[04:47] anyone facing slack & skype crashed on cosmic ? [04:50] https://bugs.launchpad.net/ubuntu/+source/glibc/+bug/1790966 [04:50] Ubuntu bug 1790966 in glibc (Ubuntu) "Electron apps segfault on glibc 2.28 (cosmic)" [Undecided,Incomplete] [06:50] good morning [06:51] Salut didrocks, le week end était bon? [07:05] Morning didrocks, jibel [07:05] jibel, do you have a secret stash of older cosmic images? I need something older than 5 Sept === pstolowski|eow is now known as pstolowski [07:05] Well, I have some myself but need other people to be able to download an older image for testing [07:10] ^ for mvo [07:10] duflu, I don't. sorry [07:10] Hmm. Wonder if I can upload somewhere [07:12] duflu: you could probably just use people.c.c to make the specific one available? and unfortunately I also don't have older comisc images available [07:13] I keep a collection for this reason, but haven't ever had to share one before [07:14] mvo, can I scp to that? or some other protocol? [07:17] duflu: yeah, scp should work [07:17] duflu: only from the vpn though [07:19] salut jibel, oui, le week-end allait, et toi ? [07:19] hey duflu [07:20] Ah! mvo try this (I know 20180822 was a good one): http://mirrors.piconets.webwerks.in/ubuntu-mirror/ubuntu-cdimage/daily-live/20180822/ [07:22] Their published SHA1SUM matches my original authentic one [07:24] ... which is 7dcc2e28de480bb14a958de4132d3b5a323ef3e7 [07:25] duflu: nice [07:26] duflu: this is one that will expose the upgrade hang? [07:26] * mvo downloads [07:31] good morning! [07:43] Morning andyrock [07:44] hey andyrock [07:45] mvo, I assume so. I think I have installed a couple of machines from 20180822 and then hit the bug [07:45] Until today that was the newest "current" image I knew of [07:45] Now it's old :) [07:46] duflu: cool [07:46] duflu: I mean, cool that I'm (hopefully) now able to reproduce it [07:46] As cool as we can hope [07:46] duflu: people got quite personal about this bug [07:46] Did they? [07:46] duflu: I got mails like "can't you test your upgrades" [07:47] duflu: or "why don't you break snapd" [07:47] Ignore them I guess. We need as many testers as possible, even if sometimes they get angry [07:48] duflu: yeah, I'm not fuzzed, its just unusual, hasn't happend in a long time [07:48] duflu: anyway, now that I have a test image I will try to get to the bottom of this. thanks again for your help! [07:49] On that note, my own experience with angry users in bugs is somewhere around 1 in 400. And someone else recently told me they see about the same [07:54] morning [07:56] hey willcooke [08:00] Hi willcooke [08:00] yo [08:01] morning didrocks Laney. Afternoon duflu [08:01] Hi Laney [08:03] what ho [08:05] hey Laney [08:16] good morning desktopers [08:16] hey seb128 didrocks [08:16] what is going on [08:16] hey seb128 [08:17] hey Laney, willcooke [08:17] did you have a good w.e? [08:17] Hi seb128 [08:18] hey duflu [08:31] duflu: yay, I can reproduce [08:32] Good-ish [08:33] and the winner: seeding :/ [08:34] so it looks like the seeding is not complete on this image, I wonder if there is something wrong with the order, looking now. anyway, thanks duflu that was the crucial missing piece of information [08:35] mvo, you're welcome but I don't understand which type of seeding you're referring to. Also don't need to [08:35] Deb seeding maybe [08:36] duflu: snap seeding is broken on the cd [08:36] duflu: well, "broken" [08:36] duflu: but yeah, no worries unless you want to I will not bother you with the details :) [08:37] mvo: you've seen bug 1785388 right? [08:37] bug 1785388 in snapd (Ubuntu) "seeds.yaml needs updating in cosmic in order that all snaps are initialised" [Undecided,Confirmed] https://launchpad.net/bugs/1785388 [08:38] seb128: it looks like you tagged but didn't upload totem-pl-parser in Debian [08:40] did I? [08:40] jibel: now I did, I did not connect the dots earlier [08:41] jbicha: -^ [08:41] jibel: