| oSoMoN | good morning desktoppers | 06:30 |
|---|---|---|
| willcooke | jibel, can you see if you can recreate https://bugs.launchpad.net/ubuntu/+source/gdm3/+bug/1710637 | 06:30 |
| ubot5 | Launchpad bug 1710637 in systemd (Ubuntu Xenial) "Input falls through to gdm3 and terminates the session on Ctrl+C after udevadm trigger is executed under wayland" [Undecided,New] | 06:30 |
| jibel | Hi willcooke | 06:34 |
| jibel | willcooke, I'll have a look. On which release? xenial? | 06:35 |
| jibel | good morning oSoMoN | 06:35 |
| oSoMoN | salut jibel, ça va? | 06:38 |
| jibel | oSoMoN, ça va bien. Trop de pluie ce w-e en Normandie donc je suis allé chercher le soleil du coté de chez toi | 06:40 |
| didrocks | good morning | 06:50 |
| oSoMoN | salut didrocks | 06:51 |
| oSoMoN | jibel, tu étais où? | 06:51 |
| didrocks | salut oSoMoN ! | 06:52 |
| jibel | oSoMoN, un peu plus au nord que toi à Cadaques | 06:52 |
| didrocks | et jibel | 06:52 |
| jibel | salut didrocks | 06:52 |
| oSoMoN | c’est joli Cadaqués | 06:53 |
| duflu | jibel, it's 18.04 (bug 1767918) | 07:13 |
| ubot5 | bug 1767918 in gdm3 (Ubuntu) "Login password is shown in plain text when shutting down " [High,New] https://launchpad.net/bugs/1767918 | 07:13 |
| duflu | Morning oSoMoN, willcooke, jibel didrocks | 07:13 |
| didrocks | hey duflu | 07:16 |
| oSoMoN | good afternoon duflu | 07:19 |
| seb128 | good morning desktopers | 07:32 |
| willcooke | morning all | 07:32 |
| duflu | Hi seb128, willcooke | 07:39 |
| didrocks | hey seb128, willcooke | 07:39 |
| duflu | Do you want to do bluetooth today? (and koza?) | 07:39 |
| seb128 | duflu, hey, I skip it's best to skip | 07:41 |
| seb128 | hey didrocks | 07:42 |
| duflu | I also skip so | 07:42 |
| seb128 | +1 | 07:42 |
| duflu | Since cosmic is emerging we might have a 5.49 on the way next week... | 07:42 |
| seb128 | duflu, the ubiquity issue, I'm not sure it started with/is an ubiquity issue, e.u.c doesn't make it easy for us to know if those started after a gtk or webkitgtk update | 07:42 |
| duflu | seb128, I know. I've been on it all day. Narrowing down to some failing access control code in the Xorg server (or maybe kernel) | 07:43 |
| seb128 | I saw your comments | 07:43 |
| duflu | Interesting 17.10 does 200% scale fine. So that's not really new, just a trigger | 07:44 |
| duflu | seb128, BTW I would love to bisect daily images. Do we keep the old ones anywhere? | 07:46 |
| jibel | we don't | 07:46 |
| duflu | I've asked this in the past | 07:46 |
| duflu | :( | 07:46 |
| jibel | We used to keep them. I'll do it for C | 07:47 |
| jibel | just needs some disk space | 07:47 |
| seb128 | duflu, i'm a beta2 iso, I'm going to try on it | 07:47 |
| seb128 | I've* | 07:47 |
| duflu | jibel, you're my hero | 07:47 |
| seb128 | duflu, did you try to make an easier "test case", like trying to display the slideshow in a standalone pywebkitgtk process? if not I'm going to try to play with that | 07:48 |
| seb128 | easier than having to go through the installer every time | 07:49 |
| duflu | seb128, I looked and could not figure out how to run it locally | 07:49 |
| jibel | duflu, there is a script called test-slideshow.sh to test the slideshow at its name suggests. I don't know if it would be enough to reporduce the bug | 07:51 |
| jibel | it's in the source package | 07:51 |
| jibel | hm, it uses zenity so probably not the right tool | 07:52 |
| jibel | ah no, it may work zenity is only to select which flavor then it's a webkit view | 07:54 |
| duflu | jibel, which source package? Not ubiquity | 07:56 |
| duflu | Oh, maybe ubiquity-slideshow-ubuntu | 07:57 |
| seb128 | yes | 07:58 |
| duflu | I don't think that script is working right now | 08:01 |
| seb128 | no it's not | 08:03 |
| Laney | moin | 08:19 |
| seb128 | hey Laney, how are you? | 08:20 |
| didrocks | hey hey Laney | 08:20 |
| seb128 | duflu, I fixed that script but doesn't trigger the issue on a lowdpi machine with GDK_SCALE=2 | 08:27 |
| GunnarHj | Good morning, and a Happy New Cycle! | 08:27 |
| seb128 | hey GunnarHj | 08:27 |
