[05:20] hi [05:58] Good morning [06:02] good morning desktoper [06:02] hoi pitti! had a good w.e? [06:03] seb128: bonjour ! oui, samedi était calm, et le voyage à Athens aussi [06:03] oh, tu es en Grèce ? [06:12] seb128: ogra et moi avons regardé le match hier soir :) [06:12] seb128: oui, kernel/foundations/security sprint [06:12] there are worse places indeed [06:12] there's just almost no internet [06:13] well, it slowly drips through the wire(less) ... just a matter of patience [06:13] :) [07:12] pitti, well done on winning your first match yesterday ;-) [07:12] well ... it wasnt pretty though [07:13] (but yeah, the result counts in the end) [07:15] I heard we also won our first match [07:47] moin [07:57] hey Sweet5hark, how are you? had a good w.e? [07:58] Nice weekend, saw a good soccer game yesterday ;) [08:02] morning [08:02] good game Sweet5hark [08:04] moin [08:04] morning Laney [08:07] morning people [08:07] well, the best thing about the game was that there had been some rightwing nutjob ranting about players with immigration background along the lines of "nobody wants boateng as a neighbor". The whole net exploded with "Neuer (our goalie) is very happy to have boateng as neighbor" memes when boateng did his magic defending the goal ... [08:07] hey Trevinho [08:07] :) [08:07] hi willcooke [08:07] what up [08:07] back to EU, at least... In London, so maybe still UE for just two weeks more :'-( [08:07] good morning u.k [08:08] oh, Trevinho already up? [08:08] * Trevinho feels uk too [08:08] Sweet5hark, Bringing on Schweinsteiger at 90mins only have him score 1 min later was great fun [08:08] seb128: yeah, for a while too :) [08:08] seb128: although I'm already back on the correct timezone [08:08] good, aligned on the european, keep it this way! [08:09] not sure how long it will last [08:09] europeans [08:09] :D [08:09] how was the hackweek? [08:09] did everybody had a good w.e? [08:10] we had a nice sunny saturday [08:10] willcooke: oh, yeah. his first goal in a national game since 2011. Go in, make a goal with the first ball contact, game over. [08:10] nice, on friday we also di the "run unity8 in your desktop now" session, and we got some nice stuff working including telegram app, file manager and other stores apps... plus libertine stuff [08:10] which is good because this weeks is going to be rainy [08:10] * Trevinho doesn't see sun since two weeks [08:10] :-/ [08:10] or well, real sun. [08:10] Trevinho, saw that - excellent! I have a plan though..... [08:10] Trevinho, well, telegram&co ... the touch or desktop versions? [08:10] Trevinho, lets get xmir working in mir on x, and then fire up u7 :) [08:10] seb128: yeah, touch versions [08:11] why wouldn't those work? [08:11] on unity8 [08:11] willcooke: yeah, that was my plan too :-D [08:11] they are built for it [08:11] inception [08:11] Trevinho, \o/ [08:11] Trevinho is at Cimis now :P [08:11] it's the new London office? ;-) [08:11] soon andyrock will join too [08:11] so... Mini sprint at cimis :-D [08:11] I can kick him out of the wifi if he doesn't behave D [08:11] seb128: wanna join? [08:11] :D [08:12] Trevinho, that would be nice :p [08:12] I have faster internet than the office and much better coffee B) [08:12] you guys are welcome [08:12] seb128: yeah, they do work... but.... Desktop and phone experience might be different, so.... Not everything always went ok in stock mode [08:12] k [08:12] seb128: and there are various crashes, or things bringing down the entire shell... So, things to do. [08:13] I can imagine [08:13] I have a telegram click for amd64 if anyone wants [08:13] but overall, things are growing... So... [08:13] upload it to the store! [08:13] yeah sprint was positive [08:13] cimi: do that [08:13] really good [08:13] productive [08:14] nice vibe [08:14] and best food in a sprint ever [08:14] prague 2010 excluded [08:14] that sprint had excellent food [08:14] you were young.. a freshment sprint for you... you'd misunderstand for sure :) [08:15] it set the bar way too hight :) [08:15] *high [08:15] and then the following year we had Dublin :') [08:15] seb128, you might remember dublin [08:16] ah, seb128... back in