sorry [08:41] jibel: wrong nick completion but still "good morning" :) [08:41] maybe we'll wait for Ken, but any idea why the snaps don't use the system theme icons? [08:42] jbicha: we work on this on our side now but note that I don't agree with the rational to leave this open in comisc and let our users run into a hard to recover failure scenario when there is a trivial workaround availalbe (but directed at you juts a general comment) [08:42] didrocks, they .desktop has an absolute path for Icon= no? [08:43] mvo: hey, maybe this is in your realm? As we define an icon via the generated .desktop file in snapd, it doesn't match the theme (and people open bug reports about it) [08:43] Icon=/snap/gnome-calculator/222/meta/gui/gnome-calculator.png [08:43] that wasn't the case before, correct? I remember to have seen our old theme icon before [08:43] or did just Ken used the Humanity one in the past and changed? [08:43] (in the snap) [08:43] didrocks: I'm not aware of this change but let me look at the code [08:43] didrocks, back to GNOME idea that theming needs to be killed for icons :p [08:44] yeah :p [08:44] didrocks: what is the expected behavior? Icon=gnome-calculator without a full path? [08:44] maybe it's just Ken changing the icon within the snap [08:44] mvo: I *think* it should be the expect behavior if we don't provide an icon in the snap [08:44] expected* [08:45] ah, maybe, that was what Ken was doing, and he got bug reports on other distros? unsure [08:45] that, or if nothing changed on snapd side, he may just have used the Humanity icon and changed it [08:45] does snapd let you "not provide an icon"? [08:45] unsure [08:47] there was a technical reason they used the full path but I don't remember which one [08:47] well, I can that kind of makes sense if you want to enforce your icon as part of your snap [08:47] think* [08:48] but maybe we don't want for our default snaps which can be rebranded? unsure [08:48] the issue is that if we replace the icon with the Suru/Yaru one, it will be the default on all distros and env [08:49] ah, I remember the issue [08:49] snaps can't install an icon in /usr/share/icons [08:49] so if they just specifiy a name and not a path, the icon they provide can't be found [08:50] we would need some hook that installs icons in a xdg_path like for the .desktop [08:50] (https://irclogs.ubuntu.com/2018/02/22/%23ubuntu-desktop.html#t09:03 was a conversation about that) [08:50] how did it work in the paste? We only shipped the Humanity icon? [08:50] (because the icon changed) [08:50] past* [08:50] grrrr, can't type [08:51] we are doing the full path thing for ever afaik [08:51] so, I guess Ken just changed the icon, let's confirm once he's back [08:51] do you have an issue about a specific snap? [08:52] no, they only mention: https://community.ubuntu.com/t/call-for-participation-an-ubuntu-default-theme-lead-by-the-community/1545/1708 and https://forum.snapcraft.io/t/gnome-calculator-theming-issues/7290 [08:52] I don't know where to direct them to open an issue on our snaps, do we have now a "distro" in launchpad for this? [08:52] gnome-calculator then [08:53] yep [08:54] well https://gitlab.gnome.org/GNOME/gnome-calculator/commit/1afbe22e [08:54] [08:54] Icon=gnome-calculator [08:54] [08:54] Icon=org.gnome.Calculator [08:54] upstream [08:54] sorry, gitlab copy/paste of diff fail [08:54] :p [08:54] but they changed the icon name upstream [08:54] ah, that explains then! [08:54] funny that in the "about" dialog, they are using the old icon name [08:54] (and you have the Yaru one ;)) [08:55] :) [08:55] great findings! unsure how we are going to fix this, but nice [08:55] so, I'm puzzle how with the correct name, it's picking the right one [08:56] as it's hardcoded [08:58] we are sure it was picking the themed icon for the snap? [08:58] that sounds weird to me [08:58] yeah, it definitively was. I remember to have seen twice the same icon (deb