=== fginther is now known as fginther|away [04:19] Good morning === thumper is now known as thumper-afk === dpm-afk is now known as dpm === RAOF_ is now known as RAOF [08:06] good morning desktopers! [09:21] blarg [09:22] Laney, good morning [09:22] i rebooted the server [09:23] that'll solve everything [09:23] how are the pimms? [09:23] hi seb128! [09:23] hey ;-) [09:23] pimms is usually served with lemonade which means I don't like it :( [09:24] * Laney is a fizzy drink hater [09:24] oh [09:28] Laney, darkxst, others: robert_ancell has a look at making u-s-d include the old xrandr code, https://code.launchpad.net/~robert-ancell/unity-settings-daemon/xrandr/+merge/224548 if you fancy having a look/commenting [09:28] oh [09:28] is this more selective? [09:29] more selective than? [09:29] forking the whole thing [09:29] yes [09:29] it's just the xrandr part of gnome-desktop [09:29] yay [09:29] the lib does other things [09:29] like keyboard, thumbnail, etc [09:29] I read some conversation the other day about forking it [09:29] https://code.launchpad.net/~robert-ancell/unity-control-center/xrandr/+merge/224547 [09:29] is the u-c-c side [09:29] well [09:30] the options are/were still to either do what robert did there [09:30] or to ship gnome-desktop3.8 as a different source and build u-s-d/u-c-c with it [09:31] you think this approach might not work? [09:31] well, he mp the changes, so he got it to work [09:31] seems nicer to me [09:31] anyway I'll try it out [09:32] yeah [09:32] it just looked quite some code/renaming [09:32] and I was unsure if some of those functions would call other gnome-desktop functions and rely on those to not change as well [09:32] but it doesn't seem to be the case [09:32] so +1 from me [09:32] if it turns out to really be an issue over time we can still revisit the approach then [09:38] seb128, idle monitor is not included ? [09:39] darkxst, don't ask me, I didn't review it yet, robert_ancell said he would have a look and put those up for review this night [09:39] I'm just sharing the info [09:39] that would be a good question/comment to put on the merge request ;-) [09:42] darkxst: who uses that? [09:44] oh, u-s-d does, but it's not removed from gnome-desktop 3.12? [09:45] Laney, it was moved from gnome-desktop 3.10 into mutter [09:45] what's http://sources.debian.net/src/gnome-desktop3/3.12.1-1/libgnome-desktop/gnome-idle-monitor.h ? [09:46] Laney, the api is still in gnome-desktop, but changed to dbus calls into mutter [09:47] ah [09:47] then yes, please comment saying that [09:48] https://git.gnome.org/browse/gnome-desktop/commit/?id=545c774e967fa2f21f9b8ee1acbb50bffcee01d0 [09:51] Laney, ok, added a comment to MP [09:51] ty [09:52] thanks [09:52] we should bribe Trevinho into implementing this dbus service ;-) [09:52] isnt the whole point of copying the legacy code to avoid a dbus service? [09:58] I think it was more a concern about whether breaking it out would be stable enough [09:58] but having both implement the same interface would be okay, the issue there is lack of development resources [09:58] so this is the path of least resistance [09:58] AIUI anyway [10:01] right, my main concern was to add a piece of software that is new/can have bugs without anyone having slots to maintain it [10:01] if the unity team is wanting to allocate resources to implement that interface and maintain it, that would be great [10:01] but that's not the case atm and we didn't want to block Ubuntu GNOME to get the new gnome-desktop until that happens [10:05] seb128, I am happy for you guys to do it this way, but don't forget upstream are maintaining the new code (and the patches to de-couple it from mutter were pretty trivial) [10:06] we should maybe try that standalone daemon yeah... [10:06] I had concerns mostly for the LTS because we don't like to introduce new code/architectures just before a LTS [10:07] but we are still early in this cycle so we could try it out and see how it goes [10:08] seems like it would be nicer [10:09] darkxst: got a PPA or something? [10:13] pitti: what's up with these broken pipe adt failures I keep getting for chromium? [10:13] (also, am I really still TIL for that?) [10:14] Laney: it's a bug somewhere in the qemu runner, I already retried this [10:14] yeah [10:14] I already spent some two hours trying to reproduce and understand it, but so far no success :/ [10:14] I've had it quite a few times though [10:14] okay :( [10:15] Laney, so you maintain chromium and libreoffice now? [10:15] * Laney is the master of trivial diffs [10:15] tkamppeter: have you seen that cups has a failing