topic... I just remembered we didn't land https://code.launchpad.net/~azzar1/gtk/lp-1574693-shadows/+merge/293082 yet. [08:17] after a couple of days I thought they were just freezing and then defrosting the food we didn't eat, so one day left a mark on a slice of cake and found the very same slice the day after :))) [08:17] seb128: as for the USD thing I was waiting some gsd reviews, but upstream isn't much in the mood of approving the whole thing, so I'd just land our side for now [08:17] * Trevinho has to improve a patch, though [08:19] Trevinho, k for landing usd, we can always improve things later after review, and that would unblock some SRUs [08:19] cimi, I remember Dublin, not sure about the food being the best though [08:20] oh, you said the good food was Prague, indeed :p [08:27] seb128, it was the worst [08:42] Hello world ! [08:48] morning alexarnaud [08:49] willcooke: how are you ? Could you reply to my mail please :) ? [08:50] alexarnaud, did you resend it? Cos I still dont have it [08:51] willcooke: really? Yes, I can. [08:51] alexarnaud, send it to my other address, I'll msg it to you [08:54] alexarnaud, got it! [08:54] willcooke: thanks you. I've re-send it. [08:54] alexarnaud, got it to my Canonical account as well this time [08:55] willcooke: nice [09:50] Laney, when you get a mo, please could you take a look at this review? https://code.launchpad.net/~willcooke/ubuntu-themes/progress-bars/+merge/297169 [09:50] I think I got all the niggles out now [09:50] would be nice to have that in 16.04.1 [09:54] it's on my list already [09:54] but thanks for the poke [09:55] oh nice one, thanks Laney [09:55] I was thinking about making a PPA for some wider early testing [09:56] when it comes to version numbers for the PPA - should I use something like "14.04+16.04.20160415-0ubuntu50" [09:56] (I've never made a PPA before) [09:57] oh, reading a bit more, sounds like "3" would be wiser [09:57] ah, https://help.launchpad.net/Packaging/PPA/BuildingASourcePackage [10:01] seb128: hey, I wonder what's the status of nm-applet's 1.2.0 SRU? cyphermox said he uploaded something, but it's not in -proposed yet === hikiko is now known as hikiko|ln [11:28] happyaron, hey, the SRU queue is public, https://launchpad.net/ubuntu/xenial/+queue?queue_state=1&queue_text= [11:28] but yeah, the SRU team hasn't reviewed much past week it seems [11:29] I can't really help there, maybe try to ping the SRU member of the day from https://wiki.ubuntu.com/StableReleaseUpdates#Publishing [11:38] anyone seeing issues alt-tabbing or alt-graveing between Nautilus windows? [11:44] not me [11:44] what sort of issues? [11:44] but I don't use the by-app switcher [11:44] when copying files from a USB drive I can't alt-tab to the copy progress dialog [11:45] I'll start a fresh session later a open a bug if it's real [11:45] oh [11:45] that's bu g#1575452 [11:46] that's bug #1575452 [11:46] bug 1575452 in unity (Ubuntu) "Copy/move dialog cannot be switched to" [High,Confirmed] https://launchpad.net/bugs/1575452 [11:46] oh, indeed it is [11:46] I showed it to Trevinho and andyrock in Prague :-) [11:46] wonder why I couldnt find it in LP [11:46] thanks seb128 [11:46] yw! [11:46] and I even added it to trello [11:46] ?! [11:46] well, you run the import script [11:47] It was a "feature"... :-) [11:47] that's what you said in Prague indeed :p [11:51] not that I like, eh... it just that there's even some custom code in order to have that behavior, although, maybe alt-tab should be skipped [11:51] or everything === hikiko|ln is now known as hikiko [11:54] g'morning, all. [11:55] hey qengho [11:56] Trevinho, unsure what was the rational but I think in Prague you agreed it was wrong excluding the copy dialog from the alt-tab list? [11:56] hello qengho [12:11] I believe bug 1573052 needs to be re-opened [12:11] bug 1573052 in gnome-software (Ubuntu) "[packaging] gnome-software provides a broken symlink of /usr/lib/gs-plugins-9/libgs_plugin_xdg_app_reviews.so on Ubuntu 16.04" [Low,Fix released] https://launchpad.net/bugs/1573052 [12:22] seb128, ping :) [12:22] could you help me a little with the ppa again? [12:22] hikiko, contentless ping