| duflu | seb128, does it look different at all? I noticed that var didn't work when I tried it | 08:27 |
| duflu | Happy New Cycle GunnarHj | 08:27 |
| seb128 | yes, it's much bigger | 08:27 |
| seb128 | new cycle? stop the crazy talk | 08:27 |
| GunnarHj | ;) | 08:28 |
| seb128 | we have plenty of LTS work still to do :p | 08:28 |
| duflu | seb128, for some years | 08:28 |
| duflu | Hence the "LTS" | 08:28 |
| seb128 | right, but we should focus/be full time on that for a while | 08:29 |
| seb128 | before starting with new crazyness | 08:29 |
| GunnarHj | seb128: Should we stop im-config from setting GTK_IM_MODULE on GNOME? Bug #1761554 and bug #1760818 speak for that. So does bug #1760308 even if another fix was applied for the latter. | 08:29 |
| ubot5 | bug 1761554 in gnome-shell (Ubuntu Bionic) "[bionic] Extended characters in GNOME screen keyboard don't get entered" [Medium,Triaged] https://launchpad.net/bugs/1761554 | 08:29 |
| ubot5 | bug 1760818 in mutter (Ubuntu) "gedit and gnome-calculator transparency/graphics corruption issue" [Undecided,Confirmed] https://launchpad.net/bugs/1760818 | 08:29 |
| ubot5 | bug 1760308 in ibus (Ubuntu Bionic) "Unicode input not working on Bionic with CTRL-SHIFT-U" [High,Fix released] https://launchpad.net/bugs/1760308 | 08:29 |
| seb128 | GunnarHj, I don't understand what's the difference between GTK_IM_MODULE=ibus and it unset | 08:30 |
| seb128 | so I can't comment until I understand what that means exactly | 08:30 |
| duflu | seb128, try running the script with sudo | 08:30 |
| GunnarHj | seb128: My understanding is that when it's unset, configuration happens dynamically by GNOME somehow. After all GNOME does not set it. | 08:34 |
| seb128 | well I would expect they do load some immodule though | 08:34 |
| seb128 | so it's not the ibus one they use | 08:34 |
| seb128 | if not which one? | 08:34 |
| duflu | seb128, can you please share the script fix? | 08:35 |
| seb128 | duflu, yeah, I'm working on it, a min | 08:36 |
| duflu | Or propose it :) | 08:36 |
| GunnarHj | seb128: They set always "QT_IM_MODULE=ibus" and "XMODIFIERS=@im=ibus". | 08:36 |
| Laney | hey seb128 didrocks GunnarHj duflu | 08:38 |
| Laney | doing alright here thanks | 08:38 |
| Laney | finally going back home later | 08:38 |
| duflu | Hi Laney | 08:38 |
| duflu | seb128, btw I get the feeling the Xorg error happens server-side before the slide show starts. So maybe the test won't work | 08:39 |
| GunnarHj | seb128: Probably "XMODIFIERS=@im=ibus" is sufficient in the GNOME environment to do the right thing for GTK apps. | 08:40 |
| seb128 | duflu, http://paste.ubuntu.com/p/hb5fCCY9qb/ is my local tweaks | 08:41 |
| seb128 | GunnarHj, weird | 08:42 |
| seb128 | Laney, did you go back over the w.e or just stay in London? | 08:42 |
| seb128 | duflu, GDK_SCALE=2 sudo -E ./test-slideshow.sh works | 08:43 |
| seb128 | displays a scaled UI without XError | 08:43 |
| seb128 | but maybe GDK_SCALE is not enough for webkitgtk/not the right way | 08:44 |
| seb128 | would be interesting to know if that reproduce on your hidpi machine | 08:44 |
| Laney | seb128: stayed here | 08:44 |
| duflu | seb128, yeah the euid of the gtk_ui process changed between 17.10 and 18.04 and it looks like Xorg is manually checking that | 08:44 |
| seb128 | Laney, you didn't do any ubiquity test on your hidpi machine before release? | 08:44 |
| seb128 | duflu, ? | 08:45 |
| Laney | I didn't do an install | 08:45 |
| seb128 | :/ | 08:45 |
| Laney | but I did run ubiquity some times | 08:45 |
| Laney | probably not far enough? | 08:45 |
| Laney | I also did do test-slideshow | 08:45 |
| seb128 | Laney, looks like the installer XError on hidpi machines in the slideshow | 08:45 |
| seb128 | that's the suck | 08:45 |
| Laney | sounds bad | 08:46 |
| seb128 | can't believe we didn't do at least one test install on an hidpi machine :/ | 08:48 |
| duflu | seb128, we definitely did. Many times. Just on the wrong days | 08:48 |
| duflu | With earlier images | 08:48 |
| seb128 | well, I wonder at which point it changes/started being buggy | 08:49 |
| seb128 | changed | 08:49 |
