the link ;) [14:44] dbarth_: we are speaking about it for the past 3 days :) [14:44] dbarth_: there is a workaround now in natty package [14:44] as a distro-patch [14:44] but it makes the menu not appearing under some conditions [14:45] so smspillaz will look at the proper solution [14:45] ronoc, why don't they install a desktop entry in the standard xdg dir? [14:45] seb128, I don't know :) [14:45] agateau, ^ [14:46] * agateau reads backlok [14:46] backlog [14:46] agateau, thx [14:46] the amarok desktop file cannot be loaded by a system desktop file lookup because its not in a standard xdg lib [14:46] weird [14:47] all of kde is in /usr/share/applications/kde4 [14:47] yeah, why is this ? [14:47] probably a remaining of the kde3 to kde4 transition [14:47] when kde3 versions were in /usr/share/applications and kde4 in /u/s/a/kde4 [14:47] i need to be able to get hands on that desktop file inorder to let the app register [14:48] ScottK probably knows about this [14:48] agateau, makes sense [14:48] (ScottK: the reason for the kde4/ dir I mean) [14:49] ronoc: maybe there is a way to add a dir to the standard xdg dirs? [14:49] didrocks would know this I am sure [14:50] agateau, ronoc: well you can tweak XDG_DATA_DIR I guess [14:50] agateau: not sure we want to add the kde4 dir to the GNOME session [14:50] apps that keep their menus should be reported as bugs right? [14:50] lamalex, what do you mean? keep their menus? which one? [14:50] didrocks: of course you want, kde apps love you! [14:51] agateau: apparently, you don't share your desktop files and put them in a dedicated directory! :) [14:51] ronoc, is amarok in the GNOME menus? [14:51] didrocks: :) [14:52] ronoc, seems the KDE guys just don't want amarok to be used under GNOME [14:53] seb128, don't know running unity and cant find anything :) [14:53] ronoc, try running alacarte [14:53] see if it's listed in there [14:54] seb128, yes it is [14:54] any idea when the dash comes back? i miss it :) [14:55] ronoc, grep -i amarok /usr/share/applications/* [14:55] ronoc, does it find anything? [14:57] seb128, yeah its in the cache [14:57] seb128, usr/share/applications/desktop.en_GB.utf8.cache:[kde4/amarok] [14:57] kvalo: i'm on it [14:57] the lookup must not be using the cache [14:58] ronoc, if you remove the cache is it still listed by alacarte? [14:58] didrocks: is the corresponding menu issue tracked independently? there are bugs for menu and panel items not appearing, but i thought that was separate [14:58] kvalo: wow, Andrew is most productive! :-D [14:59] seb128: doesn't applications.menu add /usr/share/applications/kde4/ ? [14:59] dbarth_: normally, once the workaround removed and the crash issue correctly fix, we should not see anymore the menu issue [14:59] Amaranth, no [15:00] Amaranth, grep kde in /etc/xdg/menus -> nothing [15:00] kamstrup: he is. and using vala for settings app is really good imho. [15:00] ah, perhaps we recursively look through the dir [15:00] seb128, yep it is [15:00] kvalo: I take it you've been test running this yourself? [15:00] ronoc, seems buggy [15:00] * Amaranth tries to find that bit in gnome-menus [15:00] kamstrup: it doesn't do anything else than opens some windows [15:00] ronoc, gmenu-simple-editor lists it as well? [15:01] seb128: it should, they both use gnome-menus [15:01] kvalo: ok [15:01] kamstrup: but yes, I have been testing it a bit myself. now I start adding backend code to it [15:01] Amaranth, hum, are subdir supposed to be read? [15:01] seb128: pretty sure, yes [15:02] ronoc, what glib api do you use to find the .desktops? [15:02] seb128: the spec doesn't say for sure either way iirc but everyone does it [15:03] seb128, gmenu-simple-editor does aswell [15:03] ok [15:03] well anyway seems gmenus read the subdir then [15:03] but the api you are using doesn't [15:04] brb [15:05] seb128, http://www.valadoc.org/gio-unix-2.0/GLib.DesktopAppInfo.html [15:05] seb128, must be the scenario alright [15:07] ronoc: you mean g_desktop_app_info_lookup_get_default_for_uri_scheme? [15:07] ronoc: turns out the desktop file Name field was a bug. [15:08] ronoc: it does a pkg-config check and sets the Name accordingly. i didn't include that dep in build-deps, so all i did was add that dep [15:08] ronoc: is it urgent enough to require an upload immediately or can i wait for the next release? [15:08] Amaranth, I'm using vala [15:08] hyperair, release on the 12th is fine [15:09] ronoc: alright then [15:09] seb128, monodevelop [15:09] keeps its appmenu rather than putting it on the panel [15:12] check open indicator-appmenu and appmenu-gtk bugs [15:14] lamalex: we have a distro patch for monodevelop and eclipse IIRC for that, because otherwise it tries to export it and it's never drawn [15:15] so for both, there should aready be a bug about it [15:15] ok [15:15] http://launchpadlibrarian.net/56822992/monodevelop_2.4%2Bdfsg-3ubuntu1_2.4%2Bdfsg-3ubuntu2.diff.gz [15:15] is libreoffice going to be patched this cycle for menus? [15:16] depends if someone wants to tackle it :) [15:16] my sister hopes somebody does hehe [15:17] ooh tedg's back. [15:17] tedg: did you see those memleak patches? [15:17] i would tackle it, but i am no developer [15:17] hyperair, Not yet... still catching up on e-mail... [15:18] tedg: okay, take your time. [15:18] hyperair, Which project? [15:18] tedg: dbusmenu, indicator-applet, and indicator-messages, i think [15:18] hyperair, Oh, wow. Cool! [15:19] vala bug definitely [15:19] tedg: but don't apply the dbusmenu patch just yet, the latest version seems to cause some hard to debug memory corruption issues, so i'm redoing it [15:20] tedg: it's mostly stuff that you get from dbus as a client that you don't free, but dbus expects you to free. the surprising bit is why only *i* seem to be seeing these memory leaks. [15:20] ronoc: vala doesn't seem to have a way for you to use the function I mentioned [15:21] tedg: oh there's a libindicate patch and no indicator-applet one [15:21] hyperair, Do you think we should do these for Maverick? I mean, that most of the dbus stuff is changing with the port to GDBus. [15:22] tedg: what's GDBus? [15:22] hyperair, The new dbus bindings that are in GIO vs. the external lib dbus-glib. [15:22] morning tedg [15:22] Amaranth, no it doesn't but it should search the child directories of any xdg data dir [15:22] tedg: there hasn't been much noise about memleaks in the indicators [15:23] tedg: if the dbus stuff is going to be changing, then i suppose we should just drop the patches [15:23] Amaranth, quoting the glib api for the g_desktop_app_info constructor "GIO is looking for a desktop file with this name in the applications subdirectories of the XDG data directories " [15:23] Morning lamalex [15:23] hyperair, Well, if they're causing memleaks I think they'd be candidates for SRU, no? [15:24] ah tedg, lamalex: good morning [15:25] there are a couple of new/undecided bugs in the appmenu and indicator inbox [15:25] tedg: thing is, nobody except me seems to be facing it, even for long running systems. i suppose we could do an SRU, but the problem is that my patches don't completely fix the memleaks, and i'm still seeing some severe ones i can't track down [15:25] could you have a pass at those to now were we stand here? [15:25] http://reports.qa.ubuntu.com/reports/team-assigned/canonical-dx-team-assigned-bug-tasks.html [15:25] tedg: i could give you a massif log, but i can't figure out what's leaking what [15:26] morning dbarth_ [15:30] hyperair, I realize, memleaks are like that. Makes the SRU process harder though :-/ [15:31] tedg, well obvious leak fixes are easy to sru [15:31] seb128, That's the problem, they're not obvious. [15:31] tedg, we just need some users to run the new version for a while and confirm it doesn't crash [15:31] tedg, well usually a leak fix is a one liner to free things which you forgot to free [15:31] seb128, I thought an SRU could only be confirmed by someone who could recreate the bug originally? [15:32] tedg, if that's implies rewrite of code it's harder to sru indeed [15:32] If that's the case, we'll have to be extra careful that hyperair's house doesn't get hit by lightening :) [15:32] tedg, well for cases not easy to get confirmations we are usually fine with several users confirming the fix doesn't create any issue [15:32] like the fix seems right from a code review and users don't get a new crash is ok [15:33] Okay, cool. [15:33] Thanks seb128! [15:33] kvalo: done :-) i'm glad that most of it where deletions :-) [15:33] tedg, you're welcome [15:33] dbarth_, Was that link for me? [15:33] tedg: hey! happy new year :-) [15:33] Happy new year kamstrup! [15:33] tedg: yeah, you and lamalex [15:34] dbarth_, Okay, looking now. [15:34] tedg, it's basically me who subscribed