and snap) [08:58] and people just recently (on those forum threads) are mentioning the same, they would have mentionned it earlier [08:59] so why did it work? puzzling :/ [08:59] (that was why I was pinging m_vo about potential snapd changes) [09:02] didrocks, the previous snap had "Icon=gnome-calculator" but the snapcraft.yaml doesn't have any icon mention [09:03] I bet snapd/snacraft have some magic when icon_name == snap_name [09:03] which stopped being true [09:03] jibel, I think the black screen when remote controlling might have had something to do with not having a screen plugged in on the server. Or a reboot fixed it. On a related note, the reason my HDD failed was because, I think, the case fans were jammed up with dirt and dust and the old HDD over heated and died. [09:03] mvo, hey, I did a [09:03] $ snap refresh --beta gnome-calculator [09:03] Automatically connect eligible plugs and slots of snap "gnome-calculator" | [09:04] but it's spinning endless for some minutes now on that line [09:04] seems stuck [09:04] should I ctrl-C it? [09:04] (I tried beta because that has 3.28.1) [09:04] ah [09:04] it finished [09:04] wonder why it took minutes to do that [09:04] mvo, unping [09:05] seb128: ahah, ok, so something really changed, let's see then ;) [09:05] didrocks, well, "something" is upstream that renamed their icon yes :) [09:05] thanks for looking! [09:05] np [09:05] yeah [09:05] let's wait for kenvandine once he's around (Ken, please backlog ^) [09:06] I like the fact that they renamed it but didn't change the about dialo [09:06] dialog* [09:06] something to check for the one doing the update, if there are broken icons due to that [09:07] it has been poorly coordinated imho [09:07] didrocks, https://gitlab.gnome.org/GNOME/gnome-calculator/commit/05fa1ab3 [09:08] urgh, indeed [09:09] hum, where my n-m-applet has gone? [09:09] no email, uploaded an hour ago though [09:10] didrocks, someone else said the same thing last week [09:10] hum, network-manager-applet_1.8.16-1ubuntu1_source.ubuntu.upload [09:10] let me reupload, just in case [09:11] ah, got the reject email this time :) [09:12] and the second one at the same time, weird [09:12] * didrocks reuploads with -sa [09:20] and n-m-applet which is FTBFS now. Was passing on my install on Friday which was mostly up to date… [09:20] error: macro "G_UNAVAILABLE" passed 3 arguments, but takes just 2 [09:21] who is changing macro # of args? grrr [09:22] I guess cherry-picking 1.12.2-2 for n-m will be needed [09:24] ah, n-m 1.12.2-0ubuntu3 was published on 2018-09-05 (I guess due to glibc) [09:24] so, I wasn't up to date when building n-m-applet. Anyway, there is a patch in debian, will first confirm the FTBFS locally, backport the patch and see [09:26] I see that https://launchpad.net/ubuntu/+source/gnome-screenshot/3.26.0-0ubuntu1 is stuck due to FTBFS but there is 3.30 out in debian [09:26] didrocks, n-m was blocked for a while because of netplan.io autopkgtest regression, which foundations investigated/decided to skip (it was just due a format change in the nm command output) [09:26] can we try force-syncing that one? or does it need a proper merge? [09:26] xnox, no, it needs proper merge [09:27] awww shame [09:27] ok [09:27] seb128: ok, but anyway, at least, I have my explanation why it was working here [09:27] sorry about that [09:27] and sounds like it will be easy to fix, so let's try [09:27] well, bad timing, not your fault :p [09:28] if you didn't already, I'll backport as well 1.12.2-3, sounds worthy [09:28] ( * Apply patch from upstream to fix FTBFS with GLib 2.57 (Closes: #906965)) [09:29] didrocks, I didn't, thx! [09:33] xnox, do you need the new gnome-screenshot for some urgent reason? [09:34] not really no. [09:34] xnox, there are plenty of foundations packages that need to be updated if you are borred btw, no need to look at our desktop ones :) [09:34] the one in release pocket is fine [09:34] i