[00:00] well, say thanks to James Hunt for finding and fixing the problem ;) === m_conley_away is now known as m_conley === m_conley is now known as m_conley_away === duflu_ is now known as duflu [06:17] Laney, hey, can you give codesearch a bump! [08:04] it's already in progress [08:04] also morning [08:07] good morning desktopers [08:07] hey Laney, how are you? [08:07] what is in progress? [08:07] evening ;) [08:07] 08:17:33 darkxst | Laney, hey, can you give codesearch a bump! [08:07] seb128: ^ [08:08] didrocks, thanks [08:08] yw! [08:09] hey seb128, I'm alright thanks [08:09] it's finally not raining [08:10] how are you? [08:10] Laney, same here have had 3 days in row without rain ;) [08:10] I'm good thanks [08:10] *days*! [08:10] no rain here either, but it's still grey [08:19] bregma, seb128 i think there was a mistake in the last unity release [08:19] https://code.launchpad.net/~andyrock/unity/lp-1281058/+merge/215331 should not be there [08:20] andyrock, hey, why not? was it not ready? [08:21] yeah but what about the translations? [08:21] andyrock, well, that landed to utopic, not trusty [08:21] https://launchpad.net/ubuntu/+source/unity/7.2.0+14.10.20140512.4-0ubuntu1 [08:21] ah yeah [08:21] sorry :D [08:21] just wake up ^_^ [08:21] I guess you need to target lp:unity/7.2 if you want a SRU [08:21] no worry ;-) === tsdgeos_ is now known as tsdgeos [11:23] Sweetsha1k, hey, did you see that somebody pointed to a commit for that libreoffice segfault bug you upstreamed yesterday? is that commit in our current version? [11:25] http://ubuntuone.com/5OeS9F9yz8DogeokQRYCWY [11:26] Laney, is that an unity8 iso? ;-) [11:26] yeah [11:26] launching things doesn't work ... [11:26] oh some things do [11:26] kind of [11:26] do you have unity8-desktop-session-mir installed? [11:26] using x11 [11:27] oh, k [11:27] i should try mir [11:27] Laney, sorry, I didn't push the seed changes yesterday, Rick pinged me for slidedeck work and I noticed that I lost my seed when I whipped out my "isobuild" directory on friday (when your script was failing for me) [11:28] it's okay [11:28] I made a quick one by copying touch [11:28] Laney, but I basically copied the touch one, added unity8-desktop-session-mir instead of the custom session, and dropped everything that looked like android related [11:28] cool [11:28] lp:~laney/ubuntu-seeds/ubuntu-unity8.utopic I think [11:29] seb128: I saw that, didnt check yet on which branches it is. === MacSlow is now known as MacSlow|lunch [11:29] did mir work in a vm? [11:29] Sweetsha1k, ok, let me know how it works [11:30] Laney, I didn't try, might be a good question for #ubuntu-mir, my gut feeling is that it's not going to work... [11:31] you could cheat and use the x86 touch emulator ;) [11:31] Failed to create GBM device [11:31] yeah I think this is a dead end [11:31] (since thats just a qemu VM after all) [11:32] doesn't that do android and stuff? [11:32] it uses the andorid drivers inside the image and a special translation layer that makes qemu use desktop GL [11:32] seb128: just checked, seems to be on master only (aka for release with LibreOffice 4.3.x), so yeah, would need backporting upstream or as vendor patch. [11:33] Sweetsha1k, great, at least it's not "we already have the patch, so it's not resolving the issue" [11:34] seb128: yes [11:34] Laney, I think not working in a VM is alright === Sweetsha1k is now known as Sweetshark [11:34] well "alright", that's not a blocker [11:34] even if having VMs working would be nice [11:34] * Sweetshark reboots for release update ... [11:35] I think I need to get a USB key [11:35] amazingly I don't think I have any [11:35] how do you usually install Ubuntu on new machines? [11:35] haven't done that in a while [11:36] probably had one back then [11:36] k, makes sense [11:36] I don't use mines a lot, but I've a stack of those [11:36] like I get a few at conferences, though they are old one slow/1G style [11:37] needs changes to /etc/upstart-xsessions & unity8.conf to work at all though, /me uploads those [11:40] bregma: did you get anywhere with the unity8-x11 session? [11:40] It doesn't work very well for me [11:42] Laney, bregma: do we care about unity8 on xorg? [11:42] it doesn't look like something we want to spend time on [11:42] we better start on Mir [11:42] I'm just wondering [11:42] imho we should just drop it [11:43] we are not going to have stuff like clicks working on xorg [11:43] due to the security model [11:43] so let's just focus on Mir [11:52] Laney, your seed looks mostly fine to me [11:53] Laney, we should probably add those "camera-app, friends-app, gallery-app, mediaplayer-app, notes-app, webbrowser-app" [11:53] those are the apps unity8-desktop-session-mir suggests, they are all debs in the archvie [11:53] (touch install them as clicks) [11:54] are the debs maintained? [11:54] yes [11:54] they go through CI train [11:54] well, that's temporary until we get click support on the desktop I guess === alan_g is now known as alan_g|lunch === oCrazyLemn is now known as CrazyLemon [12:19] pitti, hey [12:20] pitti, could you have a look to https://bugzilla.gnome.org/show_bug.cgi?id=728205#c4 , not sure if you know about those changes/have a suggestion on what needs to be adapted? === MacSlow|lunch is now known as MacSlow [12:22] hum, http://sourceforge.net/p/gphoto/bugs/974/ has some details [12:26] hum [12:26] http://ubuntu-codesearch.surgut.co.uk/ gives a 502 [12:30] * ogra_ thought he saw Laney mention it above [12:30] (that it just gets updated, not that it is down thoough) [12:41] it will return soon [12:41] mlankhorst: the qtbase SRU with 1307701 fix is now in unapproved queue. [12:41] just napping [12:42] hmm, the gstreamer SRU updates are also still in unapproved :( [12:43] Mirv: goodie :-) === alan_g|lunch is now known as alan_g [12:48] Mirv, do they fix any issue you are waiting to see resolved? [12:51] seb128: just the gst-libav bug #1290368 when seeking h.264 files which I was happy to see Laney's upload early on [12:51] oh, right [12:51] maybe nag on #ubuntu-release about it? [12:51] since it crashed also eg. thumbnailers [12:51] * seb128 is also unhappy about the lack of action on the SRU queue [12:51] yep, I'll give a note there about at least that one [12:56] Mirv, thanks [13:01] seb128: queueing that for investigation (sorry, in meeting / busy with other stuff) [13:01] pitti, no hurry, thanks [13:02] seb128: my phone mounts quite happily, but that might be because it's got a special rule or so [13:02] pitti, your phone is a mtp device, not a ptp one, right? [13:02] seb128: right; but my camera mounts fine, too [13:02] pitti, that issue is about libgphoto/ptp cameras [13:02] (and that's ptp) [13:03] k, mine doesn't [13:03] yeah, will have a look, and perhaps drop the cam specific rule [13:04] danke [13:20] tseliot: Hi! We've been seeing more complaints about black borders around windows after resuming from Suspend w/ Nvidia gfx: https://bugs.launchpad.net/ubuntu/+source/compiz/+bug/1292830 . I thought there was a patch provided by nvidia that helped alleviate this. [13:26] ChrisTownsend: I think the nvidia driver is already doing the right thing on resume. The symptoms look a little different from the ones in the other bug report [13:28] tseliot: Hmm, ok. Is that nvidia patch supposed to switch vt's upon resume or am I not understanding what it is supposed to do? === m_conley_away is now known as m_conley [13:28] ChrisTownsend: I don't know if this is relevant but it looks at least suspicious (in dmesg): compiz[16459]: segfault at 0 ip 00007f4bd4adb96e sp 00007fff51885ae0 error 4 in libopengl.so[7f4bd4aa5000+4a000] [13:29] tseliot: I also fixed the routine in Compiz to force a full redraw of 10 frames when resuming which was broken for a release or 2 due to upower no longer sedning a mesaage that it is resuming. [13:30] ChrisTownsend: right, that's the "other bug" I was talking about (I was the one who suggested the fix there) [13:30] tseliot: Yep:) [13:31] tseliot: So I guess I'm struggling to figure out why folks are still seeing black borders upon resume. I unfortunately have no nvidia hardware to debug this. [13:33] ChrisTownsend: so, yes, the patch for nvidia is just to require a vt switch on resume. Your patch reinitialises FBOs when dealing with resume. I'm not sure what we're missing here. Let me try again on my laptop [13:34] tseliot: The compiz segfault you pointed out might be a Red Herring since there would be no Launcher, Panel, etc. if Compiz was dead. [13:35] tseliot: Ok, thanks! [13:37] ChrisTownsend: unfortunately it works well here and I can't reproduce it. I have a hybrid system here, so maybe this can be reproduced only on non hybrid systems (i.e. with only