pong [12:22] sure [12:23] I think I've done something wrong when I built the packages [12:23] because when I try to run the dput command [12:23] I get this error: [12:23] http://pastebin.com/dkmGGU9V [12:24] I don't have a .sig or .asc [12:24] I used this command to package compiz for example: [12:24] $ dpkg-buildpackage -us -uc -S -nc -d [12:25] hikiko, do you have a gpg key associated to the email you used in debian/changelog? [12:25] oh, no :) my launchpad key is associated to my gmail and I used the @canonical.com in changelog [12:26] k [12:26] so, what should I do? create another key or change the email? [12:26] well you can debsign -k *.changes [12:26] to sign [12:26] only the .changes file? [12:26] that does sign the .changes and .dsc [12:26] or the dsc/tarball too? [12:27] the tarball is not signed [12:27] but its md5 is in the changes [12:27] you can also use the -k option on the dpkg-buildpackage line [12:27] but not need to rebuild the source to sign only [12:27] right, but it's good to know for next time [12:27] hikiko, oh, also you used "-us -uc" in the command you gave, which are the explicit options to not sign [12:28] \m/ [12:28] copied that from a tutorial [12:28] that's good for local builds [12:28] but if you want to upload to an archive/ppa you need to sign [12:29] dpkg-buildpackage -k is ok? [12:29] or should I add other options too? [12:32] hikiko, you need the -S" if you want to build a source [12:33] maybe I should remove -nc too to build the whole tree [12:34] and -d [12:36] you could yes [12:36] but not needed [12:36] mmm I had this line in .devscripts: DEBSIGN_KEYD= but dpkg-buildpackage ignored it :/ [12:36] KEYID* [12:37] anyway :) [12:37] also seb128 one final question: [12:39] there are instructions on how to delete/add packages to ppa but I wonder if there's some way to update the existing packages next time I have a change (or I just dput the new ones when I have a change?) and if there's any tool local or lp side that can show the ppa changes [12:39] hikiko: just dput the new one with increased changelog version [12:40] for example the changelogs of each package in each commit or something that would help me to see what I push everytime [12:40] hikiko: or you can delete the old src from ppa web ui [12:40] hikiko, try DEB_SIGN_KEYID [12:42] thanks :) [12:43] btw how can someone track the changes in the ppa? for example if I update it every week, is there any tool that could show me something like ppa "snapshots" [12:44] so that I keep track of the progress in this ppa? [12:44] week 1: package 1 state, package 2 state, package 3 state [12:44] I don't think so [12:45] hikiko, the tool used to update the changelog is "dch" [12:45] you usually "dch -i" [12:45] or -v if it's a new upstream version [12:45] then rebuild/sign/upload [12:46] so I can update the changelog and then use it to track the changes :) [12:47] seb128: https://requests.ci-train.ubuntu.com/#/ticket/1482 works for me, so when you want you can hit the publish button [12:47] seb128, thank you very much :) [12:47] hikiko, yw! [12:47] Trevinho, want me to give it a try? [12:48] seb128: if you want to double-check.. [12:48] seb128: mostly the screensaver parts, but they seem to work fine here [12:50] Trevinho, I'm unsure how to test those [12:51] seb128: things like that when screen is locked there's no auto-mount [12:52] k [12:52] seb128: or well, check where screensaver or session-manager proxies are used. I tested most of them, but if you want a quick check too, feel free [12:53] k [13:14] pitti: I don't know if this is still your area, but do we still have a reason not to have poppler-data on the DVDs but install it after install? [13:14] see bug 1591528 [13:14] bug 1591528 in poppler (Ubuntu) "Merge poppler 0.44.0-3 (main) from Debian unstable (main)" [Undecided,New] https://launchpad.net/bugs/1591528 [13:55] seb128: ok, pinged infinity [14:00] happyaron, thanks [14:13] happyaron, seems it worked ;-) [14:17] thanks seb128 happyaron [14:17] willcooke, yw! [14:28] happyaron, would you reply to Chih-Hsyuan's email? [14:28] happyaron, just to say its in progress [14:29] ok will do [14:29] was always