| duflu | seb128, yeah that's why I would like to be able to bisect daily images, all 6 months worth | 08:50 |
| seb128 | could be the slideshow content change, or a gtk/webkit update or something else | 08:50 |
| seb128 | duflu, did you try my patch to the standalone slideshow test script? | 08:50 |
| seb128 | does it xerror for you? | 08:50 |
| duflu | seb128, I will get to that. Still doing other tests | 08:50 |
| seb128 | k | 08:51 |
| didrocks | jibel: so, if I do some changes to do-release-upgrade and want to test it before pushing to proposed, it sounds like there is no option to disable local download of the tarball of new upgrader and so it means commenting the call to fetchDistUpgrader() manually, do you know of any other tips? | 08:52 |
| xnox | Laney, do you, per chance, have ubiquity changelog finalised & 18.04.14 tag committed locally and could you please bzr push? or shall I just tag and push? | 08:53 |
| Laney | hey xnox | 08:53 |
| Laney | probably, let me see | 08:53 |
| Laney | there you go | 08:53 |
| jibel | didrocks, let me check | 09:04 |
| didrocks | jibel: I think I can just generate directly the "bionic" wrapper | 09:04 |
| didrocks | which bypass the first stage (download/reexec) | 09:05 |
| didrocks | meaning, just executing DistUpgrade.DistUpgradeMain.main() directly | 09:05 |
| jibel | didrocks, yes you can do that | 09:05 |
| didrocks | will do! Thanks for confirming jibel :) | 09:06 |
| jibel | didrocks, for testing, I usually download the tarball and run the wrapper | 09:07 |
| didrocks | yeah, but for iterating, if I can avoid downloading the tarball, applying the diff each time and such… | 09:07 |
| seb128 | duflu, I tried editing /usr/lib/ubiquity/ubiquity/fronted/gtk_ui.py and deleted the "while(self.pagesindex < len(self.pages)):" section | 09:14 |
| seb128 | so ubiquity directly starts on the slideshow | 09:14 |
| seb128 | no xerror even with a scale of 2 | 09:14 |
| seb128 | so I don't know if that works around it or just if I can't reproduce on a lowdpi config even enabling scaling | 09:15 |
| jibel | duflu, I cannot reproduce bug 1767918. I see the console for a tenth of a second but not enough to see and read any password | 09:17 |
| ubot5 | bug 1767918 in gdm3 (Ubuntu) "Login password is shown in plain text when shutting down " [High,New] https://launchpad.net/bugs/1767918 | 09:17 |
| duflu | jibel, no problem for now. Just keep it in mind. If one bionic hits that bug then others will follow I guess | 09:17 |
| seb128 | duflu, when you say that the uid of the process changed between 17.10 and 18.04, of what process is that? we shouldn't have an uid change between those... | 09:20 |
| duflu | seb128, the python3 process ending in "gtk_ui". Seems to be dropping perms now, which I think is confusing Xorg | 09:21 |
| seb128 | ubiquity does some drop/restore privilege, the slideshow is in a drop privilege section but it has been since xenial | 09:22 |
| seb128 | def start_slideshow(self): | 09:22 |
| seb128 | # WebKit2 spawns a process which we don't want to run as root | 09:22 |
| seb128 | misc.drop_privileges_save() | 09:22 |
| seb128 | in gtk_ui.py | 09:23 |
| seb128 | also if the issue was the uid why would it be only a problem when scaling? | 09:24 |
| duflu | seb128, yeah that last part makes no sense | 09:24 |
| duflu | Side-effects | 09:24 |
| duflu | seb128, also scaling worked in 17.10 | 09:26 |
| Trevinho | Morning... | 09:33 |
| * Trevinho back in the cold Europe | 09:34 | |
| duflu | Morning Trevinho. Oh noes, the comedown begins | 09:34 |
| duflu | Oh nose, even | 09:36 |
| seb128 | hey Trevinho, back to the old world? | 09:38 |
| seb128 | how are you? | 09:38 |
| Trevinho | seb128: hey seb... all good, you? | 09:38 |
| seb128 | I'm good thanks | 09:38 |
| seb128 | duflu, let me know if you have a chance to test if the standalone script or the gtk_ui.py hack to go directly to slideshow | 09:39 |
| seb128 | I would like to know if those don't hit the issue or if it's not as simple as enabling scaling | 09:39 |
| duflu | seb128, about 2 seconds before you asked, I did. And it works thanks | 09:39 |