nvidia, and no intel graphics) [13:38] tseliot: Hmm, yeah, maybe that's the key. Ok, thanks for checking! [13:39] ChrisTownsend: np [13:53] seb128: hm it seems that gtk+ can do grabs too :/ [13:53] ChrisTownsend: one more thing, the original reporter wasn't using an nvidia driver with the fix (available in >= 331.38-0ubuntu5) [13:54] tseliot: Right. My concern is that there are still comments being added to the bug. I've asked for more info from people that are still seemingly affected by this. [13:54] like if you do crazy stuff on the background [13:55] ChrisTownsend: yes, I asked more information in the bug report to see what they're running [13:56] tseliot: Heh, I didn't refresh the page, so we both asked. At least people will see we are trying:) [13:56] ChrisTownsend: :) [13:57] seb128: what are the grabs in gtk+ for? [13:57] mlankhorst, menus do grabs at least [13:57] for keyboard navigation [13:58] afaik [14:00] i was still hitting some bugs because of the xigrabdevice calls [14:00] probably less harmful on the master device, though.. [14:04] Trevinho, bregma: do you guys plan to keep unity-SRUs-for-trusty coming? we didn't get that many fixes landing since release :/ [14:04] at least I hope it's on the master device [14:05] seb128, yes, the latest landing was blocked by some issues (now resolved) and by vacations [14:05] bregma, great ;-) [14:05] bregma, I hope you enjoyed the deserved time off ;-) [14:06] * bregma cues up Pharrell William's "Happy" [14:06] ;-) [14:07] seb128, do you anything about translations for SRU bug #1281058 ? [14:08] that's the logout message bug [14:08] bregma, yes, I told andyrock that I would update the template, but we are waiting on design to confirm that the string is what they want [14:09] bregma, let's not include that change yet though, we need to give translators some time before landing the code change (and a langpack update with the updated translations) [14:10] seb128, Canonical OEM want to know when it will be ready, that particular issue is giving them an itch [14:10] pitti, dpm: do we have a schedule of langpack updates for trusty? [14:12] seb128, unfortunately not, I no longer have the time to coordinate those updates :( That said, we should do at least a one-off post-release update [14:12] bregma, can you check if OEM would be happy with early june? [14:13] bregma, that seems a reasonable timeframe if we need to collect translations and get langpack updates [14:13] bregma, works for you? [14:14] seb128: ok nautilus appears to grab master, I can live with that :-) [14:14] mlankhorst, what issues does that create? [14:14] seb128, does the SRU depend on the langpack update, or is it the other way around? [14:14] * bregma is not an expert on i18n in Ubuntu [14:15] seb128: not many, though it would explain the warnings I was seeing in the Xorg log, I'll make a test for that :P [14:16] bregma, well, we want langpacks to be updated before the code land, the translations are made on launchpad unity/ubuntu, which means they are not going to be in your unity upload, but in the next langpack export [14:17] bregma, if we land the code first it's fine, but we introduce english strings to users [14:17] mlankhorst, k [14:17] bregma, if we export the langpacks the translations get available on disk, and when unity's code changes land we get nicely translated strings on the screen ;-) [14:20] Hi. Not sure if this is the case already. It isn't on windows and should be: a progress spinner / hourglass cursor should take exactly one second to rotate [14:20] That way we can visually count time === alan_g is now known as alan_g|tea [14:20] Is this the right place to advocate such an idea? [14:23] davidsong, hey, well first step would be to check if that's the case already or not === alan_g|tea is now known as alan_g [14:46] seb128: /me is happy that he was right for one with the spurious libreoffice on precise fail. [14:47] Sweet5hark, ;-) [14:49] seb128, bregma i emaild JohnLea [14:49] *emailed [14:50] waiting for an answer [14:50] saw that, thanks ;-) [15:05] andyrock; when did you email me? (and about what?) - sorry, I get too much email! [15:05] JohnLea, last friday IIRC [15:22] andyrock, seb128; just replied to the email - good you pinged me here, I had completely missed it - please ping me on IRC for anything urgent [15:22] JohnLea, thanks [15:24] bleh, not