wondering how to reply his emails... [14:30] seb128: yeah [14:39] happyaron, I think just a "it's in progress" is fine [14:40] ok [14:49] seb128: could you mark bug 1573052 as triaged? the fix isn't in Ubuntu yet (or in attente's PPA) [14:49] bug 1573052 in gnome-software (Ubuntu) "[packaging] gnome-software provides a broken symlink of /usr/lib/gs-plugins-9/libgs_plugin_xdg_app_reviews.so on Ubuntu 16.04" [Low,Fix released] https://launchpad.net/bugs/1573052 [14:51] changed to fix commited [14:51] which we usually use when there is a fix in upstream git [14:53] thanks, that's an annoying bug because g-software isn't bugfree yet and a lot of people see that error and think it might be responsible for whatever problem they're having [15:06] back to stockholm :D [15:06] this city wants me here [15:07] did the u.k refuse you? [15:10] seb128: nope i never left :D all swedish SAS pilots are basically on strike [15:10] i'll try again tomorrow :D [15:10] lol [15:47] seb128: do you know what debian/outfile is? dh_make seems to output it, but i'm not sure what it's for [15:47] attente, no idea sorry [15:47] just delete it? ;-) [15:47] or ask Laney, he knows those details usually [15:49] not heard of that one [15:49] sounds like it could be a bug [15:49] don't see it as a string in dh-make either [15:49] like it's supposed to be a variable [15:49] I call shenanigans [15:49] * Laney stares at attente [15:51] it's definitely producing debian/outfile... [15:51] Laney, attente, https://anonscm.debian.org/cgit/collab-maint/dh-make.git/commit/?id=ec6863b4ad75e83bab2c5720aca122c1a6e4ca09 [15:51] attente, what dh-make version do you have? [15:51] I guess could be that bug/fix ^ [15:52] seb128: 2.201604 [15:52] k [15:52] that's it then [15:52] Hah. [15:52] lol [15:52] update to 2.201605 [15:52] so that should be debian/source/format? [15:52] yes [15:52] we should probably SRU that update to the LTS [15:52] if you want to do it... ;-) [15:54] sure :) [15:59] bah [16:00] seb128: any chance you can make me an admin of ubuntu-desktop? [16:00] or make a gtk320 ppa [16:00] Laney, enjoy your new admin rights ;-) [16:01] leaving the dmb has some downsides [16:01] thanks! [16:01] seb128: do we sru the version in yakkety or just the one commit? [16:02] attente, depends if the other changes are SRU worth [16:02] Laney, yw! [16:02] seb128: what do you think? https://paste.fedoraproject.org/378560/83376414/ [16:03] don't know enough about packaging to know what's important or not [16:04] i can look into it a bit later [16:04] attente, depending how much you want to bother, I would probably SRU 201605 [16:04] the SRU team is going to want a bug with testcase by logical change [16:05] so depends if you want to create some extra ones [16:05] it's a bit on the too-much side to my taste [16:05] but it enforce testing things rather than including a bunch of fixes to figure out later than one has a regression [16:08] Laney: if you're using GTK 3.20, you'll need to rebuild webkitgtk, webkit2gtk and aisleriot [16:13] That's the tip of the iceberg [16:13] (I've been using it for a while already) [16:15] you haven't been using the gnome3 staging xenial ppa? [16:28] and it needs firefox 48 Beta (or the patched FF 47 in the staging ppa) [16:29] libreoffice-gtk3 needs to be fixed but it's not installed by default [16:33] be calm [16:33] there's going to be time to fix things, no need to brain dump now [16:34] (will certainly appreciate help!) [16:39] hi #u-d! [16:40] hackfest is going awesome so far [16:40] hey desrt! [16:40] nice to read [16:40] what are the topics? [16:40] this morning was really useful [16:40] we came up with a plan for gtk4 [16:40] ...and gtk5, and gtk6, and gtk7... [16:40] i'm gonna write a blog post soon [16:40] going to bump the major version on regular timelines? [16:41] once per two years [16:41] cool [16:41] and gonna have like 4.8, 5.8, 6.8, etc as "forever stable" versions, fully parallel installable [16:41] caveat: 4.0 → 4.2 → 4.4 → 4.6 are going to be unstable (but we do soname) [16:42] but we are going to be very upfront about that and basically say "this is for gnome -- everyone else use the stable one...." [16:42] if