autopkgtest? [10:16] aha maybe -6 fixes this? [10:16] looks like it will [10:17] Laney, https://github.com/darkxst/displayconfig [10:17] and one for mutter? [10:18] that code is ripped from mutter [10:18] yes [10:18] its the important bits that u-s-d/g-d need [10:18] presumably you need to turn it off there [10:19] if they plan to use the standalone service under gnome-shell, which I don't think they do? [10:19] that is supposed to be a replacement for !shell sessions [10:20] right, https://github.com/darkxst/displayconfig/blob/master/src/main.c#L381 [10:20] just quites if shell is running [10:20] quits [10:20] ah [10:20] you should rather check for the name on the bus [10:20] how is it activated? [10:20] dbus activation [10:20] what name? [10:20] in my "GNOME" !shell session I'd want this service too [10:21] unity steals the shell name on the bus [10:21] the name is org.gnome.Mutter.something [10:21] k, that we don't steal afaik [10:21] Laney, no, in GNOME !gnome3 session you want u-s-d [10:22] I have XDG_CURRENT_DESKTOP=gnome [10:22] so you can't use this to differentiate between shell and not [10:25] ok, but you can't use g-s-d anymore, if it even still works! [10:25] it works GREAT! [10:26] but yes I think the plan was to have me using usd [10:26] doesn't seem to have happened yet [10:26] Laney, I have already dropped most of the ubuntu legacy hacks from g-s-d [10:26] the only problem I've noticed is that brightness isn't restored when I come back from suspend [10:27] brightness api did change slightly, but think that was maybe in 3.10 [10:28] maybe you want to look into the upstart jobs to see which sessions get g and which get u [10:28] seems they need tweaking [10:31] okay lemme play with this displayconfig stuff [10:31] I'll give you a patch that makes the exit condition be if the name is already on the bus [10:32] Laney, have seen it, pitti told me, and OdyX has already fixed it in -6, which will auto-sync into Utopic soon. [10:33] yeah, saw [10:33] Laney ok [10:33] darkxst: did you give it an inactivity timeout? [10:34] the daemon? no [10:34] kay [10:35] I don't think thats a particularly good idea, it needs to run the entire session anyway [10:35] is -s-d using this API the whole time? [10:35] why does it need to run the entire session? [10:35] its pretty transparent to -s-d [10:35] it feels like that should only be needed when changes happen? [10:36] or is it tracking idle as well? [10:36] seb128, idle monitor! [10:36] k [10:36] then it'll never timeout ;-) [10:36] -s-d still calls into gnome-desktop [10:37] but gnome-desktop uses dbus calls [10:40] did you clean up configure.ac? [10:41] i.e. does it still need the x libs? [10:41] oh yes I see it [10:42] Laney, cleaned up most of the configure.ac [10:42] its still making X calls, so mostly needs x libs! [10:42] yeah duh [10:53] Laney, also see bug 1228765 [10:53] bug 1228765 in unity-control-center (Ubuntu) "[FFe] Implement DisplayConfig dbus interface and transition to gnome-desktop 3.10" [Low,In progress] https://launchpad.net/bugs/1228765 [10:53] some patches are required on u- side but they are all just cherry-picks [10:53] what about that last comment? [10:55] I never looked into that, however it was certainly working when I tested it months earlier === vrruiz_ is now known as rvr [10:57] my guess would be the deamon was not actually running [10:57] okay lemme try it [10:57] what cherry picks are these? [10:59] - git_xrandr_adapt_to_new_displayconfig_api.patch: [10:59] Adapt to new display config d-bus api used by gnome-desktop 3.10 [10:59] - fix-cursor-gnomedesktop-api.patch: [10:59] minor api fix [10:59] - git-power-update-backlight-api.patch: update gnome-desktop backlight [10:59] api [10:59] - git_power_dont_look_for_screens_no_lid.patch [10:59] - git-gnomerr-output-is-laptop-change.patch [10:59] - git_wacom_remove_deprecated_call.patch [10:59] - git_wacom_gnomerr-api-changes.patch [10:59] - git_xrandr_remove_applyConfiguration_dbus.patch [10:59] - git-power-simplify-support-external-monitor.patch [10:59] - git_color_adapt_to_gnomeRR_api.patch [11:00] or I should just say look at the braches attached to that bug! [11:00] you can just apply them directly [11:00] Laney, they are applied directly on the u- branches there [11:01] ok [11:01] well they should be [11:07] Laney, https://code.launchpad.net/~noskcaj/unity-settings-daemon/gnome-desktop-3.10/+merge/208911 might still apply without change [11:07] mostly [11:07] there's a conflict [11:12] Laney, hmm, no apparently I can't read, dialog is provided by gnome-shell, so that bit will need reverted to use