making much out of it yet, touch sucks [15:24] I think that will be my status update for this week [15:25] "who wants to touch things anyway?" [15:27] Laney: https://www.youtube.com/watch?v=gxm_mwz9sRo [15:28] * Laney gets hammerrolled [15:28] Hammerzeit! [15:29] good idea [15:29] * KombuchaKip waves [15:30] oh, it's meeting time [15:30] ah he got the hint :D [15:31] qengho, Sweet5hark, mlankhorst, Laney, tkamppeter, desrt, attente, larsu, kenvandine(?), KombuchaKip: hey, it's meeting time! [15:31] Yo! [15:31] * desrt is getting uncomfortably close to the middle of the list [15:31] JohnLea, thanks! [15:31] desrt, ;-) [15:31] ok, let's get start [15:32] qengho, hey, there? [15:32] Hey! [15:32] Er, not ready. I'll go next. [15:32] k [15:32] Sweet5hark, hey [15:32] - Bumped LibreOffice 4.2.4/trusty in the PPA [15:32] -- note: there were some Qt issues with the precise backport [15:32] - Bumped LibreOffice 3.5.7-0ubuntu6.1 for SRU to precise-proposed [15:32] -- thanks to seb128 and everyone else enduring that endless /pain with me [15:32] - Bumped LibreOffice 4.1.6/saucy in the PPA [15:32] - Build >180 full builds of LibreOffice covering the range from 4.2 branch-off to 4.3.0~alpha1 for https://wiki.documentfoundation.org/Bibisect [15:32] - also: Hammerzeit. [15:32] EOF === gatox is now known as gatox_lunch [15:33] Sweet5hark, how is the menu issue going? still tracking it down? [15:34] seb128: didnt get to it much this week, but have an idea how to fix it. [15:34] ok, let's see next week then ;-) [15:34] Sweet5hark, thanks [15:34] qengho, ready or want to be next next? ;-) [15:34] - in-progress: chromium bugs. 1) sub-window placement. 2) tabs disappearing. 3) [15:34] - done: two sick days end of last week. [15:35] EOF [15:35] qengho, I hope you are feeling better [15:35] er, 3, I was looking for notes. Still, one and two are big enough that I problably won't get to it. [15:35] I am, thanks, seb128 [15:35] good [15:35] qengho, thanks [15:35] mlankhorst, hey [15:36] hey [15:36] tons of pointer bugs, and other xorgy bugs [15:36] oh wait my status update was supposed to be [15:36] "touch sucks" [15:36] ^D [15:37] haha [15:37] oh and lts-trusty landing [15:37] good work on the qt/xorg/dnd-segfault-on-touch issues [15:37] but mostly touch sucks [15:37] np :-) [15:37] mlankhorst, thanks [15:37] Laney, your turn [15:37] was with dandrader's help though [15:37] • Working on a desktop flavour with unity 8 and touch apps. Try a build with http://people.canonical.com/~laney/random-scripts/build-ubuntu-iso & livecd-rootfs from ppa:ubuntu-desktop/unity8-flavour [15:37] ∘ Next TODO: try on real hw (Mir doesn't work in a VM apparently) [15:38] ∘ Then get dailies built using the cd image infrastructure [15:38] • u-s-s reviews and test fixes [15:38] • cross-building fixes for ofono-qt and signon so that u-s-s can be x-built in utopic again [15:38] • cheese 3.12, MIR appdata-tools [15:38] • vte3 0.36.1 with rewrap on resize [15:38] • Resurrect a lost gnome-terminal patch to take the default transparency value from the theme. We used to have a default purple & slightly transparent terminal but that got disabled at some point so it was just solid purple. In desktop PPA, not distro yet as it's on by default so people will see their settings changed (in other words, I'm scared). [15:38] • Start on webkitgtk 2.4.2, probably will SRU too as the current one is still waiting in trusty-proposed [15:38] • DMB: start sil2100's email application [15:38] ☁ [15:38] oh, speaking about DMB [15:38] is Sweet5hark's one still being a thing? [15:38] I tried to hassle [15:38] let me email them right now [15:38] somebody said it would be handled after the election iirc [15:39] thanks [15:39] Laney, thank you for getting the unity8 iso bootstrapped, I'm going to start a build/test on a real hw after the meeting as well [15:39] I don't know if it needs to include drivers or something [15:39] bregma can maybe assist at this point [15:40] feel free to update the seed in the ppa until it works [15:40] ok [15:40] Laney, thanks [15:40] tkamppeter, hey [15:40] cups-filters: Started working on rastertopdf filter to make a CUPS server emulate an IPP Everywhere printer [15:41] cups-filters, ghostscript: Bug fixes [15:41] min12xxw: Synced with Debian [15:41] Bugs. [15:41] tkamppeter, thanks [15:41] desrt, hey [15:42] hey [15:42] Solved