| seb128 | "works" as trigger the bug? which one(s)? | 09:40 |
| duflu | seb128, no, no bug | 09:40 |
| seb128 | :( | 09:40 |
| didrocks | good morning Trevinho :) | 09:41 |
| Trevinho | didrocks: hey! :-) | 09:51 |
| seb128 | duflu, did you also try to edit gtk_ui.py to delete the pages section so it directly goes to the slideshow? | 09:52 |
| seb128 | duflu, I wonder if something is happening async and taking more time when scalling and the callback by the time we are out of the drop_privilege section | 09:55 |
| seb128 | but it's weird | 09:55 |
| duflu | seb128, that suggests sync at 100% might trigger it. I will try | 09:56 |
| duflu | Then I will make dinner | 09:56 |
| duflu | Also, more comments in the bug | 09:57 |
| seb128 | "sync at 100%"? | 09:59 |
| seb128 | thx | 09:59 |
| duflu | seb128, Yes. You win. It crashes at scale 100% if you slow it down with GDK_SYNCHRONIZE=1 | 10:01 |
| duflu | So just a race | 10:01 |
| seb128 | ah | 10:02 |
| seb128 | did you try full install? | 10:02 |
| seb128 | or did you hack to skip the pages? | 10:02 |
| duflu | seb128, that's a full install. I can reproduce the bug at low DPI using that | 10:03 |
| duflu | Although "full install" failed | 10:04 |
| duflu | So it sounds like a race between Xorg and the perms dropping | 10:04 |
| duflu | which then breaks Xorg's ability to authenticate XShmAttach | 10:04 |
| seb128 | k, thanks | 10:05 |
| duflu | seb128, I hope to see a fix tomorrow ;) | 10:06 |
| seb128 | haha | 10:06 |
| seb128 | let's see :p | 10:06 |
| duflu | We all do | 10:06 |
| duflu | Night... | 10:09 |
| seb128 | have a nice evening duflu | 10:10 |
| seb128 | thx for the ubiquity debugging | 10:10 |
| duflu | (I will test a proper low DPI machine before tomorrow too) | 10:12 |
| duflu | o/ | 10:12 |
| GunnarHj | seb128: Found this: | 10:20 |
| GunnarHj | https://unix.stackexchange.com/questions/260601/understanding-setting-up-different-input-methods | 10:20 |
| GunnarHj | "If GTK_IM_MODULE is not set, GTK selects a built-in IM on the basis of configurations in /etc/gtk-2.0/gtk.immodules. GTK 3.0 looks in /usr/lib/gtk-3.0/3.0.0/immodules.cache which is generated by gtk-query-immodules-3.0." | 10:20 |
| GunnarHj | (the correct file paths seem to be /usr/lib/x86_64-linux-gnu/gtk-2.0/2.10.0/immodules.cache and /usr/lib/x86_64-linux-gnu/gtk-3.0/3.0.0/immodules.cache) | 10:20 |
| seb128 | so ibus and autodetect should be equivalent? | 10:20 |
| GunnarHj | seb128: Probably only in GNOME by default. For instance, I think we should keep setting GTK_IM_MODULE for other frameworks, and for other flavors. | 10:23 |
| seb128 | GunnarHj, well, if GNOME ends up selecting ibus when that env is unset then I don't understand what is being fixed/acting differently and why | 10:23 |
| GunnarHj | seb128: I mentioned a few bugs where it has proved to make a difference. | 10:24 |
| seb128 | right, but I don't understand *why* | 10:25 |
| seb128 | if we say GNOME is supposed to pick one and the one is "ibus" | 10:25 |
| GunnarHj | seb128: Probably the one is not *always* ibus. As in bug #1761554 and bug #1760308. | 10:29 |
| ubot5 | bug 1761554 in gnome-shell (Ubuntu Bionic) "[bionic] Extended characters in GNOME screen keyboard don't get entered" [Medium,Triaged] https://launchpad.net/bugs/1761554 | 10:29 |
| ubot5 | bug 1760308 in ibus (Ubuntu Bionic) "Unicode input not working on Bionic with CTRL-SHIFT-U" [High,Fix released] https://launchpad.net/bugs/1760308 | 10:29 |
| seb128 | the upstream osk issue report that it should be "wayland" under wayland | 10:30 |
| GunnarHj | seb128: True. And still it's sufficient if it's unset. | 10:34 |
| seb128 | yeah | 10:34 |
| seb128 | anyway lunch time, need to drop from IRC | 10:35 |
| seb128 | but I had that to my backlog for later | 10:35 |
| seb128 | that comes after that ubiquity issue though | 10:35 |
| didrocks | enjoy your lunch seb128 | 10:35 |
| GunnarHj | Have a nice meal. Ok. | 10:35 |
| seb128 | didrocks, thx | 10:35 |
| Saviq | seb128: https://bugzilla.mozilla.org/show_bug.cgi?id=1390015 | 11:33 |