a gnome app following unstable starts getting unmaintained then other gnome people can step in and fix it... [16:50] seb128: is there a way to sru a new package for xenial? https://github.com/ubuntu-core/snap-desktop-links for lp:1580740 [17:06] attente, yeah, I think uploading a new package is fine [17:06] attente, see https://launchpad.net/ubuntu/xenial/+queue?queue_state=0&queue_text= as another example [17:06] dinner, bbl [17:07] seb128: thanks [17:07] yw [19:41] night [19:48] qengho: hey, I was told to ping you regarding the chromium-browser package in Ubuntu. In Debian, the chromium package is at 51, but in Ubuntu, it's still at 49, and 50 is FTBFS. I would like to help with getting the chromium-browser package in Ubuntu working again and updated to 51. Would it be wise to fix the FTBFS then update the package to 51, or vice versa if applicable? dobey on [19:49] #ubuntu-devel said that you would be the one to ping [19:49] tsimonq2: Hi! [19:49] hey :) [19:50] oh, but to clarify, Xenial has 50, but with security updates [19:50] and that passed fine [19:50] i was about to say. i'm pretty sure i have 50 installed :) [19:51] tsimonq2: I upload to a PPA here and test and copy. It's going okay for all except Y. I hoped to fix soon. https://launchpad.net/~canonical-chromium-builds/+archive/ubuntu/stage [19:51] yeah, I'm on Yakkety, I've been using saiarcot895's PAA for the dev branch [19:51] *PPA [19:51] oh okay [19:52] tsimonq2: If you can figure out that build error on Y, it will be easy. https://launchpad.net/~canonical-chromium-builds/+archive/ubuntu/stage/+build/9852104 [19:52] qengho: so do you have it under control or can I help? [19:52] oh okay :) [19:52] qengho: then can we get 51? [19:52] tsimonq2: I would be happy if you discover it. If not, I'll get to it soon. [19:52] tsimonq2: Yes! :) [19:53] qengho: awesome :) [19:54] the offending line fwiw: dh_strip.pkg-create-dbgsym: strip --strip-debug --remove-section=.comment --remove-section=.note --enable-deterministic-archives debian/chromium-browser/usr/lib/chromium-browser/obj/net/libnet_extras.a returned exit code 1 [19:56] qengho: how would you usually update releases? do you do it automatically or is it all manual? if it's the former, I'd like to try locally to see if it builds [19:57] tsimonq2: https://code.launchpad.net/~chromium-team/chromium-browser/yakkety-working [19:57] OH I see [19:59] tsimonq2: Note that the X (yay!) and Y (FTBFS) debian/ are pretty much the same. [20:01] qengho: well how do I build it locally? the README instructions aren't working [20:01] let me be more specific, hold on... [20:01] make[1]: *** No rule to make target 'get-packaged-orig-source'. Stop. [20:01] that pretty much sums it up ^ [20:01] tsimonq2: "bzr bd" [20:02] \o/ thanks [20:02] tsimonq2: If it works on your machine but not on the build servers, then you are in the same place I am. :) [20:02] qengho: so is the README outdated or am I just reading it wrong? :) [20:02] qengho: maybe I'll upload to a PPA and try then [20:03] because if it works in a PPA, maybe it just needs a rebuild in the server [20:03] tsimonq2: Nice. Beware it might be too large for a private P A. And rebuild didn't work. And all architectures failed. [20:04] tsimonq2: I hope I'm not scaring you off. :P [20:04] qengho: totally fine, I've dealt with worse :) [20:05] and hey, I'm happy to be working on a package that I use and know how to use quite a bit :) [20:07] qengho: while I'm waiting for this, I might as well mention that in Chromium, there is a PR pending that only keeps support for supported releases and adds Yakkety in src/build/install-build-deps.sh [20:07] that i authored [20:07] so I'm sorta familiar with the Chromium workflow, not entirely [20:07] Oh, for upstream? Nice. [20:08] yep :) [20:09] O_O I ran out of space on my /home partition [20:11] * qengho passes tsimonq2 the liquor. [20:11] qengho: I can't, I'm 14 ;) [20:12] Oh man. [20:14] qengho: but I appreciate the thought :) [20:21] stupid me, I was building in a tmpfs... :/ [20:21] working fine now! :P [20:23] desrt: what do you think distributions are going to ship in your new scheme? [20:24] if we ship