[01:17] <callmepk> good morning
[01:18] <duflu> Hi callmepk 
[01:18] <callmepk> hi duflu 
[06:08] <oSoMoN> good morning desktoppers, happy Friday!
[06:09] <jibel> Good morning 
[06:17] <oSoMoN> salut jibel 
[06:21] <jibel> salut oSoMoN 
[06:37] <didrocks> good morning
[06:43] <oSoMoN> salut didrocks 
[06:47] <didrocks> salut oSoMoN 
[06:51] <jibel> salut didrocks 
[06:53] <didrocks> salut jibel 
[08:04] <laney> \o
[08:05] <duflu> Salut oSoMoN, jibel, didrocks, laney 
[08:50] <jibel> hi duflu 
[10:00] <seb128> hey IRC, how is everyone doing today?
[10:14] <oSoMoN> hey seb128, I'm doing good and you?
[10:31] <seb128> oSoMoN, I'm alright thanks!
[10:38] <laney> moin seb128 
[10:39] <seb128> laney, hoi, wie gets?
[10:42] <laney> it's ok, looking forward to the weekend!
[10:42]  * juliank updates to proposed gnome-shell 40 - living dangerously
[10:43] <juliank> Also pipewire still only works with one of my headphones, sigh, gotta update the upstream issue for the WF1000XM3
[10:43] <juliank> * BT headphones
[10:44] <laney> do it
[10:44] <juliank> So awkward to have desktops move side by side
[10:45] <juliank> but looking good
[10:46] <juliank> Sometimes app titles overflow in the dock thingy at the bottom in vanilla gnome session
[10:47] <juliank> but apart from that, it works https://imgur.com/a/q1Bs1VA
[10:47] <juliank> :D
[10:48] <juliank> Wonder if Chrome has Wayland support, I should try Wayland again
[10:50] <juliank> Hmm seems gdm is dead now
[10:51] <juliank> ok / was filled up
[10:57] <juliank> One day I shall figure out why my Eclipse app starts with a bright menu bar when autostarted and a black one if I restart it. I guess autostart happens before gnome-settings-daemon has properly propagated the theme choice
[11:10] <KGB-1> gnome-control-center signed tags 4f3edf8 Sebastien Bacher ubuntu/1%40.0-1ubuntu2 * gnome-control-center Debian release 1:40.0-1ubuntu2 * https://deb.li/3WkQ2
[11:10] <KGB-1> gnome-control-center ubuntu/master 8a6b820 Sebastien Bacher debian/ changelog gnome-control-center-data.install patches/debian/info-overview-Use-the-Vendor-logo-prominently.patch ubuntu-logo-icon.png * use a vendor logo icon which is available in Ubuntu (lp: #1934564) * https://deb.li/YNui
[11:10] <KGB-1> gnome-control-center ubuntu/master b51991b Sebastien Bacher debian/changelog * upload to ubuntu * https://deb.li/3AKSh
[11:45] <didrocks999> laney: do you know why fonts-cantarell is back in universe? You got a MIR in vidid to promote it: https://bugs.launchpad.net/ubuntu/+source/fonts-cantarell/+bug/1387599 and I don’t see it anywhere in main on our supported LTSes
[11:45] <didrocks999> context: gtk4 is recommending it
[11:52] <didrocks999> Trevinho: libgtk-4-common deps on libgtk-3-common, any reason?
[11:52] <didrocks999> same libgtk-4-bin deps on libgtk-3-bin
[11:52] <didrocks999> that will make the removal (one day!) of gtk3 more complex
[11:54] <didrocks999> (on fonts-cantarell, I think we can just not ship gtk-4-examples)
[12:01] <seb128> we can probably make the ubuntu font an alternative recommends there or lower to a suggest
[12:02] <didrocks999> well, -examples isn’t mandatory, I would like though the exaplanation on dep on gtk3 for common and bin and how a future transition would be handled
[12:03] <seb128> git blame ends up with that explanation
[12:03] <seb128> +  [ Jeremy Bicha ]
[12:03] <seb128> +  * Have libgtk-4-bin and libgtk-4-common and gtk-4-examples depend on
[12:03] <seb128> +    their gtk3 counterparts until we figure out how to handle the file
[12:03] <seb128> +    conflicts there. The conflicted files have not changed in gtk4.
[12:03] <seb128> +    See GNOME bug 774912.
[12:03] <seb128> http://bugzilla.gnome.org/774912
[12:03] <seb128> https://bugzilla.gnome.org/show_bug.cgi?id=774912
[12:04] <seb128> which was migrated to https://gitlab.gnome.org/GNOME/gtk/-/issues/707 and marked as resolved
[12:05] <didrocks999> interesting, so soname-versioned and common files…
[12:06] <seb128> from a current build log
[12:06] <seb128> Installing /<<PKGBUILDDIR>>/gtk/tools/gtk4builder.its to /<<PKGBUILDDIR>>/debian/install/deb/usr/share/gettext/its
[12:06] <seb128> so sounds like we should be able to just remove those depends
[12:06] <seb128> Trevinho, do you want to handle that or should I?
