=== LucidFox is now known as lucidfox [07:03] Is it possible to tell programmatically if the appindicator menu is open? [07:04] I have a program using a tray icon that updates its menu every time it is opened, and polls for its items every X seconds, so porting it to the indicator API causes flicker when the menu is open [07:07] I also wonder what is the Right Way(tm) for a third-party application to distribute separate indicator icons for dark and light themes [07:07] monochrome ones, that is [07:12] And... are there any plans to patch gnome-shell for indicator support? [08:54] What would be the best way to indicate on https://wiki.ubuntu.com/NotificationAreaTransition/CompatibilityFixes that an application is fixed? [08:55] For example, I ported liferea, epiphany-browser, guake and linuxdcpp, so they should no longer be on that list [09:42] kenvandine: davidbarth: could one of you re-assign this https://bugs.launchpad.net/ayatana-ubuntu/+bug/599844 to the canonical desktop team? [09:42] Launchpad bug 599844 in Ayatana Ubuntu "Port the keyboard indicator in gsd to use the app. indicator protocol (affected: 1, heat: 12)" [Medium,In progress] [09:42] I've updated the patch, no more memory leaks and all the other issues brought up have been addressed [09:47] klattimer: hi, ok checking that now [09:53] closed a bunch of gsd bugs [09:53] and it's much nicer to use now [09:53] feels a bit sharper :) [09:59] klattimer, hey, thanks for the g-s-d fixes, please don't set the bug fix released though [09:59] klattimer: you want to reassign to get confirmation of the fixes and then have it landed? [09:59] seb128: it's not [09:59] klattimer, use fix commited rather, they will be closed when the package with the update is uploaded [09:59] seb128: it's still 'in progress' [09:59] right [09:59] [Bug 620571] Re: [Maverick] Move keyboard layouts out of the submenu [09:59] Launchpad bug 620571 in gnome-settings-daemon (Ubuntu) "[Maverick] Move keyboard layouts out of the submenu (affected: 1, heat: 8)" [Wishlist,Fix released] https://launchpad.net/bugs/620571 [09:59] ** Changed in: gnome-settings-daemon (Ubuntu) [09:59] Status: Confirmed => Fix Released [09:59] [09:59] I just got that by email [09:59] ah [10:11] seb128: sorry, I thought that the best way of pushing the bugs upto the main bug [10:13] seb128: so does that help to have dx-team reassigned, or can we just check the packages and you take the proposed ones into maverick on monday? [10:13] klattimer, no worry [10:14] davidbarth, dx-team reassigned? [10:14] davidbarth, klattimer: I just uploaded the g-s-d update [10:14] klattimer, just set the bugs you fix to fix commited so I know which ones to list in the changelog for the upload [10:15] klattimer, do you have other changes that need upload? [10:18] davidbarth, mpt, klattimer: btw can we drop the keyboard icon? [10:18] drop? wdym? [10:18] ie bug #620331 [10:19] Launchpad bug 620331 in gnome-settings-daemon (Ubuntu) "please drop icon from keyboard indicator (affected: 1, heat: 10)" [Low,New] https://launchpad.net/bugs/620331 [10:19] seb128, no, not without changing the API. [10:19] And it would only be a temporary change until the icon generation is implemented. [10:19] hum ok, it has very weird spacing compared to other indicators and the icon waste space [10:19] temporary like for this cycle in the stable version having something nice to use === gord_ is now known as gord [10:21] seb128: we'd have to do something in libappindicators and libindicator I think, which might break abi [10:21] it's possible to do it without breaking abi I suppose but that could be messy [10:21] e.g. passing a NULL icon to indicator_new [10:23] seb128: no other changes to upload no [10:23] klattimer, mpt: ok thanks, don't bother then, could you add a comment in this bug saying that? [10:23] seb128, mpt: i guess pitti asks more for a way to hide it even if you do have multiple layouts configured [10:23] klattimer, ok, what do you work on next? [10:23] seb128, mpt: which would be consistent with your (mpt's) design [10:23] davidbarth, no, pitti just