your team or assigned it to a stack of appmenu issues [15:34] kamstrup, Hey, you might know. I thought there was a ZG data provider for Firefox. Do you know where that went? [15:34] tedg, that's what we discussed before the holidays, someone needs to deal with appmenu issues [15:34] seb128, Hmm, I don't love you anymore :) [15:35] like you did like me before? ;-) [15:35] DBO: hey dude, when you're there: can you take a look at https://bugs.launchpad.net/unity/+bug/691651? [15:35] you tried to make me dry texan wine, I know since that you don't like me :p [15:35] dry -> drink [15:35] Careful seb128, you're coming to Texas. We don't have laws here :) [15:36] haha [15:36] dbarth_: i'm wondering who should try and fix that between you and neil [15:36] no laws but guns? ;-) [15:36] Law of the jungle only. [15:36] dbarth_, sure no problem [15:36] oh, you're there; hi ;) [15:36] yeah [15:37] rule #1 of ayatana, do not speak until spoken to :P [15:37] DBO never sleeps, he just pretends to [15:37] :-D [15:37] kamstrup: thanks a lot! [15:38] dbarth_, bamf will signal the information [15:38] will make the signaling possible, thank you :) [15:38] so? up to neil to use it, right? [15:39] i have to make sure the info is available to neil [15:39] but yes [15:43] tedg: http://people.ubuntu.com/~hyperair/massif-output === spikeb1 is now known as spikeb [15:43] tedg: this is one which i've spent a whole week staring at before giving up and waiting for you to return to the internet. =p [15:44] tedg: i suspect it's leaking dbusmenu_menuitem objects, but i can't be sure [15:46] does anyone have any idea when the stacking order issues with unity are going to be sorted out [15:46] as not being able to see any menus is getting a little old [15:47] DBO: hmm ;) assigning to you and then hand over to him once bamf shares all what it should in its signal [15:47] dbarth_, fair enough :) [15:47] oh and the screen saver appearing _under_ my background so i don't know its there is also getting old [15:47] tedg: when will stuff be ported over to GDBus? [15:58] tedg: i think it's in lp:zeitgeist-datasources or something... [16:00] tedg: sorry... that was lp:zeitgeist-dataproviders [16:00] there is one for chrome as well I think [16:00] tedg: but beware that I don't know if it respect pr0n mode yet ;-) [16:00] kamstrup, Ah, cool. Perhaps I should harass kenvandine about packaging it? I'm scared to compile XUL stuff :) [16:01] * dbarth_ restarts with a new xorg-edgers mesa; wish me luck [16:01] kamstrup, Heh, those would be some interesting ZG queries... ;) [16:04] tedg, ugh... please don't ask :) [16:04] tedg, hi. My name is Cristian, you replied to my mail here: https://lists.launchpad.net/ayatana-dev/msg00052.html [16:04] haha kenvandine [16:04] kenvandine, Oh, I thought you were working on those data providers, no? [16:04] i messed with that during UDS... distcheck didn't work and i fixed a bunch of it [16:04] evaluate, Howdy! [16:04] tedg, I'd like to mention one more thing that's broken, that is underscores in the menu items. [16:05] evaluate, Yeah, we assume that they're mnemonics. You need to use "__" for underscore. [16:06] tedg, hmm. [16:07] tedg, would that mean that mnemonics are allowed in menus? that's pretty awkward... [16:08] I mean, a user has to click the icon to open up the menu, doesn't he? what sense does it make to create a shortcut for a menu that is opened with a mouse click anyway? [16:11] evaluate, Usually you can hit those character to move quickly through the menu. [16:12] tedg, btw, if this is implemented, that it's not implemented correctly, because if I actually try to use an underscore in a menu, the letter that is preceded by the underscore doesn't get underlined or anything (there is no visual hint that that letter is the shortcut) [16:17] evaluate, Hmm, that's odd. [16:20] evaluate, i tedg, that might be a unity panel rendering bug? [16:20] lamalex, No, the menus are still rendered by GTK. [16:20] there are bugs about that i've seen [16:20] ah right [16:21] evaluate: could you file a bug for that [16:21] tedg, also, is there any reason why you wouldn't allow formatting on text in the menus? [16:21] lp should find some matches and see if there are existing bugs that cover your problem [16:22] dbarth, I can file a bug against the fact that the underscores are not rendered properly (as in, they are not