just looked into it, as it was part of the reverse stack. but it doesn't appear to be "broken because omg ubuntu is broken" just by itself borked [09:35] right, andyrock is looking at this upate for us, it's going to be resolved probably this week [09:36] seb128: I've already a branch ready [09:36] andyrock, ah ok, it's not linked/mentioned on the launchpad bug? [09:36] not a MP yet, I'll do asap [09:36] https://bugs.launchpad.net/ubuntu/+source/gnome-screenshot/+bug/1760192 [09:36] Ubuntu bug 1760192 in gnome-screenshot (Ubuntu) "Update gnome-screenshot to version 3.30" [Low,In progress] [09:36] k [09:36] great [09:36] and hey :) [09:36] thx andyrock, happy monday btw :) [09:37] seb128: btw I dropped the patches to remove headerbar in unity [09:38] seb128: I guess it's not worth [09:38] other apps are broken anyway [09:38] seb128: and I added this: https://git.launchpad.net/~azzar1/ubuntu/+source/gnome-screenshot/commit/?id=25553d04c67f0169b7a36edf230e8c817d4d8ca4 [09:38] andyrock, seems fine [09:39] what is that? save dialog when? [09:39] in unity if you click e.g. PrintScreen you get a save dialog [09:39] ah, printscreen in GNOME just does nothing visible right? [09:40] in gnome-shell it just saves the screenshot under "Pictures/" [09:40] right [09:40] does unity session include "ubuntu"? [09:40] or should we check for either unity || ubuntu? [09:40] ubuntu includes both [09:41] let me check again [09:41] thx [09:41] (yep, ubuntu includes both) [09:41] it imapcts MATE though as well [09:41] impacts* [09:42] if they don't want to be impacted by ubuntu choices they shouldn't claim to be an ubuntu session I guess? [09:42] thx didrocks [09:42] or said differently, I think it's fine [09:42] agreed, or they can add their own override [09:42] I agree, just wanted to mention it :) [09:43] seb128: I also prepared a MP to upload gnome-screenshot 3.30 in debian [09:43] not sure what's going on but our gnome-software gbp'config is broken [09:44] the pristine-tar and upstream/latest branches are not in sync [09:44] andyrock, hey! Thats great, thanks for working on that. [09:44] I was preparing the branch to support XCancelUbuntu in gnome-softwares shell-search-provider and I noticed that [09:45] *gnome-software 3.30 in debian [09:45] willcooke: hey hey [09:45] np! [09:56] * didrocks did 2 dputs for network-manager as well, and no upload proceeded, no email, let's do that after lunch [09:56] Hey, who do i need to ask to review this change for ubuntu's gnome-software, is it Robert or is anyone on the desktop team able to do it? https://gitlab.gnome.org/Community/Ubuntu/gnome-software/merge_requests/5 I've reported an ubuntu bug here as well https://bugs.launchpad.net/ubuntu/+source/gnome-software/+bug/1791553 [09:56] (at least, that fixes n-m-applet FTBFS locally) [09:56] Ubuntu bug 1791553 in gnome-software (Ubuntu) "Update preferences is not shown when flatpak plugin is installed" [Undecided,New] [09:56] Ubuntu issue (Merge request) 5 in gnome-software "gs-shell.c: Show update preferences when flatpak plugin is enabled" [Opened] [09:56] ahayzen: I would say Robert [09:57] ok thanks :-) [10:36] andyrock, sorry got sidetracked and forgot to reply here, gnome-screenshot got updated in Debian meanwhile by someone who didn't check pending MRs :/ [10:38] ahayzen, Robert is best but others can upload/review if needed [10:39] seb128, ok thanks for the info :-) [10:54] seb128: have I asked you to upload the new, and buildable, colord yet? [10:54] If not... 