say the icon is useless and use space [10:24] seb128: jorge said apport [10:24] klattimer, no ibus fixing? [10:24] i don't think so [10:24] seb128, sure [10:24] by implementing the option to hide the indicator, the space is saved, and that's part of mpt's design [10:24] that's still broken in maverick [10:24] is it? [10:24] hmm [10:24] you updated your ppa [10:24] but you didn't send any updated patch for us to upload [10:24] i tested in my vm and the package in the ppa was working [10:24] that's why I'm asking [10:24] i'll test again [10:24] well you didn't ask any sponsoring [10:25] is your update patch to upload? [10:25] updated [10:25] or was that just a testing ppa to debug? [10:25] there's no updated patch [10:25] just an updated build and removing the distro patch [10:25] well you have a ppa version working according to comment [10:25] yeah, that's a rebuilt package [10:25] same patch as already in the bug report [10:25] which is the same as in maverick? [10:26] oh hang on [10:26] there is something [10:26] davidbarth, the point is that the icon is useless [10:26] I'm testing it now, nearly forgot about this [10:26] there's still a problem with the activated signal not being sent from menu items [10:26] davidbarth, it eats space compared to what we had an other systems have, ie just the label [10:26] in python [10:26] need tedg to look into that testing my ibus package which shows the problem [10:28] seb128: right, so that's why i say, we should push a little further and add mpt's checkboxes to be able to hide the icon [10:28] davidbarth, but seems it's not trivial so let's not bother for this cycle [10:28] having the spacing fixed would be nice though [10:28] seb128: a gconf key and a dialog update [10:29] it's a bit weird right now [10:29] davidbarth, ok, that would be ok I guess [10:29] spacing [10:29] ? [10:29] klattimer, http://launchpadlibrarian.net/53938769/ibus_1.3.7-1ubuntu2_1.3.7-1ubuntu3.diff.gz is weird [10:29] for the ibus one? [10:29] davidbarth, did you try the layout indicator? [10:29] weird? [10:29] or the gsd one? [10:29] what's weird about it? [10:29] i'm testing it right now (ibus is fubared on my system) [10:29] klattimer, did you drop the indicator support of was there some cruft in the maverick version? [10:29] there was cruft in the maverick version [10:30] klattimer, I'm trying to understand what you did in your ppa and what we should do in maverick to it [10:30] oh ok [10:30] a patch which removed part of my patch [10:30] so we need to clean that? [10:30] yeah [10:30] ok, that's what I was asking before [10:30] thanks [10:30] if you look I removed debian-changed-1.3.7-1ubuntu2 [10:30] davidbarth, gsd one [10:30] klattimer, right, ok, makes sense now [10:30] klattimer, thanks [10:30] davidbarth, there is no left margin [10:30] davidbarth, but there is quite some space between the label and the right border [10:31] seb128: there is for me, i have the network icon on the left [10:31] the connman one that is [10:31] seb128: in GSD? [10:31] hmm right [10:31] It's hard to get the guide size right on that [10:31] I used XXX1 [10:31] davidbarth, click on it and watch the rectangle [10:31] on the right you mean, the right padding is too wide [10:31] yeah, because of the guide [10:31] the spacing next to the label and icon are not identical [10:31] as different layouts will be different widths, and changing width doesn't appear to be the right thing to do [10:32] davidbarth, yes [10:32] ok, I've to run for some errands, will be back in an hour [10:32] ok [10:35] seb128, done [10:36] to be nitpicking, the guide could be calculated according to the declared layouts, so that if none are 4 char wide, the padding could be adjusted [10:38] davidbarth: that's true [10:39] do you want me to do that and produce a new patch? [10:41] davidbarth: the other thing with it, would be that it's centrally aligned within the allocated space [10:41] if we're going to have an icon we should align to the left against the edge of the icon + spacing [10:48] Cimi: ping [10:48] klattimer: align to the left? sorry, re-reading [10:50] klattimer: the label you mean, but then that's something to patch in indicator-application [10:51] yeah that's right [10:51] I'm just saying [10:51] I can patch the guide size now if you like? [10:53] klattimer: that'd be cool, yes [10:53] klattimer: if you think that's a safe change [10:54] klattimer: ie, if there aren't parts to update all across the module [10:54] klattimer: i'm looking at the dialog thing, to see if mpt's checkbox could be also an easy win [10:54] k [10:55] ronoc: pong [10:55] cool [10:55] klattimer: can't find where the ui for the layout preferences tab is located [10:55] :/ [10:55] Cimi: any joy with the colours etc. ? [10:55] klattimer: do you know where that is? [10:56] not yet [10:56] nope [10:56] not yet [10:56] :) [10:56] will find out in a sec [10:56] ok, i'll keep digging, don't worry [10:56] Cimi: okay let me know if you need anything [10:58] klattimer: seems to be in gsd-keyboard-xkb.c [10:58] ? [10:58] the layout preferences? [10:58] nooo, surely not [10:58] klattimer: ah, where then? [10:59] I've been through that file all day and haven't seen it [10:59] not sure [10:59] gimme a sec [10:59] it calls popup_menu_launch_capplet from gsd-keyboard-xkb.c [11:00] ah a capplet of course [11:00] so there's a capplet for it called gnome-keyboard-properties [11:00] yeah, hmm [11:02] davidbarth: your patch is on the bug report now [11:02] using the shortest guide width [11:03] klattimer: wow! impressive! ;) [11:03] not really :) [11:03] right onto the preference item [11:05] davidbarth: the file gsd-keyboard-plugin is the capplet [11:06] hmm noooo [11:06] :/ [11:06] ok, this is getting a bit quirky [11:08] klattimer: ah ok, so it's only for "extended" layouts, cool, the changes looks ok [11:09] klattimer: it's in gnome-control-center [11:09] ah bugger [11:09] and the ui is in the capplets-data package [11:11] gnome-keyboard-properties-dialog.ui [11:12] yeah looking at it now [11:13] well, a few things need to be done, add the necessary XML, add the handler for it which writes to gconf and then have that gconf change detected in the applet and the hide to be done [11:13] ... is there not a way to hide applets from showing in libindicator? [11:13] I'd assume not [11:13] klattimer: hmm, it's going to be a fun one, with a schema definition to share between the capplet and the gsd plugin [11:14] lovely [11:14] :/ [11:20] klattimer: hmm, a new way with the approver mechanism, but that's even more complex, because that would mean fiddling with bamf, for something it's not meant to be [11:21] klattimer: can you get the gconf key first, because if time runs out, at least that provides an option for "people who know" (like pitti) [11:22] i'll have a look at what can be done [11:24] mpt: https://bugs.launchpad.net/ubuntu/+source/gnome-settings-daemon/+bug/620331/comments/7 [11:24] Launchpad bug 620331 in gnome-settings-daemon (Ubuntu) "please drop icon from keyboard indicator (affected: 1, heat: 10)" [Low,New] [11:25] davidbarth, neat [11:25] klattimer: awesome, thanks; assigning the task to you then; i know you loooove that indicator ;) === MacSlow is now known as MacSlow|lunch [12:08] mpt: I've updated ibus to include a radio menu item, but the string "Turn off input method" should probably be changed to "Input Method Off" as it's a member of the radio group [12:08] when is string freeze? [12:08] klattimer, excellent choice. String freeze is Thursday 26th. [12:09] urgh :/ [12:09] I've got a bunch of string fixes to do before then [12:09] :( [12:10] ok, well I'll change the string [12:10] what's the procedure when I've changed the code? [12:10] should I just leave it for translators to just get to work? [12:10] I don't know, sorry. [12:10] k [12:11] well seb128 do you know ^^ [12:11] David Planella (dpm) might also know. [12:12] klattimer, you can change the strings they will show up in launchpad to translate [12:12] oh, don't need to change any other files? [12:12] that's pretty cool ;) [12:13] the problem is, there's something wrong somewhere in the python bindings for the indicators [12:13] the radio group doesn't even change on click [12:13] there's definitely something wrong here :/ [12:13] * mpt wonders why nothing appears in the panel when he turns on IBus [12:13] anyway, I'll update the patch [12:13] kamstrup, open a bug and assign to ted [12:13] mpt do you have the ibus from my ppa [12:13] the current one in maverick is broken [12:14] the fixed one will be uploaded today [12:14] seb128: did you mean me? [12:14] seb128: I'm cooking a new patch for ibus with the menu item changes mpt wanted [12:14] will add it to the bug report [12:14] klattimer, yes [12:14] kamstrup, sorry that was for klattimer [12:15] klattimer, if you have issues due to the indicator stack please open bugs [12:15] so they can be assigned to ted [12:16] k will do [12:23] klattimer, no, I'll try your PPA [12:24] I see "System" > "Preferences" now contains (1) "Input Method Switcher" (2) "Keyboard" (3) "Keyboard Input Methods" (4) "Keyboard Shortcuts" [12:25] One day we'll fix that, but for now it's a mighty depressing example of uncoordination in open source [12:25] seb128: can you assign https://bugs.launchpad.net/ubuntu/+source/ibus/+bug/621075 to tedg [12:25] Launchpad bug 621075 in ibus (Ubuntu) "Python app indicator fails to send "activate" signals (affected: 1, heat: 6)" [Undecided,New] [12:25] mpt: I know, I hate that too :/ [12:26] what's the right thing to do however, is mighty difficult [12:26] sorry guys disturbing your work but.. mpt when will we have an ubuntu-control-center that handles all this kind of stuff? yea, like mac [12:26] merging of gnome-settings-daemon->keyboard with ibus in some way [12:26] klattimer, done and reassigned to one of the components he works on so he can't ignore it [12:26] c10ud, http://live.gnome.org/Design/SystemSettings/ [12:26] awesome :) [12:27] no maverick i guess [12:28] certainly not [12:28] UI freeze is six days away [12:29] heh, btw looking at all those visual improvements you've been doing lately, a 1080 dropdown menu for preferences is certainly a punch in your eye :p [12:30] c10ud, yes, especially when bug 578768 makes it near-impossible for me to choose the later items. :-) [12:30] Launchpad bug 578768 in light-themes (Ubuntu) ""Main Menu" submenus scroll badly with Ambiance/Radiance theme (affected: 4, heat: 24)" [Undecided,New] https://launchpad.net/bugs/578768 [12:32] aww [12:35] "To be removed: xserver-xorg-video-i740" -- I hope that isn't anything important [12:35] mpt: doubtful unless you've got a graphics card made by intel a decade ago? [12:35] Mine's Intel but 2007~2008 vintage [12:35] it'll be fine :) [12:37] mpt: btw https://bugs.launchpad.net/ubuntu/+source/light-themes/+bug/578768 should be marked as a duplicate of https://bugs.launchpad.net/bugs/548652 [12:37] Launchpad bug 578768 in light-themes (Ubuntu) ""Main Menu" submenus scroll badly with Ambiance/Radiance theme (affected: 4, heat: 24)" [Undecided,New] [12:40] ronoc: the code is there [12:40] we just need to decide which colors we might use [12:41] Cimi, sure for now I have the colours hardcoded in the play-button.c, can you use these for ambiance? [12:41] Cimi: which branch is it ? [12:42] ronoc: I mean [12:42] I am actually getting the colors fromgtk [12:42] apply a shade [12:42] and draw with them [12:42] now the question is: [12:43] okay [12:43] which colors we might use? I suggest bg[NORMAL] for the background, and fg[NORMAL] for the symbols on top [12:43] sounds good [12:43] does it support prelight etc etc? [12:44] not yet, never spec'd but at some point we might need it [12:44] lets refactor what is there to start with and then add stuff as we need it [12:46] which is the design to implement? [12:50] chaotic: [13:14] klattimer, with the keyboard menu from your PPA, it now shows the correct title ("USA2" instead of "USA" as it did before), but the menu has zero items. [13:16] kenvandine: reading about libgwibber, is it possible or will it be possible to retrieve statuses? [13:16] kenvandine: I would very like to see that for some sexy integration into smuxi [13:20] ronoc: bzr branch