| ubot5-ng` | bugzilla.mozilla.org bug 1390015 in Folder and Message Lists "Large emojis displaying in message subject" [Normal, Resolved: Invalid] - Assigned to nobody | 11:33 |
| ubot5 | Mozilla bug 1390015 in Folder and Message Lists "Large emojis displaying in message subject" [Normal,Resolved: invalid] | 11:33 |
| ubot5 | bug 1390015 in OpenLP "Service Item Notes on Stage view" [Low,Fix released] https://launchpad.net/bugs/1390015 | 11:33 |
| seb128 | Saviq, thx | 11:33 |
| jbicha | Saviq: I think I heard the emoji issue is fixed in the latest Thunderbird Beta. I expect 18.04 to get Thunderbird 60 in several weeks | 11:51 |
| Saviq | heh, the bug still claims it's a font issue, but ok ;) | 11:52 |
| seb128 | Saviq, could you try from https://launchpad.net/~mozillateam/+archive/ubuntu/thunderbird-next ? | 11:54 |
| Saviq | will do | 11:54 |
| seb128 | Saviq, https://bugs.launchpad.net/ubuntu/+source/thunderbird/+bug/1759286 | 11:56 |
| ubot5 | Launchpad bug 1761844 in thunderbird (Ubuntu) "duplicate for #1759286 Thunderbird: emojis is displayed in the email list" [High,Confirmed] | 11:56 |
| seb128 | can use that if you need a launchpad bug to follow/comment on | 11:56 |
| Saviq | yeah emojis are fine there, calendar extension incompatible though | 12:03 |
| seb128 | good, at least it means they fixed that one | 12:05 |
| GunnarHj | jbicha: I thought we were talking about g-i-s. Is it a Ubiquity string? | 12:09 |
| jbicha | (for context, this is https://github.com/CanonicalLtd/desktop-design/issues/102 ) | 12:11 |
| ubot5-ng` | CanonicalLtd bug 102 in desktop-design "first-login: "location you choose" or "location you chose"" (comments: 6) [Closed] - Assigned to matthewpaulthomas (Matthew Thomas) | 12:11 |
| ubot5 | Error: Launchpad bug 102 could not be found | 12:11 |
| jbicha | we are talking about gnome-initial-setup | 12:11 |
| jbicha | for the common usecase of single user install done from a clean install, I am thinking there will be a mismatch between the language pack and gnome-initial-setup | 12:12 |
| jbicha | which language pack does ubiquity install if there is network access? the one from the ISO or the latest one available? | 12:13 |
| jbicha | btw, I think it would be nice if ubiquity would at least upgrade gnome-initial-setup at tne end of the install so that SRUs for gnome-initial-setup make more sense | 12:14 |
| seb128 | why does it matter? | 12:14 |
| jbicha | but we're stuck until at least 18.04.1 now | 12:15 |
| seb128 | it does it you click "install updates during installation" which is default | 12:15 |
| seb128 | also it covers people upgrading | 12:15 |
| jbicha | seb128: the ubiquity option is "download updates during installation" not "install" (!) | 12:15 |
| jbicha | I was confused by that for years and I think a lot of people are confused by that | 12:16 |
| seb128 | when are they installed? | 12:16 |
| jbicha | after the user reboots and logs in, update-manager pops up and asks the user to install updates | 12:16 |
| seb128 | k, well upgraders still get the SRU | 12:17 |
| ahayzen | Hi, does anyone know if there is a reason why autopkgtest's don't run for gnome-software, gnome-software-plugin-{flatpak,snap} ? http://autopkgtest.ubuntu.com/testlist#index-g And is my understanding right that a regression in snap/flatpak that causes a test failure in gnome-software-plugin-{flatpak,snap} would be picked up by autopkgtest and prevent the update from going into the archive? (if it was configured correctly) | 12:17 |
| GunnarHj | jbicha: I thought otherwise too, and get surprised each time. :( | 12:17 |
| seb128 | ahayzen, Laney might be able to help you with that | 12:17 |
| jbicha | I'd really like to either 1) have an additional checkbox to install updates too (instead of just downloading) or 2) change the existing option to install instead of download | 12:19 |
| ahayzen | seb128, ok :-) | 12:19 |
| jbicha | the downside is that it's my understanding that it would make the install take much longer once a release has been out for several months | 12:19 |
| seb128 | ahayzen, looks like gnome-software has no debian/tests / autopkgtest? | 12:19 |