problem of evince not displaying some files correctly with Marek Kasik and seb128. [15:42] brought the XDG_CURRENT_DESKTOP saga to a close by implementing proper support for it vs. OnlyShowIn/NotShowIn in GAppInfo. committed, bug closed, everyone seems happy [15:42] great ;-) [15:42] there were some issues there caused by an old gnome-session patch that added this support which broke things due to the ':' in some sessions now (ie: GNOME-Classic:GNOME) [15:43] also landed the Implements= support, so now you can find all desktop files that do a particular interface [15:43] larsu: for you ;) [15:43] : [15:43] :) [15:43] landed the patches for (nullable) and (optional) support in gobject-introspection and that caused some fallout so i spent a day or so chasing around issues and fixing them [15:43] also did some reviews of kdbus stuff [15:44] finally, did some initial looking into a gtk backend for mir -- there is existing work here, but it's not particularly complete, and there are some hard problems we face, particularly going forward [15:45] i think we may want to seriously consider looking at using the wayland backend in gtk and creating some sort of a bridge for wayland/mir instead of (essentially) copypasting the wayland backend to create a new mir one, doing a huge sed patch, then fighting to keep up to date with constant gtk changes [15:45] will discuss that with robert as soon as i see him... [15:46] the good news is that the gtk guys are relatively openminded about mir support -- only concerned about maintainership (which is a concern i share) [15:46] ok, noted, I'm going to take that one up for discussion, trying to get a feeling if that's an option or if that would make people unhappy at us [15:46] so even if the wayland backend needs some tweaks to accomodate our mir bridge, i think they'd be open to it [15:46] great [15:46] and we'd still be in a better place than having an entirely separate backend that none of the upstream developers ever compile/test [15:46] well, let's see how other people are feeling about that [15:46] do you have any idea at what level that bridge would be? [15:47] would it be a Mir thing? like a library that makes wayland clients talk to Mir? [15:47] i'd ideally want the mir guys to make this part of mir, but rick is concerned that it would slow down their development [15:47] desrt: whats the timeline on this btw? One way to bring LibreOffice to MIR native is completling the gtk3 backend. Sooo that would be the fun of stacking an alpha gtk3 backend for LibreOffice on a beta MIR-backend for gtk? [15:47] so the idea for now is to see if we can possible to it as a third component, or maybe some library-type thing that we could load [15:47] they were talking about having that compat layer [15:47] ok [15:48] i'm not crazy about shims [15:48] we could just use X if we wanted to do that, right? [15:48] but at least this will be a much thinner shim... [15:48] ideally, for the future, we could natively connect wayland clients to a mir-based display server [15:48] needs discussion, though, for sure... [15:48] right [15:48] there's another side of that, as well: [15:49] the "application logic" stuff is a bit more involved in the mir world than in wayland (where it basically doesn't exist) [15:49] even with the bridge, we'd probably want to get gtkapplication participating a bit more actively [15:50] so even if we don't get a mir gdk backend, i suspect we'll want to have a mir gtk(application) backend anyway... [15:50] ah, that would be nice [15:50] doesn't seem like something that is going to change a lot either [15:50] could maybe push some of our ideas at upstream there as well, since it seems that they basically don't consider the problem at all yet... [15:50] (e.g should be easy to keep up with upstream on that, right?) [15:50] ya... [15:50] i recently did a substantial refactoring of GtkApplication to make backends easier and more self-contained [15:50] was turning into #ifdef hell in there.... [15:51] cool [15:51] anyway... that's enough detail, i think [15:51] yeah [15:51] desrt, thanks for the detailed update ;-) [15:51] some of those topics are going to be easier dealt with at the client sprint at the end of the month [15:52] rick wants some sort of a vague idea of a plan before the sprint [15:52] (we are going to keep updates coming for people who are not