lp:~cimi/indicator-sound/use-gtk-colors [13:20] thanks Cimi, will take a look after lunch [13:21] kenvandine: with retrieve I mean to fetch the status per API not UI [13:32] Cimi, did I hear you wanted to do rounded corners on the album art ? [13:32] :) [13:37] meebey, not yet, but that is planned [13:37] the goal is to provide everything gwibber knows about through libgwibber [13:37] and then the gwibber client will use it [13:37] and hopefully lots of other apps [13:38] meebey, i would love some specific use cases you are after though, it will help in implementation [13:38] kenvandine: I have an issue to solve and havent decided yet [13:38] kenvandine: rewrite of the twitter engine in smuxi using twitterizer 2.0 (which has oauth support) [13:39] kenvandine: or use libgwibber if it can do the job :) === MacSlow|lunch is now known as MacSlow [13:39] what is twitterizer going to do about the api key issue with oauth? [13:39] kenvandine: I have one use-case issue though, I would need pure libgwibber support... a headless linux server with smuxi server has no X server etc [13:40] seb128: can you reassign this bug to canonical desktop team, the patch I have is working nicely; https://bugs.edge.launchpad.net/ubuntu/+source/indicator-application/+bug/558841 [13:40] Launchpad bug 558841 in indicator-application (Ubuntu Lucid) "bluetooth "devices" menu item not working in bluetooth indicator (affected: 20, heat: 121)" [Low,Triaged] [13:40] kenvandine: the client app needs to provide the key [13:40] right... but the key can't be in the source [13:40] klattimer, ok [13:40] kenvandine: it has to be in the source for opensource software [13:40] kenvandine: there is no issue ^^ [13:40] klattimer, btw the g-s-d one still has issues [13:40] seb128: like what? [13:40] meebey, but twitter won't allow that [13:40] kenvandine: everyone can use the key of course [13:41] kenvandine: what? [13:41] meebey, we are going though right now with gwibber... [13:41] it is a huge issue [13:41] klattimer, the menu selection is wrong sometime, it shows a layout selected which doesn't match the label [13:41] kenvandine: thats the standard practice to my knowledge [13:41] klattimer, the label is right though [13:41] seb128: not with the updated patch it shouldn't [13:41] klattimer, but it does with the patch from today [13:41] meebey, they told us they will shutdown the key if they find it in any published source [13:41] or is there a specific circumstance that "sometimes" is [13:42] I mean, if you change it by keyboard shortcut does it go out of sync? [13:42] meebey, i am working with them right now [13:42] klattimer, yes, I've 3 layout, I select one, do both alt together [13:42] the label is correctly updated [13:42] but the list selection is not [13:43] but the list selection isn't made [13:43] ah... I get ya [13:43] meebey, it might be good enough to obfuscate it by requiring the distro packager to include the key in the packaging and install it in a separate location on the filesystem [13:43] I know why that'll happen, but it's a tough one to fix [13:43] lemme tinker with it [13:43] klattimer, thanks [13:43] meebey, they are thinking about that... we have gone back and forth a lot now... getting me stressed! [13:44] i need to get it resolved and SRU'd for lucid before the end of the month when they shutdown basic auth [13:48] kenvandine: yes I am also in stress because of the basic auth shutdown [13:49] kenvandine: to be honest, if they dont allow the key to be used in the source code, I might consider to simply drop twitter support [13:49] meebey, that is what ryan wants to do in gwibber... but twitter is just to popular, and ubuntu users are going to want it. [13:49] kenvandine: I kind of doubt the packaging obfuscation will help, each distro needs to obtain their own key then? [13:50] meebey, yeah... the solution they are giving to open source developers is to drive each user to dev.twitter.com to register their own application and enter their own key [13:50] kenvandine: well, if enough apps drop twitter support they might change their stupid EULA [13:51] the