| GunnarHj | jbicha: But in that case, isn't the conclusion that the string will be translated for first runners until the first langpack update has happened? | 12:20 |
| seb128 | jbicha, or change the wording to be more explicit about what the option does | 12:20 |
| seb128 | but that's sidetracking | 12:20 |
| seb128 | what was the real topic? | 12:20 |
| seb128 | what string? | 12:20 |
| jbicha | seb128: https://github.com/CanonicalLtd/desktop-design/issues/102 | 12:20 |
| ubot5-ng` | CanonicalLtd bug 102 in desktop-design "first-login: "location you choose" or "location you chose"" (comments: 6) [Closed] - Assigned to matthewpaulthomas (Matthew Thomas) | 12:20 |
| ubot5 | Error: Launchpad bug 102 could not be found | 12:20 |
| ahayzen | seb128, so it might just simply be missing the debian packaging to do so. I know they run tests in CI on gitlab, i was wondering how i could get autopkgtests running :-) | 12:21 |
| seb128 | jbicha, GunnarHj, no string change in a SRU unless really needed, seems that's one of those cases | 12:22 |
| jbicha | my suggestion is to keep the old string, add the new string and copy the translations for the old string to the new string so that translations should work now and with the next langpack update | 12:23 |
| seb128 | why do we need to fiddle with strings? | 12:24 |
| seb128 | it's not a critical thing to change wording | 12:25 |
| GunnarHj | seb128, jbicha: I'm assuming that we will do a full langpack update before 18.04.1, and install all the languages irrespective of whether they have been tested or not. That's what happened before 16.04.1, and according to Martin it was kind of practice. (With that said, I wouldn't encourage unnecessary string SRUs...) | 12:26 |
| jbicha | grumble grumble, I really tried getting this done before the release :( | 12:27 |
| seb128 | why does it matter? | 12:27 |
| seb128 | you still reply to that | 12:27 |
| seb128 | it's just wording of a random string | 12:27 |
| seb128 | +didn't | 12:27 |
| jbicha | it's poor grammar in English and makes Ubuntu look just a bit less professional | 12:27 |
| jbicha | I understand if we think the change is too risky | 12:28 |
| jbicha | I think it might be workable so that's why I made the suggestions | 12:28 |
| ogra_ | jbicha, just blame the french ;) | 12:30 |
| GunnarHj | seb128, jbicha: The langpack update before 18.04.1 will make a big difference for g-i-s (you know why), so in the light of that a string change may be a minor thing. It won't be really good until 18.04.1 anyway. | 12:33 |
| GunnarHj | For example, those intro pages are currently untranslated in Chinese and German (fixed in LP now, but no new langpacks). | 12:37 |
| seb128 | we could go back to not strip translations for it | 12:39 |
| seb128 | unless until .1 | 12:39 |
| seb128 | at least until* | 12:40 |
| GunnarHj | seb128: Yeah, that would both help some languages and make it possible to fix that string without a long interruption. | 12:42 |
| seb128 | there are other strings changes we need to do | 12:42 |
| GunnarHj | seb128: In g-i-s? Trying to recall who said "why do we need to fiddle with strings". :) | 12:45 |
| seb128 | :p | 12:45 |
| seb128 | well, we have strings that have wrong content, not only imperfect english | 12:46 |
| GunnarHj | seb128: Sounds like disabling the lang stripping before 18.04.1 is well motivated then. (+ communication on ubuntu-translators) | 12:47 |
| seb128 | right, I need to look at that | 12:48 |
| GunnarHj | seb128: Second thought... We are talking about almost only Ubuntu specific strings, right? Would those language exports be done as patches then? | 12:50 |
| seb128 | yes | 12:51 |
| GunnarHj | Is it worth it? Or better wait for translations to be ok in 18.04.1? | 12:52 |
| Laney | ahayzen: It doesn't define any tests. | 12:54 |
| ahayzen | Laney, ok, so i should look at contributing some, I'll have a read of the docs :-) | 12:55 |
| Laney | ahayzen: It has an upstream testsuite; might be good to look at running that if you can convince it to run against an installed gnome-software rather than the in-tree one. | 12:57 |
| Laney | Like we do for GLib & other things. | 12:57 |
