=== JanC_ is now known as JanC === maclin1 is now known as maclin [06:21] good morning [06:35] Morning didrocks [06:35] hey duflu [06:37] good morning desktopers, and happy friday [06:37] hey duflu, how are you? [06:37] re didrocks [06:37] Happy Friday seb128 (he says while balancing a heat pack on his sore head) [06:37] How are you? [06:38] re seb128 [06:39] I'm good, it's friday, weather is nice (sunny but not too hot) and I'm playing tennis tonight! [06:51] duflu, bug #1705369 ... can we reliably tell in a video card/gpu is connected to something? [06:51] bug 1705369 in gdm3 (Ubuntu) "Ubuntu 17.10 boots to black screen when using Nvidia drivers (on a desktop with an Intel GPU)" [High,Triaged] https://launchpad.net/bugs/1705369 [06:51] morning all [06:51] hey willcooke! how are you? [06:51] seb128, yes we did it in Mir, and any display server has to do it. Just takes effort [06:51] Hi willcooke [06:52] duflu, ok, I was unsure if the system always had enough informations from the card/driver to tell e.g if there is a monitor connect to the vga output [06:53] we did it as well in Nux [06:53] maybe just blacklist unconditionnally nvidia drivers? [06:53] I mean the display server does what it can with the info it has [06:53] hey willcooke [06:54] didrocks, the dual gpu case is a bit tricky [06:54] yeah, I think [06:54] didrocks, I added MATE to that list [06:54] like "which one took over" [06:54] that list = trello card for themes [06:54] willcooke: you didn't lie, clicked on the tab, so it added :p [06:54] willcooke: good one, if you come on with any other… [06:55] s/so it/was/ [06:55] who knew that themes would be a so trendy topic in 2017 [06:55] it's the nice thing! [06:55] didrocks, I'll have a think :) Also, I'm just rebuilding that snap having moved it to the desktop launcher [06:55] now, about that input system which needs theming… :) [06:55] willcooke: ah, you are handling the cleanup of that snap? [06:56] :-p [06:56] so, moving to the cloud parts [06:56] seb128, it's "easy". Just standard KMS logic for detecting if there is a monitor connected [06:56] didrocks, welllllllllllll. Lemme see if I can make it work first - I might have questions [06:56] sure :) [06:56] willcooke: once we'll add all the themes, think about running "snapcraft update" or snapcraft will never pick the new definition [06:56] didrocks, ack [06:56] (but if we changed the remote code, it will pick it, which has interesting side-effects) [06:56] and thanks for looking at that didrocks, seb128 [06:56] snapcraft should really just tell you to refresh? [06:57] …* [06:57] yw! [06:57] np! [06:57] didrocks, how does it tell that code has changed? [06:57] seb128: it doesn't, but source: https://github.com/… [06:58] so, it always pull [06:58] however, the part definition is cached [06:58] I had the issue once when I changed the remote code (the make parameters to be passed) and thus, the parts definition [06:58] got a lot of bug reports because people were rebuilding their snap without refreshing the remote part [06:59] and so, failure to build [06:59] right [06:59] it's a bit tricky [06:59] yep [06:59] it should cache everything, included pointed resources [06:59] or nothing [06:59] not half of the thing [06:59] hitting the internet at every build to check if the parts are outdated would be suboptimal as well [07:00] yeah, that's why caching more could be an answer [07:00] then, it can hit the Internet, in the background [07:00] so that next command tells you "you have new stuff to refresh" [07:00] that would be an improvement already IMHO [07:00] log a bug, I'm sure Sergio will help [07:01] willcooke: it's logged, discussed [07:01] on the backlog list [07:01] oh cool [07:01] we even discussed it last october again IIRC [07:01] I'll go speak to Sergio and see if he'll do us a solid ;) [07:02] willcooke: that's why, even if we add those new themes, people not using the platform snap and rebuilding a snap locally (we won't have that issue on b.snapcraft.io) may still miss the themes in [07:02] I prefer to be explicit about gotchas we can encounter :) [07:14] Can someone in an Xorg gnome-shell recreate this? https://bugs.launchpad.net/ubuntu/+source/gnome-shell/+bug/1705458 [07:14] Ubuntu bug 1705458 in gnome-shell (Ubuntu) "Mouse clicks pass through to window underneath" [Undecided,New] [07:14] It feels like one of those "only me" bugs again [07:16] willcooke: can't reproduce, indeed [07:16] :) [07:17] however, I have some other interesting bug [07:17] like, if I reduce the window [07:17] the popup goes a little bit over [07:17] on the bottom [07:17] but what's after is cut [07:18] I guess it's because the popupover isn't its own X window [07:23] bah, today it's gone. Even though I could re-create it after reboots yesterday [07:25] seb128, Laney - polite reminder, could you update the newsletter GDoc today please? I can post to Wordpress if needed, but Laney also has an account I think [07:30] didrocks, confirmed I think: http://imgur.com/a/4FgtF [07:30] good morning [07:31] talking about weird bugs, bug 1705459 is annoying. Has anyone seen it with another app? [07:31] bug 1705459 in gnome-shell (Ubuntu) "Cannot switch to another workspace when an sqlitebrowser window is open" [Undecided,New] https://launchpad.net/bugs/1705459 [07:31] hey andyrock [07:31] willcooke: yeah [07:31] willcooke: I think it's the X window size at max (+ transparent borders) [07:32] jibel: never (yet) [07:43] didrocks, logged: https://bugs.launchpad.net/ubuntu/+source/gnome-shell/+bug/1705647 [07:43] Ubuntu bug 1705647 in gnome-shell (Ubuntu) "File Open popover only shows half the popover" [Undecided,New] [07:45] Morning desktopers [07:46] hey andyrock jibel flexiondotorg [07:48] (confirmed) [07:56] morning seb128 [08:01] beeeeeeeeee gooooooooooooooooooooooooooooooooooodddddd [08:02] heyyyyyyyyyyyyyyy Laneyyyyyyyyyyyyyyyyy [08:02] goooooood moooorrrning Laney [08:05] https://www.youtube.com/watch?v=Dqtxl4Qc_7A [08:08] heh [08:08] Laney, how are you today? [08:14] hey seb128 [08:14] good thanks, happy it's friday! [08:14] you? [08:15] same, friday and tennis day! [08:16] seb128 Federer [08:18] haha [08:27] morning guys [08:28] I need your publishing powers for https://bileto.ubuntu.com/#/ticket/2874 [08:29] hey Trevinho [08:29] hey seb128 how are you? [08:29] the guy who only says hello to us when he needs something :-) [08:29] Trevinho, I'm good! you? [08:30] Good... Working in a seaside pine forest today :) [08:30] nice [08:30] Trevinho, what's that .upgrade doing? [08:32] seb128: it's updating the default settings for an user that has already installed unity [08:33] so basically if the profile has already been created the settings are updated [08:33] basically a migration script [08:33] compiz-style [08:33] k [08:34] Trevinho, published [08:34] ta [08:34] yw [08:51] could somebody try this to me on artful [08:52] on the gdm screen (either before log in, or lock screen and click "use different user") [08:52] type a wrong password and when the error is displayed hit "esc" directly on your keyboard while it's still on screen [08:53] do you end up with a password entry on top of the user description as well? [08:54] seb128: checking [08:55] seb128: yep [08:56] andyrock, thanks [08:58] jamesh, nice snapcraft post about desktop needs! [08:58] seb128: thanks. Is there anything you'd like me to add? [08:59] I tried to start each task with what we need rather than railroad it to a particular solution. [09:01] jamesh, I think you covered the important topics, I'm looking at https://bugs.launchpad.net/ubuntu//+bugs?field.tag=snap-desktop-issue but I'm not sure other things are worth mentioning [09:02] jamesh, we eventually need to resolve bug #1583250 at some point, not sure if that's worth putting on the list though [09:02] bug 1583250 in Snapcraft "upstream use of build-time defined PREFIX or DATADIR incompatible with snaps relocation" [High,Triaged] https://launchpad.net/bugs/1583250 [09:02] jamesh: yeah, that one needs to be listed in ^ [09:02] jamesh, we currently build our GNOME snaps with an hacked prefix=/snap/$name/current/usr and use organize [09:03] to move things back to the right location [09:03] I heard that snap location are different on other distros [09:03] so the same snap would fail [09:03] but that makes things depends on /snap to be the location [09:03] better to have it said before :) [09:03] which seems is not true on other distros [09:05] seb128, didrocks: I'll add that. [09:06] thx! :) [09:07] jamesh, thanks [09:15] didrocks: https://forum.snapcraft.io/t/ubuntu-desktop-team-priorities/1397 <- does that look okay? [09:16] jamesh: on the themes, you can maybe reference our google doc? [09:16] sure. [09:16] thx [09:17] jamesh, on the relocation the snappy team position is that "/snap" is an implementation detail and could change and that we shouldn't rely on it [09:17] and