[12:06] <didrocks999> yeah, that would be good and not something to do in a second pass when we really want to transition
[12:07] <didrocks999> that was my only things I spotted, otherwise, good for promotion and we shouldn’t promote -examples IMHO
[12:07] <seb128> wfm
[12:07] <seb128> thanks for the quick review!
[12:10] <didrocks999> yw :) np!
[12:11] <didrocks999> take that as a +1, so if you happen to be around for promoting before myself, feel free
[12:12] <seb128> k, will do, thanks again!
[12:57] <Trevinho> Trevinho: oh, let me check but I feel it's fixed by new revision no?
[12:57] <Trevinho> seb128: sorry self pinged :D
[12:57] <Trevinho> that was for you
[12:57] <Trevinho> I've 4.3.1 ready, but i've not the powers for gtk4 it seems, not under gnome team
[13:01] <Trevinho> btw s390x is ALWAYS correct in failing... love it! https://gitlab.gnome.org/GNOME/gtk/-/merge_requests/3740/commits
[13:03] <seb128> Trevinho, hey! how are you? 
[13:03] <seb128> Trevinho, what do you mean by 'not under gnome team'? it's in https://salsa.debian.org/gnome-team/gtk4
[13:24] <Trevinho> Trevinho: it is in the repo, but not in the debian metadata, so my upload last night was rejected...
[13:24] <Trevinho> I've asked laney to include that package :)
[13:24] <Trevinho> seb128: btw all good, you?
[13:28] <laney> didrocks999: looks like it was a build depends and we stopped requiring those to be in main so it would have been demoted then
[13:28] <laney> I would guess examples can stay in universe
[13:50] <didrocks999> Trevinho: so, we need your expertise on gdm-session-worker.c & PAM… :)
[13:51] <didrocks999> Trevinho: if we have a session PAM module which is required/requisite and return an error…
[13:51] <didrocks999> on focal & impish, GDM stops, while starting to open a session
[13:51] <didrocks999> detect that pam_sm_open_session failed
[13:52] <didrocks999> try to switch back to VT1 (but nothing has to be done as we never moved from VT1)
[13:52] <didrocks999> and we end up with a black screen
[13:52] <didrocks999> as no X is restarted
[13:52] <didrocks999> then, the whale is failing, blablabla
[13:56] <didrocks999> some pastebin: https://paste.ubuntu.com/p/5TjJBmCstQ/
[13:57] <didrocks999> I think the issue is due to the VT switch not switching back to VT1
[13:57] <didrocks999> juil. 09 15:30:54 client01 gdm-password][3765]: GdmSessionWorker: jumping to VT 1
[13:57] <didrocks999> juil. 09 15:30:54 client01 gdm-password][3765]: GdmSessionWorker: first setting graphics mode to prevent flicker
[13:57] <didrocks999> juil. 09 15:30:54 client01 gdm-password][3765]: GdmSessionWorker: VT mode did not need to be fixed
[13:57] <didrocks999> (as it never left it at first)
[13:57] <didrocks999> or maybe for gdm to wait for the session to really succeed in PAM before starting it
[14:01] <jibel> Trevinho, also with login or ssh for example, if PAM open session is not successful, then session initialization is aborted correctly
[14:05] <Trevinho> didrocks999: oh, sorry missed it... let me read the backlog...
[14:11] <jibel> Trevinho, a simple way to reproduce the issue is to add "session requisite pam_deny.so" early in /etc/pam.d/gdm-password
[14:11] <Trevinho> didrocks999: mhmh, log isn't fully clear to me, I guess I'll have to get dirty in my hacked gdm version
[14:13] <didrocks999> Trevinho: a simple test case is to set "session requisite pam_deny.so" in /etc/pam.d/gdm-password like after "@include common-account"
[14:24] <Trevinho> didrocks999, jibel... ok got it... so the whole session gets stopped and never restarted for gdm itself
[14:25] <didrocks999> Trevinho: exactly
[14:25] <Trevinho> back to vt2/vt1 sets it back... mhmh
[14:25] <didrocks999> yeah, you never quitted vt1 from what I see
[14:25] <Trevinho> might look something related to systemd too 
[14:25] <didrocks999> then, the fallback try to reset it to vt1 but nothing happens ofc as  you were already on it
[14:26] <didrocks999> from my understanding…
[14:28] <Trevinho> yeah, seems so
[14:29] <didrocks> I wonder if this switch to vt is called:
[14:29] <didrocks>         if (g_strcmp0 (worker->priv->display_seat_id, "seat0") == 0) {
[14:29] <didrocks>                 if (worker->priv->display_mode == GDM_SESSION_DISPLAY_MODE_NEW_VT) {
[14:29] <didrocks>                         jump_to_vt (worker, worker->priv->session_vt);
[14:39] <Trevinho> didrocks: let me check