g-s-d [11:13] (according to the bug comment atleast) === bschaefer_ is now known as bschaefer === oCrazyLem is now known as CrazyLemon [11:45] darkxst: no cursor, forgot my resolution :) [11:46] Laney, is it actually running? [11:46] after I opened the panel (presumably activating the service) they were restored [11:46] no cursor means, no idle monitor [11:46] but my monitors are the wrong way around [11:47] it uses the same monitors.xml [11:47] no dialog on apply [11:47] got a g-c-c crash [11:47] u-c-c [11:49] * Laney gets a trace === alan_g is now known as alan_g|lunch === MacSlow is now known as MacSlow|lunch [12:09] darkxst: comments for you ;-) [12:14] Laney, odd, dbus activation should fire on login [12:16] it's probably easier if you/jackson just try it at this point [12:17] Laney, yes I will try it, but won't be until the weekend, possible Jackson messed up the merges [12:17] or me, there was a conflict [12:18] or something changed since the branch was proposed and the merge is borked now [12:18] (i.e. it needs rebasing) [12:18] or 3.12 is bad and 3.10 would have worked [12:18] havent tested with 3.12 [12:18] the dialog thing just needs doing though [12:19] yes I will look at that dialog [12:19] anyway there we go ;-) [12:22] Laney, I never saw anything like your crash in testing, probably 3.12 related [12:22] I will run through and get it all working over the weekend [12:23] it could be something that expects the service to be running [12:23] I just saw it in the background, don't know when it happened [12:23] but cool, thanks [12:23] * Laney goes back to good old utopic for now :P [12:24] Laney, well atleast assuming you guys want to use it [12:25] * Laney is pro [12:25] sounded like seb was too [12:25] so thanks for working on it === fginther|away is now known as fginther [12:33] Laney, ok, I won't have time to polish it right up, but I will rebase my work and fix any obvious problems === MacSlow|lunch is now known as MacSlow === alan_g|lunch is now known as alan_g [14:18] bregma: hey! [14:19] bregma: we discussed the UI scaling stuff in Malta; I've been using a scaling factor since (previously I Was just xrandr-ing to a lesser resolution and leaving the Ubuntu defaults alone) [14:20] bregma: It seems to me as if the scaling factor is applied over and over again on each boot; it feels like the fonts are getting bigger over time; now it's to the point I had to reduce the scaling factor under 1 and I get a mix of decent fonts and small UI elements; could it be that something is accumulating the factor somewhere? [14:20] bregma: is this known and/or reproducible for you? === Estilanda_ is now known as Estilanda === alan_g is now known as alan_g|tea [14:38] lool, yes, it's known and I get it all the time, it's caused by a race condition somewhere and we're trying to track it down === alan_g|tea is now known as alan_g === m_conley_away is now known as m_conley [15:04] TheMuso: someone in another channel just linked http://askubuntu.com/questions/278693/how-do-i-stop-orca-screen-reader which notes that it's quite hard to switch orca off if you start it from the desktop file [15:05] looks like most things look for the screen-reader-enabled gsettings key which isn't set on if you start it manually [15:09] so you have to toggle it on and off to make orca quit (that's all it looks for to shutdown afaics) [16:56] pitti: have you thought about testing reverse-build-depends or reverse-test-depends on upload? [16:56] I just saw that pep8 made some stuff fail but they're generally not runtime dependencies [16:57] I guess test rather than build [16:57] or build for build-needed maybe === alan_g is now known as alan_g|EOD [17:06] * didrocks waves good evening [17:06] I have to buy beers for seb128 [17:06] question of requirements :) [17:06] haha [17:07] Get him on the hard stuff [17:07] well, he's not really German for that [17:07] he only likes light one [17:07] it's a shame! [17:07] * didrocks get out the sake [17:07] gets* [17:08] so, *maybe* see you tomorrow after this evening ;) [17:08] * didrocks really waves good evening now [17:08] best of luck! [17:08] man, that guy's fast === dpm is now known as dpm-afk === maxb_ is now known as maxb [22:07] Laney: Thats purely upstream... [22:09] Laney: I.e thatts a decision that was made upstream. Given the way Orca is supposed to be activated these days, I wonder if it makes sense to drop /usr/share/applications/orca.desktop. === oCrazyLemn is now known as CrazyLemon === m_conley is now known as m_conley_away === larsu_ is now known as larsu === oCrazyLemn is now known as CrazyLemon