😉 [10:55] RAOF, no you didn't, also did you decide on the version for cosmic? [10:55] 1.4.3 should be good, yeah. [10:56] I guess I'll stop hitting C++ dynamic_cast and file an ffe tomorrow. [10:57] Although, really, the colord we already have is mostly fine... [12:21] Hi. The PulseAudio packages creates the user *pulse*. I thought PulseAudio does not need that anymore, bcause it is a user(?) daemon now. [12:24] /usr/sbin/useradd -d /var/run/pulse -g pulse -s /usr/sbin/nologin -u 78 pulse [12:52] seb128: yeah saw that, but I was talking about gnome-software [12:52] is Robert a DD? [12:53] didrocks, Laney: maybe we could make this default in our pkgs too? https://salsa.debian.org/gnome-team/vte2.91/commit/e9de01fa94c0b81cc86160148c5afd55c1da0edc [12:53] sounds ok [12:54] we will get it through merging for most things no? [12:56] yeah, lgtm [12:56] https://wiki.debian.org/Gnome/Git#gbp.conf is probably a more future proof link [12:57] andyrock, no [12:58] andyrock, unsure why the pristine-tar and upstream/latest are diverging, maybe Robert doesn't pull from Debian, unsure how familiar he is with the gbp workflow and sharing with Debian [12:59] seb128: https://salsa.debian.org/gnome-team/gnome-software/merge_requests these MPs will take care of salsa [13:00] andyrock, that doesn't explain why we see divergences though? === sergiusens_ is now known as sergiusens [13:00] seb128: I guess robert forgot to push the 'pristine-tar' and 'upstream/latest' branches [13:00] https://code.launchpad.net/~ubuntu-desktop/ubuntu/+source/gnome-software/+git/gnome-software [13:00] looks to me like he forgot to put [13:00] push [13:01] not really divergence per se [13:01] seb128: the best approach is to upload them in salsa and then do a "git merge debian/3.30..." [13:01] yeah divergence was not the correct word [13:01] I though those should be handled through "git pull" and not "git merge"? [13:01] also unsure if Robert has commit access to salsa [13:02] well "git pull" will behave as a merge [13:02] maybe he didn't forget but just can't push/isn't interested by doing the extra work there [13:02] seb128: still the ubuntu branches are broken [13:03] how so? [13:03] you can push to launchpad and ask someone to push them on to salsa [13:03] Laney, you know robert? ;) [13:03] * Laney runs [13:03] Laney: I can't push to launchpad [13:03] not you [13:03] the updater [13:04] anyway nothing that urgent [13:06] andyrock: you can probably find the upstream/xxx commit he used by looking in the ubuntu/master branch for the merge commit [13:06] I mean assuming gbp import-orig was used ... [13:09] looks like not, the upstream source is out of date there [13:10] andyrock, Laney, I let you sort that one out, seems another case where I'm getting lost in git workflow details and I don't want to spend an hour today fighting with tools [13:10] seb128: kk [13:10] ok [13:10] the explanation is that he just committed the debian directory update [13:10] not so many details there [13:11] I suggest that we merge the MPs in debian before [13:11] I mean salsa [13:11] I'm unsure if Robert is familiar with the workflow to use [13:12] andyrock, "Changes 118+" [13:12] that's just too much to me to review properly today [13:12] seb128: most of the are from upstream [13:12] also I don't know enough about what Debian does/wants to do with the plugins (especially the snap one) [13:12] right [13:12] but salsa doesn't make it easy to split out the downstream changes to review [13:13] gitg does [13:13] btw it can wait [13:13] right, as said I don't want to get too much sidetracked [13:13] I can have a look this week [13:13] pushed the upstream and pristine-tar bits [13:13] but it's not trivial enoguh that I want to deal with it now [13:13] Laney, thx [13:13] didn't review the pkging yet [13:13] that can wait as said [13:14] I fixed the launchpad branch too, what that required was 'git merge upstream/3.30.0' (andyrock's tag that he pushed to salsa) [13:14] someone should improve the way launchpad shows diff [13:14] starting from merge commits [13:14] merge commits are kind of broken in gitlab too [13:15] andyrock: how broken in gitlab? the diff looks very similar in general to what github is showing up [13:18] ok I just rm -rf a git checkout :-O [13:18] git remote add lp --fetch [13:18] didrocks: I would prefer if they showed the additional diff in