[14:40] <didrocks> yeah, the issue is really if you fail between state SESSION_OPENED and state SESSION_STARTED
[14:41] <didrocks> (which is where the session pam stack is called)
[14:41] <Trevinho> don't think so though
[14:43] <didrocks> Trevinho: if you make a script failed like /etc/gdm3/PostLogin/Default or /etc/gdm3/PreSession/Default, you enter the SESSION_STARTED state (which is after SESSION_OPENED) and fails here with an exit 1, there is no VT1 and GDM isn’t killed
[14:43] <didrocks> which was why I think this may be the behavior ^
[14:43] <Trevinho> yeah, seems fair
[14:46] <didrocks> (yeah, it tries to switch to VT5 at first when you make the script fails)
[14:50] <KGB-1> gtk4 tags 76aa054 Marco Trevisan upstream/4.3.1 * Upstream version 4.3.1 * https://deb.li/aAGj
[14:50] <KGB-1> gtk4 upstream/latest f165ac6 Marco Trevisan * pushed 1954 commits (first 5 follow) * https://deb.li/3Qn6i
[14:50] <KGB-1> gtk4 upstream/latest 621e0a1 Matthias Clasen gtk/meson.build * Install gtkpasswordentrybuffer.h * https://deb.li/1Av0
[14:50] <KGB-1> gtk4 upstream/latest c2f2a45 Matthias Clasen gtk/gtktreeview.c * treeview: Documentation tweaks * https://deb.li/3BQt2
[14:50] <KGB-1> gtk4 upstream/latest 37acfcf Matthias Clasen gtk/meson.build * Merge branch 'password-entry-buffer' into 'master' * https://deb.li/ZQzs
[14:51] <KGB-1> gtk4 upstream/latest e35e7bc Matthias Clasen gtk/gtktreeview.c * Apply 1 suggestion(s) to 1 file(s) * https://deb.li/3LzwU
[14:51] <KGB-1> gtk4 upstream/latest 6a86acb Matthias Clasen gtk/gtktreeview.c * Apply 1 suggestion(s) to 1 file(s) * https://deb.li/0rsn
[14:51] <KGB-1> gtk4 pristine-tar 85027ec Marco Trevisan (Treviño) gtk4_4.3.1.orig.tar.xz.delta gtk4_4.3.1.orig.tar.xz.id * pristine-tar data for gtk4_4.3.1.orig.tar.xz * https://deb.li/iiAr7
[14:56] <Trevinho> didrocks: don't see `jump_to_vt` to be called here though...
[14:57] <Trevinho> i've only a call to Switching to vt 1 and that's it
[14:57] <didrocks> Trevinho: to the user vt? It doesn’t if you fail in gdm_session_worker_open_session
[14:58] <didrocks> but if you fail in gdm_session_worker_start_session, it will switch to the user session vt first
[14:58] <didrocks> before switching back
[14:58] <didrocks> (at least on a VM on impish, unsure if this is a X/Wayland thingy)
[14:59] <Trevinho> didrocks: I'm still testing with the pam_deny thingy, but I don't see switches happening here
[14:59] <Trevinho> this is a vm and using wayland, but shouldn't matter
[15:24] <jibel> Trevinho, our tests are also in VMs with Wayland FWIW
[15:32] <Trevinho> jibel, OK I think I got it
[15:49] <Trevinho> jibel: so this https://paste.ubuntu.com/p/nZ2SHrJndG/ will fix it, if you guys are in hurry I could prepare a gdm upload otherwise I can leave it to you :)
[15:49] <Trevinho> it's probably something to backport though
[16:05] <ted> marcustomlinson: any thoughts on using a Java module with the LO snap?
[16:05] <ted> I need to open an access database, but it seems the way you do it is with: http://ucanaccess.sourceforge.net/site.html
[16:06] <ted> Seeming like I can't do it with the snap as the CLASSPATH isn't set. But I thought I'd ask if you knew of any way to do it.
[16:18] <jibel> Trevinho, thanks, I'll test your patch, and let you know. We'll need it in focal.
[16:44] <Trevinho> jibel: the full change is https://gitlab.gnome.org/GNOME/gdm/-/merge_requests/143 fyi
[17:27] <jibel> oh, it's merged already. Awesome that was quick. We'll test on Monday
[17:27] <jibel> have a nice w-e everyone
[18:06] <juliank> Hmm, need to upgrade something else to gnome 40, I suppose:
[18:06] <juliank> gsd-media-keys[4818]: Couldn't lock screen: Cannot invoke method; proxy is for the well-known name org.gnome.ScreenSaver without an owner, and proxy was constructed with the G_DBUS_PROXY_FLAGS_DO_NOT_AUTO_START flag
[18:07] <juliank> let's try gsd 40
[18:07] <juliank> it's a bit unhappy about No such key “antialiasing” in schema “org.gnome.settings-daemon.plugins.xsettings:GNOME” as specified in override file “/usr/share/glib-2.0/schemas/20_vanilla-gnome-default-settings.gschema.override”; ignoring override for this key.
[18:07] <juliank> But let's see if it works :D
[18:09] <juliank> gnome-shell depends gnome-settings-daemon needs to be bumped to >= 40
[18:09] <juliank> :D
[18:10]  * juliank should rejoin Debian GNOME team to fix the issues he sees, but will submit PRs for now
[19:21] <k_alam> Trevinho: Hi, just pinging about the merge...