it's change on e.g fedora I think [09:17] because it's against the fhs and they put it under /run or something [09:19] but the "prefix" hack which is required with many upstream projects were the only recommendation they can give at the time [09:19] seb128: maybe we suggest people get their app working with strict confinement then? :) [09:19] who cares where the snap appears on the host file system if it is at /snap inside the sandbox? [09:20] (or is it also under /run in the sandbox on Fedora?) [09:20] jamesh, you mean? [09:20] jamesh: it's not in /snap, it's in /snap//current/ [09:20] on ubuntu [09:20] didrocks, which is under /snap :p [09:20] and somewhere else on other distros [09:20] jamesh, I don't think the sandbox does magic to remap it as /snap [09:20] correct, but still depends on snap_name and path [09:21] didrocks, seb128 was saying that snapd on fedora doesn't put snaps under /snap//current. I was wondering whether that was only on the host file system, or also in the sandbox [09:21] didrocks: since a strict confined snap won't see the host file system anyway [09:21] jamesh: ah, I guess we only talk about the sanbdox [09:21] jamesh: well, snap isn't like that [09:21] if you snap run --shell [09:21] you will see /snap/* [09:21] you just can't access to other directories [09:22] apart from /snap// [09:23] didrocks, well, the question is to know if fedora patches snappy to not create a /snappy on the fs but use /run/snappy what is the dir in the sandboxes env? [09:23] is that /run/snappy as well [09:23] or is the sandbox relocating to /snap? [09:23] only zygmunt can answer to that I guess [09:23] right [09:24] because if the sandbox relocates than the prefix to /snap works [09:24] for anything running apparmor, what you see outside is what yo uhave inside [09:24] due to how the tech work [09:24] (paths should be the same, apparmor is outside the sandbox and path-based) [09:24] I don't know if they have clever tricks for fedora [09:25] I guess not === carlolo is now known as clobrano [09:33] willcooke: mind looking at https://trello.com/c/9cYd1a02/201-add-more-themes-in-desktop-part-and-gnome-platform last comment and giving a ack/nack? [09:36] didrocks, wow. I thought it would be bigger, but not that much bigger. [09:37] currently its using my laptop's internal mic [09:37] So, I suggest we keep a base version as we have, and an "expanded" one which you've just built? [09:37] oops [09:38] willcooke: hum [09:38] willcooke: then, people will complain "my snap does work, no this one doesn't" [09:38] hey willcooke :) [09:38] and we will keep getting into those discussions [09:39] so it's either we add them, or none [09:39] IMHO [09:39] seb128: thoughts? ^ [09:39] didrocks, thats a good point, but "all" seems too big to me [09:39] willcooke: we took the popular ones, not all [09:39] mate alone is a +35M [09:39] willcooke, didrocks, it's a "can't win" situation [09:40] +1 [09:40] willcooke, having 2 versions put us back where we are today, you can bundle thems but snap author don't do it [09:40] and so, those tweets will keep coming [09:40] the equivalent post change is going to be "they picked the version with less depends to avoid the cost" [09:40] and we still are in the same situation [09:40] exactly [09:41] this is the reason why we argued for theme snaps with seb128 to be the proper solution [09:41] which requires what is on our gdoc to be implemented [09:43] this is the additional list we took: http://paste.ubuntu.com/25138289/ [09:43] you can see there isn't even elementary theme [09:44] (so it's +9 themes compared to what we have today) [09:50] kk [09:51] so I feel like that extra size is probably too much, wdyt? [09:51] * didrocks feels it too, also increased build time [09:51] willcooke, size is a bit much but I would rather be concerned by the triple build time [09:52] ack [09:52] platform snap would be ok IMHO as it's only in one spot, BUT I feel the inconsistency between "bundling everything" experience and using the platform will give strange messages [09:52] we could find a way to improve the build time issue [09:52] but not the size one [09:52] (at extra maintenance cost) [09:53] So how about just: Mate, arc, numix? [09:53] the first one is 35M by itself [09:53] until we get to the next request? [09:54] humm, maybe not worth it then [09:54] which was already what happened when shipping ambiance/radiance + adwaita dark & white [09:54] Hello