an 'unstable' gtk and then want to update it, that's an abi transition post-release [20:29] or does each unstable one still get normal point releases? [20:34] qengho: ...I'm still getting the "Cannot write: No space left on device" [20:34] error, but df -h shows: /dev/sda2 130G 65G 59G 53% /home [20:34] Laney: you will ship all of them [20:35] *shurg* [20:35] * *shrug* [20:35] just like there is gtk2 and gtk3 now, there will be gtk2/3/4/5/6 [20:35] will there be a 7 as well? [20:35] larsu: you can make a difference. please give generously. [20:35] * larsu looks around [20:36] desrt: I mean the 4.0 4.2 4.4 4.6 [20:36] we can only have one of those [20:36] yes. you will ship the latest. [20:36] well, you will do it like: [20:37] libgtk-4-dev, plus libgtk-4-0, libgtk-4-2, libgtk-4-4, etc [20:37] nope [20:37] in the usual way that soname transitions happen... you keep the old non-"-dev" package in the archive until everything is rebuilt [20:37] nope? [20:38] everything has to be rebuilt [20:38] yes... that's what i'm saying. [20:38] but the soname will help with that... you can rebuilt over time [20:38] for debian and ubuntu these will be blocked in unstable or proposed until that's complete [20:39] cool [20:39] which means everything depending on those gets backed up too [20:39] seeing as those things that get backed up are going to be gnome programs, and gnome releases on the same schedule as gtk anyway, that seems pretty reasonable [20:40] new gnome release = of course you have to build it [20:40] fwiw, smcv was an active participant in this discussion [20:42] i'm worried that it will not be this neat in reality [20:42] so are we [20:42] like if the build fixing person is on holiday for a bit [20:42] or random developer needs a new feature [20:42] we are pretty deeply concerned about bus factor [20:42] but isn't on your schedule [20:42] both in terms of upstream maintainer and package maintainer [20:42] are stable releases still normal? [20:43] 4.2.1 [20:43] yes. of course. [20:43] "SRUable" [20:43] okay [20:43] because doing that in a stable release would be fucked [20:43] and once we hit 4.6.x we will probably just keep doing micros there [20:43] so, good :-) [20:43] :) [20:44] the rest of it is still scary [20:44] this is why i wrote this "caution" paragraph [20:44] in practice this is gonna be pretty fine for gnome.... we have a release team too... [20:45] as for the others: probably better if they stick with the stable.... if they don't, you can choose not to package them [20:45] i don't control all the "you"s [20:45] we would need some policy that says that this shit is quick to remove [20:45] smcv's take on this is interesting [20:46] suggestion is: don't do it [20:46] but if an individual package maintainer wants to put themselves on the line as the "person responsible", then they can [20:46] and if upstream disappears, it falls to the package maintainer [20:46] do what? [20:46] ship independent projects using unstable gtk [20:47] his words "a bus factor of 2 is twice as good as a bus factor of 1" [20:47] i think most people will probably just want to avoid the pain... unless they have a very compelling reason [20:48] all in all, we'll see how it goes... people will figure out how it goes... policies will be made.... it'll work itself out [20:48] you won't get to random DD that's found cool-project-of-the-day and uploads it [20:48] * larsu hugs Laney [20:48] imho, that's a debian policy problem [20:48] that's what I said, we would need a policy that lets us remove that stuff [20:49] on the other hand, (again, channelling smcv), cool-project-of-the-day, when it stops building, will be marked as having a release-critical bug and dropped from testing [20:49] we can't have an Ubuntu release (milestone, whatever) blocked by that [20:49] * Laney snuggles larsu [20:49] * desrt joins the snuggles [20:49] * larsu giggles [20:49] ...this is getting weird [20:50] * flocculant is off now ... [20:50] because we're in the same room? [20:51] frittata time [20:51] * Laney periscopes himself cooking dinner [20:52] enjoy! [20:52] Laney: *hug* [20:52] it's all gonna be good :) [20:52] i've had years of