the merge commit [13:18] that gives you a bad time [13:19] andyrock: yeah, GH doesn't do as well, but I agree with you [13:19] when you merge and apply additional changes, I would prefer if gitlab highlights the additional changes (like git show does...) [14:02] Laney: do you mind pushing in gnome-software a gbp.conf file too [14:03] or maybe I should just propose the change in https://gitlab.gnome.org/Community/Ubuntu/gnome-software [14:03] and robert will take care of it [14:20] andyrock, so we have a vcs in gitlab for ubuntu and one in launchpad now? should we deprecate one of those? [14:23] seb128: let's wait for robert [14:23] robert seems to prefer to handle things without gbp [14:23] didrocks, unlucking, there is another n-m-applet version out since friday :) [14:23] andyrock, right, let sync up with it (or keep that for discussion next week) [14:24] andyrock, unless you have changes you need to get uploaded on gnome-software this week? [14:24] Laney: in terminal do we still need debian/patches/0001-Add-the-style-context-provider-with-FALLBACK-priorit.patch ? [14:24] Laney: there's no bug reference on that! [14:24] seb128: yeah I've the XUbuntuCancel one [14:24] ah, right [14:25] seb128: I'm testing it right now [14:25] it's a small one [14:25] ah debian's 787104 ? [14:26] seb128: do we really want it? :p [14:26] didrocks, it fixes regressions in g-c-c apparently... ;) [14:27] humf… [14:27] :/ [14:27] https://gitlab.gnome.org/GNOME/network-manager-applet/issues/20 [14:27] GNOME issue 20 in network-manager-applet "Cannot import VPN configuration from standard ovpn files" [Closed] [14:27] that will be for tomorrow [14:28] yeah, no hurry, it's bugfixing and minor [14:29] thanks for the head's up! [14:30] it's not "vcs in gitlab for ubuntu and one in launchpad *now*", there was always two [14:30] np, sorry I'm just a less-red-on-version's maniac atm :p (though going to shift away from that now than we are post ff) [14:30] andyrock: feel free to include your gbp.conf in your MP(S) imho [14:30] Trevinho: I pushed a commit adding that (vte not terminal, that confused me for a minute :P) [14:30] Laney, I though the fullsource in git in launchpad would deprecate the gitlab ubuntu variant, probably wrong assumption from my part [14:30] the answer is yes, probably [14:31] Laney, I guess gnome-software is good to have a discussion over with Robert next week [14:32] sure [14:32] it had a fork of upstream because there are/were so many patches that it's easier to maintain them like a fork [14:33] it used to be a branch in the actual upstream repository but that changed, not sure why [14:33] maybe to let other people commit there who don't have gnome rights [14:33] yeah, I don't know [14:34] I know that for gnome-initial-setup Robert is maintaining the diff as a big patch that he keeps rebasing since he said that from experience maintaining a fork as a stack of git isn't working well [14:34] that sounds a bit like our gbp branches [14:34] unsure if gnome-software is in the same bucket though [14:35] anyway, it's going to be easier to discuss that together next week, especially with Robert since he's not online at the same hours as we do otherwise === sergiusens_ is now known as sergiusens === Class7_ is now known as Class7 [15:40] seb128: in terminal we had "Install compatibility version of gnome-terminal.desktop", I guess I can drop that now, right? [15:42] yes those can be dropped === ecloud_away is now known as ecloud === pstolowski is now known as pstolowski|afk [16:46] Trevinho, can be removed [17:06] Could someone please steal my gtk+2.0 merge? I don't want to deal with it anymore ;) [17:06] Laney, jbicha: ^ [17:39] Let do another try to build firefox with wayland :) [18:19] 0:34.62 /build/firefox-62.0+build2/obj-x86_64-linux-gnu/dist/include/nsISupportsImpl.h:17:10: fatal error: prthread.h: No such file or directory [18:19] bleh :( [21:14] dupondje, so much optimism!