all :) ! [09:54] hey alexarnaud [09:54] morning alexarnaud [09:54] seb128, didrocks: how are you? [09:54] alexarnaud: good, thanks, yourself? [09:54] good, you? [09:54] good :) [09:58] willcooke: so, we just archive that card for now? I can push to a branch if needed. [09:58] I've a little question. Hui Wang tells on the pulseaudio-discuss ML that Ubuntu has a GUI to handle jack selection when the jack plug works for input and output. What is the name of the GUI? Where can I find it ? [09:58] https://www.mail-archive.com/pulseaudio-discuss@lists.freedesktop.org/msg17876.html [09:58] didrocks, oki thanks for looking in to it - useful stats [10:01] alexarnaud, i also have realtek sound and a dell alienware14 laptop [10:03] upon plugging in the headphone jack no sound output is present [10:03] but upon plugging in the headset i get full bass filed soundoutput [10:04] immu: do you see the pop-up? Do you know its name? [10:04] nope i don't see other then a popup sound bar type notification [10:05] none i just tried [10:05] in windows i get asked what have i plugged in [10:07] immu: As I understand you don't have the wanted behavior. [10:07] https://www.mail-archive.com/pulseaudio-discuss@lists.freedesktop.org/msg17902.html [10:07] What is the process to discuss technical stuff with Hui Wang ? [10:08] alexarnaud, email him? [10:08] or report a bug on launchpad [10:09] i had filed a bug but after something the discussion dried up [10:09] immu: are you on the pulseaudio-discess ML? [10:09] Realtek ALC3661 [10:09] seb128: thanks, good idea. [10:09] alexarnaud, Realtek ALC3661 [10:10] i had filed a bug on launchpad and bugzillaaa [10:11] alexarnaud, the corresponding code is https://code.launchpad.net/~diwic/unity-settings-daemon/what-did-you-plug-in/+merge/206915 [10:11] seb128: thanks :) ! [10:12] alexarnaud, GNOME has a similar feature but different implementation so things are worth re-testing on artful [10:12] alexarnaud, yw [10:13] alexarnaud, http://bazaar.launchpad.net/~unity-settings-daemon-team/unity-settings-daemon/trunk/files/head:/plugins/media-keys/what-did-you-plug-in/ [10:14] thanks again :). [10:17] didrocks: no more GDocs tutorials, everything has been converted to MD, just merged the last one \o/ [10:18] davidcalle: excellent! My md parser didn't need any change thus? [10:18] or missing features ? [10:18] compared to the gdoc one [10:20] didrocks: I think we have a (reasonably small) list of bugs yet to file (mostly line breaks handling I think, but we discussed that already), and no missing features, no [10:20] sweet! [10:20] the line breaks handling was a tradeoff if you remember our discussion [10:20] I do, and I explained it to the rest of the docs team [10:20] good :) [10:23] alexarnaud, do you have ALC3661 [10:23] ? [12:19] seb128 Thanks for helping alexarnaud. He is discussing this feature with MATE developers. [12:19] seb128 alex has a specific interest in a11y and this would be a useful feature for his organisation :-) [13:16] flexiondotorg, yw [14:24] kenvandine, hey, did the gedit snap got published? [14:26] * kenvandine double check [14:26] +s [14:27] seb128, whoops... i had forgotten to publish it after the manual review was approved :) [14:27] seb128, it's in edge now [14:27] kenvandine, thanks [14:27] seb128, i'm going to push them all to stable today [14:27] after a round of testing [14:28] nice [14:28] what is the list of apps you published? [14:30] gedit, gnome-sudoku, quadrapassel, gnome-dictionary, gnome-calculator, and gnome-clocks [14:30] working on gnome-maps now [14:30] also have wip of gnome-contacts [14:30] folks issues iirc [14:31] i had to backport libchamplain this morning for gnome-maps [14:33] thanks [15:11] seb128, any opinions about adding gjs to the platform snap? [15:15] All, I am trying to merge patches to gnome-initial-setup and gnome-control-center for supporting encrypted home. So far, no one has commented on the patches. [15:15] Please take a look and comment here: https://bugs.launchpad.net/ubuntu/+source/accountsservice/+bug/1699216 [15:15] Ubuntu bug 1699216 in gnome-initial-setup (Ubuntu) "Encrypted home support" [Wishlist,Confirmed] === czajkows1i is now known as czajkowski === JanC is now known as Guest32366 [17:16] night all [17:19] kenvandine, I think I would be -1 on gjs but need to look at what GNOME does and think a bit more about it [17:19] night Laney & desktopers [17:21] seb128, i had avoided adding gjs