there during that week though) [15:52] so i'll be chatting with robert :) [15:52] that sounds good [15:52] desrt, thanks [15:52] attente, hey [15:52] hi seb128 [15:52] emitting menu shows and hides using submenu-action in u-g-m, need to backport a gtk patch before it can be proposed [15:52] figured out why using shift as an input source switcher doesn't translate capital letters in the correct group properly, but wasn't able to fix it [15:52] sru the group locking fix for g-s-d, thanks seb128 [15:52] eof [15:52] desrt seb128: (punting my LibreOffice on gtk3 on MIR questions for the sprint then) [15:53] Sweet5hark, right, too much details to sort out to reply yet [15:53] attente, do you have a mp for the gtk change to backport? [15:53] or a bug reference? [15:53] https://bugzilla.gnome.org/show_bug.cgi?id=729820 [15:53] https://bugzilla.gnome.org/show_bug.cgi?id=729820 [15:54] desrt, attente: thanks, I'm going to have a look after the meeting [15:54] fwiw, the patch is obviously correct but mclasen has a nice idea that we may also want to try [15:54] attente: let's talk after [15:54] seb128: thanks [15:54] desrt: ok [15:55] desrt, let's get that patch in, we can revisit/do other changes later if needed [15:55] well, you guys keep me updated ;-) [15:55] but if not told otherwise I'm going to review/land that patch to start [15:56] seb128: there are style issues, of course.... :p [15:56] lol [15:56] attente, thanks [15:56] larsu, your turn [15:56] yay [15:56] talked a lot this week with tvoss and push notification people [15:57] I'm glad I've joined in on that discussion to make them think about the desktop use cases a bit more (thanks seb128) [15:57] also, looks like we're getting the transparency fix for gnome-terminal [15:58] rishi has been working on that [15:58] larsu, thanks for joining, I hope they are going to use your input in that discussion [15:58] did he do the theme thing? [15:58] seb128: I hope so too [15:58] what's this transparency thing all about? [15:58] Laney: don't know, I don't have a bug pointer yet and he didn't respond back after last night [15:58] people want it? people don't want it? [15:58] desrt: chpe removed transparency upstream [15:58] !chpe wants it [15:58] but we figure we'll get a huge backlash if we upgrade and don't have it anymore [15:59] seb128: I've also fixed a small issue in the theme relating to that (give toolbars a background) [15:59] larsu, I saw that one earlier, I'm going to look at what is pending and organize some landing [15:59] that's an issue in the vte test app which might also happen in g-t itself [15:59] thanks [15:59] not urgent of course [15:59] real transparency or fake transparency? [15:59] I didn't watch trunk stuff too much recently, still focussed on trusty [15:59] desrt: real [15:59] 'cause real transparency is actually kinda useless :) [15:59] ya [15:59] but people want it [16:00] I totally agree and never have it enabled - I don't even know if we enable it by default [16:00] I also did some more 3.12 testing [16:00] the theme is looking good, I couldn't find any major issue [16:00] *issues [16:00] it [16:00] is the update still blocked on the bash cwd issue to be resolved? [16:00] it's even better now, because the black bg problem is gone [16:01] seb128: yes it is. Laney said they had a fix but reverted [16:01] deferred to sprint like everything else [16:01] k [16:01] and I've started looking into theming headerbar'ed apps correctly [16:01] right now we don't have any theming for those [16:01] good [16:01] which is why those apps look _really_ bad [16:01] "some more 3.12 testing" ... that's about GTK righT? [16:02] I'm unsure if we want to patch header bars out of dialogs though [16:02] (I've been running that on my test machine for some days) [16:02] it's a tiny patch (I think) and those are used everywhere [16:02] seb128: yes [16:02] that seems like another topic for the sprint week :p [16:02] right [16:02] it runs well, I've now tried to test all the apps and stuff [16:02] especially the ones were we had theming problems last cycle [16:03] yeah, I've it running fine [16:03] out of [16:03] - the headerbar issues [16:03] - the buttons on the dialog [16:03] - the theme [16:03] the buttons in dialogs look correct for me [16:03] button, like try to shift-suppr in nautilus [16:03] well, at least the theming is not right [16:04]