| Laney | gnome's "installed-tests" stuff | 12:57 |
| ahayzen | Laney, ok thanks, i'll see if i can figure anything out :-) | 12:58 |
| juliank | I'd like to rename software-properties from "software & updates" to "software sources", matching the kde dekstop file. It's super confusing now that it's not in a settings menu anymore. | 12:58 |
| GunnarHj | juliank: I hope you are talking about some future relase and not 18.04. | 13:02 |
| juliank | 18.10 of course | 13:02 |
| GunnarHj | Puh. ;) | 13:02 |
| jbicha | juliank: check out https://wiki.ubuntu.com/SoftwareAndUpdatesSettings and talk to m_pt if you have suggested changes | 13:04 |
| jbicha | juliank: it looks like the newest version there is named "Legacy Software Settings" :| | 13:04 |
| jbicha | oh that new name is for a "Snap-based system" | 13:05 |
| seb128 | juliank, that's confusing, that Ui also does drivers and livepatch | 13:06 |
| jbicha | I look at the picture before I read the text :) | 13:06 |
| juliank | seb128: software & updates is not better, though | 13:06 |
| juliank | Software settings maybe | 13:07 |
| jbicha | with GNOME's ellipses in the Activities Overview, we've got Software Upd… and Software & Up… | 13:08 |
| juliank | yes | 13:08 |
| jbicha | 😢 | 13:08 |
| juliank | i just saw the complaint on reddit (https://www.reddit.com/r/Ubuntu/comments/8gbfjw/i_dont_always_use_the_gui_software_center_but/dyalakh/) and figured we could fix it | 13:09 |
| * juliank goes back to writing britney tests | 13:09 | |
| seb128 | jbicha, duflu/upstream have a fix for that | 13:14 |
| seb128 | juliank, that's worth considering | 13:14 |
| seb128 | juliank, we also want to try to move some of the things under settings, or it might make sense to split "Drivers", unsure why it has been added there | 13:15 |
| juliank | true | 13:16 |
| tseliot | seb128: hey, any chance this can be merged and uploaded? LP: #1753333 | 13:39 |
| ubot5 | Launchpad bug 1753333 in software-properties (Ubuntu) "Using "Software and Updates -> Additional Drivers" to Switch Fails" [High,In progress] https://launchpad.net/bugs/1753333 | 13:39 |
| seb128 | tseliot, you don't have upload rights? | 13:40 |
| seb128 | tseliot, I'm at a sprint/don't really have time for that today/this week, maybe cyphermox or didrocks can help you to get that uploaded? | 13:40 |
| tseliot | seb128: I do, but doesn't that require dealing with CI? | 13:41 |
| seb128 | what CI? | 13:41 |
| seb128 | just merge & dput imho | 13:42 |
| seb128 | there is no specific process/CI for software-properties | 13:42 |
| tseliot | seb128: ok, then I probably confused it with another package. I can do that myself, then | 13:44 |
| seb128 | right | 13:47 |
| seb128 | jbicha, k, you win, I'm going with some strings changes needed for g-i-s and tweak the template in advance. Can you open a launchpad bug if we don't have one yet for https://github.com/CanonicalLtd/desktop-design/issues/102 and add a code diff? | 13:53 |
| ubot5-ng` | CanonicalLtd bug 102 in desktop-design "first-login: "location you choose" or "location you chose"" (comments: 6) [Closed] - Assigned to matthewpaulthomas (Matthew Thomas) | 13:53 |
| ubot5 | Error: Launchpad bug 102 could not be found | 13:53 |
| tseliot | seb128: BTW, shall I upload to canimal first? | 14:05 |
| didrocks | tseliot: looking good to me, let me just do a quick testbuild before uploading | 14:06 |
| seb128 | tseliot, that's a question for the SRU team, historically they copied things over when it was that early | 14:06 |
| tseliot | didrocks: sure, thanks. BTW I merged the commit, and added a changelog entry. I haven't uploaded yet though | 14:06 |
| didrocks | for my current pending SRUs, I only uploaded to bionic | 14:06 |
| tseliot | seb128, didrocks: ok, good | 14:07 |
| didrocks | tseliot: yeah, saw that, the bug is fine, there is just a small nitpick in the syntax, changing this | 14:07 |
| didrocks | (in the generator, missing space before ]) | 14:07 |
| tseliot | didrocks: oh, I missed that one | 14:08 |
| didrocks | tseliot: always easier when someone else reread ;) | 14:08 |