packaging obfuscation is about as good as stuffing it into a binary that can just be disassembled [13:51] kenvandine: if the key would be passed at build time... [13:51] either way it just can't be as secure as they want it [13:51] kenvandine: only the packagers would need to [13:51] kenvandine: that could be an acceptable workaround [13:51] still lame, but acceptable [13:51] not build time, the current though is to stuff it into a file, maybe several files on the filesystem [13:51] and we reassemble the files to get the key at runtime [13:51] kenvandine: I would do build time [13:52] so at least people have to do a little digging to figure out how to put it together [13:52] kenvandine: use with an automake maro [13:52] --with-twitter=$apikey [13:52] yeah... but gwibber is python so it would end up in plain text somewhere [13:52] oh! [13:52] :) [13:52] I dont have that particular issue with C# [13:53] but ic, a buildtime var is not helping, it ends up as plaintext on disk [13:53] disc, hehe [13:54] yeah :) [13:54] so build time could help, if we built it into some sort of binary [13:54] as long as they are OK with it being included in the packaging as plain text [13:54] kenvandine: actually.... [13:54] kenvandine: libgwibber is C, why not that? [13:55] they definately are opposed to it being committed to the source repository [13:55] well libgwibber depends on gwibber-service [13:55] kenvandine: and at runtime gwibber reads (if needed) the api key from libgwibber [13:55] i don't want a circular depends [13:55] we could make gwibber build it into a tiny C module [13:55] yeah or that [13:56] something that is easy accessible from pyhton [13:56] they are discussing the idea of having the key in packaging but not in the source now [13:56] I don't understand why so many upstreams act downright hostile about libappindicator [13:56] "Until it appears in GNOME mainstream, no way we're supporting that" [13:56] lucidfox, yeah... i don't get it either [13:56] how backwards [13:57] good that there are also awesome upstreams like smuxi [13:57] * meebey hides [13:57] ^_^ [13:57] GNOME is not providing any solution to the tray mess [13:57] awesome upstreams! [13:57] so why would I expect them to solve it? [13:57] they didnt for the past 10 years [13:57] ... [13:58] * meebey tries to get smuxi 0.8 ready in time for maverick with all the ayatana sexyness [13:59] woot [13:59] and twitter is the showstopper on that roadmap :/ [13:59] maybe its time for a status.net replacement ^^ does it allow a brdige to twitter? [14:00] i think just posting [14:00] i haven't looked into it much [14:00] * meebey never used identica... [14:00] i want things like twitter lists and stuff [14:00] right, for full twitter support there is no way around it [14:00] Oh great, not bloody Twitter *here* too [14:00] but smuxi has no full blown twitter support yet, so... [14:00] haha [14:01] lucidfox went once crazy on #debian-cli because of using the word twitter [14:01] ... [14:01] hehe [14:52] mpt, Should we be puttin gin an override for the placement of the gsd keyboard layout switcher? [14:52] tedg, because it's not yet a system one? [14:53] mpt, Correct, so it gets place by category/id. [14:53] mpt, On my system that puts it all the way to the left. [14:53] tedg, it would make sense to override both the gsd one and the ibus one to put them at the trailing end [14:53] well spotted [14:54] mpt, Hmm, I would have guessed the opposite? Why the trailing, it seems like the should be next to power? [14:54] tedg, yes, next to the other system ones [14:55] mpt, Ah, cool. I thought you meant the other side ;) Any preference on the order of gsd, ibus, gpm ? [14:55] https://wiki.ubuntu.com/SoftwareCenter#def-trailing :-) [14:56] tedg, no preference [14:57] just as long as gpm isn't separating the other two. [14:57] mpt, Hmm, I would have said trailing is "closest to white space" as the one in the corner is anchored and everyone else is following. [14:57] It's Java GUI lingo, apparently [15:27] ronoc: [15:27] Cimi [15:28] ronoc: did ypu play with the branch? [15:28]