rendered as underscores) if you want [16:23] dbarth, I'm not going to file a bug against underscores not being rendered as mnemonics, because that is utterly useless IMHO [16:24] and it's also against the logic that GTK normally uses. I now have to make an extra function that will replace all underscores with "__" just to have the appindicator menu rendered correctly, that is counter-productive... [16:33] btw, why can't you just use "gtk_menu_item_new_with_mnemonic" to let users create menu items with mnemonics and leave the normal menu items with underscores alone? [16:33] s/users/developers/ [16:45] evaluate, We can, we just haven't yet. :) [16:54] sounds like a bitesizable patch [16:54] er, bug [17:00] agateau, got to the bottom of the desktop file loading, one sec [17:11] agateau, the desktop file name that the mpris root interface returns should be kde4-amarok so as the lookup knows to look in /kde4/.... [17:12] I will point this bug towards to you if you don't mind ? [17:12] agateau, ^ [17:17] ronoc: ok, will have a look at it (please remind me next week if I forget) [17:17] ronoc: I am off now [17:18] agateau, grand will mail [17:18] enjoy the evening [17:21] tedg, bratsche: Hi! I'm on the SWT team and we are interested in looking at what we need to do to get Eclipse menus to show up in the application menu. [17:21] ircbog, Sweet, that'd be great. [17:22] ircbog, You guys are using GTK, correct? [17:22] that is correct - and we are creating all of the menus by hand (i.e. we don't make use of GtkUIManager) [17:23] ircbog, Hmm, that shouldn't pose an issue at first blush. [17:23] ircbog, I'm not sure what the issue is though, perhaps how you're putting them into the window we're not detecting them properly? [17:23] ircbog, We watch the window and trying to find the menus that are associated with it. [17:24] We put them inside of a vbox [17:24] which is inside of the GtkWindow [17:25] Hmm, bratsche, you around? ^ [17:25] How do you extract the menus from the window? [17:25] ircbog, We load a GTK module that watches things like new windows being created and looks at the object tree. [17:26] ircbog, I'm guessing that code is somehow not finding your menus. [17:26] ircbog, http://launchpad.net/appmenu-gtk is the code that does it. [17:26] Yeah [17:26] Hey ircbog [17:26] hi! [17:27] Yeah, I never looked into Eclipse to try to understand what was going on in there. [17:28] Let me take a look at the appmenu-gtk code. [17:29] ircbog: If you have any questions about it, let me know. [17:29] ircbog: Also, there is a patch in gtk+ in Ubuntu that appmenu-gtk depends upon. [17:32] DBO, what major piece is expected to land next? [17:32] dash, places? [17:32] jono, I would expect dash next [17:33] DBO, cool - any idea when that is likely to land? [17:33] nope :) [17:35] * spikeb sighs [17:41] bratsche: Looked at the code and tried a simple snippet on our end. We believe the problem stems from the fact that the code iterates through the widget tree once (when a window is initially realized). In a SWT app, the menu is not available at that time, hence we don't show up. [17:42] ircbog: It iterates through the widget tree, but it also watches for insertions and parent changes. [17:43] * bratsche pulls up the code [17:46] ircbog: Do you insert the menubar into the window very late in the process? Like after the window has already been shown? [17:48] ircbog: Eh.. seems like if that were the case it should be caught by toplevel_notify_cb() [17:49] bratsche: it depends on the app. In our simple example, it fails if we realize the window before we insert the menu but works if insert the menu first. [17:49] Oh, hmm. [17:49] ircbog: Do you have a pretty simple test program that demonstrates this which you could send me? [17:49] bratsche: it looks like the code expects a notify signal. Are we expected to send that out? [17:50] No.. when your menubar is added to the window then I should get a notify. [17:51] ircbog: Is your example program something you can share with me? [18:04] bratsche: of course. Is the SWT code OK for you? (It's a very simple app with one menu item + one hack to stop realizing the shell before the menu is added) [18:05] I'm not familiar with SWT.. I guess I can get all the packages I need to build it with "apt-get build-dep eclipse" ? [18:08] bratsche: Let me see if I can find a C code equivalent for you [18:08]