| tseliot | true | 14:08 |
| tseliot | didrocks: are you going to push a commit for that, or shall I? | 14:09 |
| didrocks | tseliot: I'm pushing it as I do the release commit as well | 14:09 |
| tseliot | didrocks: great, thanks! | 14:10 |
| didrocks | no need to add extra step for this small fux | 14:10 |
| didrocks | fix* | 14:10 |
| didrocks | yw! | 14:10 |
| didrocks | ok, building fine, tests pass, sponsoring | 14:10 |
| didrocks | (done) | 14:10 |
| seb128 | thx didrocks | 14:10 |
| didrocks | nw seb128! | 14:11 |
| seb128 | jbicha, thanks | 14:20 |
| Oooohboy | hello all, I'm attempting to use systemd user service files to mount a drive on startup. My issue is that mount wants root permissions. Is there any way to elevate permissions for user services? | 17:59 |
| robert_ancell | kenvandine, how do you subscribe to MPs on GNOME gitlab? | 20:31 |
| kenvandine | robert_ancell, not sure actually | 20:31 |
| robert_ancell | I wonder if it uses the DOAP file to pull out the maintainers? | 20:32 |
| robert_ancell | But that can't work for private stuff, surely | 20:32 |
| robert_ancell | I'm trying to work out why I'm not notified by https://gitlab.gnome.org/Community/Ubuntu/gnome-initial-setup/merge_requests/1 | 20:33 |
| ubot5-ng` | Ubuntu bug (Merge request) 1 in gnome-initial-setup "Ensure stamp file if the user quit the wizard" (comments: 0) [Opened] | 20:33 |
| kenvandine | ah | 20:33 |
| kenvandine | there's a notifications setting on the main project page | 20:33 |
| kenvandine | defaults to "Global" | 20:33 |
| robert_ancell | oh, now I see it | 20:33 |
| kenvandine | robert_ancell is there another branch that has individual commits? | 20:34 |
| robert_ancell | kenvandine, no | 20:34 |
| jbicha | robert_ancell: good morning, I uploaded a new gnome-initial-setup to bionic/cosmic UNAPPROVED earlier today (just updates the what's new graphic) | 20:38 |
| robert_ancell | jbicha, cool. Can you do a MP to https://gitlab.gnome.org/Community/Ubuntu/gnome-initial-setup with the change? | 20:39 |
| jbicha | robert_ancell: um, I only changed debian/ and there isn't a debian/ there | 20:40 |
| robert_ancell | jbicha, that's only in there because debian doesn't support binary patches >:( | 20:40 |
| robert_ancell | it's in the source as gnome-initial-setup/pages/ubuntu-changes/ubuntu-changes.png | 20:41 |
| jbicha | it can do it if you list the patch file name in debian/source/include-binaries | 20:41 |
| kenvandine | robert_ancell, did you come up with a way to handle translations of the page? | 20:42 |
| robert_ancell | kenvandine, not yet | 20:43 |
| jbicha | robert_ancell: maybe I should just push directly if you want me to rebase your single "Ubuntu mode" commit? | 20:44 |
| robert_ancell | jbicha, I'm fine with you pushing directly | 20:45 |
| jbicha | force-pushed but now Andrea's merge request will need to be rebased | 20:48 |
| seb128 | jbicha, robert_ancell, that's not right, andyrock's change have been submitted a while ago and I made him rebase today | 22:05 |
| seb128 | you can't just come and do our stuff and ask people to keep rebasing their changes while they don't get reviewed | 22:05 |
| seb128 | should be somewhat fifo | 22:06 |
| seb128 | also now is not time to do new release changes, SRU coming first | 22:06 |
| robert_ancell | seb128, I rebased andyrocks changes - it's fine | 22:10 |
| seb128 | robert_ancell, hey, k, I just read your comment on the mp, what it solves is that using the shell panel-menu to exit should not lead to the wizard to come again on next login | 22:11 |
| seb128 | users who do that just want to not be bothered | 22:11 |
| robert_ancell | seb128, the AppMenu? | 22:11 |
| seb128 | yes | 22:11 |
| robert_ancell | ah, bug 1766685 | 22:13 |
| ubot5 | bug 1766685 in gnome-initial-setup (Ubuntu) "Gnome-initial-setup is shown on next login if appmenu is used to quit it." [Medium,In progress] https://launchpad.net/bugs/1766685 | 22:13 |
| seb128 | yes | 22:13 |
| robert_ancell | andyrock, are you online? | 22:16 |
| === BigW is now known as BigWhale | ||
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!