poppler and eds transitions [20:52] these things are sad [20:52] hope yours are less so [20:52] in some ways they will be worse [20:52] the first transition is going to _suck_ [20:52] we'll get better at it [20:54] this is something which is quite fine in the flatpak world [20:55] this is nothing new for the distro world either [20:56] Laney: I think you meant to say snappy [20:57] it's not new, but it is a pain point [20:57] larsu: oh yeah, schanppi too [21:00] oh, late evening everyone :) [21:08] ahoy ochosi [21:08] yeah, surprised to see you around at this hour :) [21:09] i just read the gtk4 post and now i'm skimming through all kinds of channel history to read some of the comments :) [21:14] desrt: this is all very interesting. with xfce we waited for a series of gtk3 releases without too many API breaks and then with gtk3.18 it felt like that had come around and then came 3.20 :) [21:14] but good to know it's really going to be 3.26 [21:14] 3.26 is an estimate [21:14] just wondering whether we should suspend our porting efforts until then [21:14] it could be 3.24, 3.26, 3.28 [21:14] but when it happens, you will know which one it is [21:14] having a moving target is just extremely painful (and ifdefy) [21:15] nod. [21:15] and xfce doesn't have the manpower anyway [21:15] if you can wait, i'd consider waiting [21:15] at the same time, not _much_ will change in terms of app-facing API [21:15] :) [21:15] we still follow the same basic rules that we've been following this whole time... we're not going to start going crazy with gtk 3 [21:15] yeah, i supposed so [21:16] so if you want to spread out the work over time, it makes sense to start now [21:16] we'll probably have to do that [21:16] you may have to redo 10% of it or something, but that's maybe nicer than saving 100% of it for later... particularly if you'll have the same manpower problems at that time [21:17] i would've just hoped not having to target a single gtk3 version for the next xfce release (ya know, we only release every two years ;)) [21:18] any suggestion from your side on that? originally i was thinking of 3.18 or something, but now it seems targetting below 3.20 doesn't make too much sense [21:18] (which means we have to redo at least all the included css :'() [21:18] man [21:19] I hope the theme API stays stable [21:19] * Laney cries [21:19] and the worst part for xubuntu as a distribution is: possibly we can ship xfce in gtk3 for one release, but not the next (if ubuntu main decides to go with the next gtk3) [21:19] * seb128 is not going to comment in that discussion ;-) [21:19] simply because xfce upstream won't be able to keep up [21:20] Laney: lol. [21:20] nobody is really [21:20] * ochosi pats Laney on the shoulder knowingly [21:20] I've been Cced on eclipse discussions speaking about going back to gtk2 as default because they can't keep up with gtk3 issues between series [21:20] once we have 3.26 or whatever, it's stable forever [21:21] seb128: tell them that gtk3 will be stable "soon" =) [21:21] seb128: that's what we're trying to fix [21:21] seb128: hi seb128 :) [21:21] hey larsu [21:21] you get to rewrite all of the theme every 2 years instead? :P [21:21] better than once per 6 months [21:22] larsu, well, you "fix" it only if you maintain old series which I doubt is going to happen in practice [21:22] if you don't you just force everybody to move over or be screwed [21:22] so in practice you don't fix anything [21:22] but let's see [21:22] so in practice everybody is gonna be screwed? :p [21:22] seb128: you just said that people are planning to go back to gtk2 because they like that better [21:23] larsu: do you miss all of this already? ;) [21:23] we're about to give people a gtk3 that's just as stable as gtk2 [21:23] why are you bitter about that? [21:23] desrt, yeah, they like it better but it sucks [21:23] it's not touch friendly [21:23] when you report bugs against it you get laught at [21:23] etc [21:23] i'm sorry... you can't have it all [21:23] "new features! fix stuff! change everything!" [21:23] "... WHY DID YOU CHANGE IT?!" [21:23] somewhat other platforms manage to move forward while not breaking in incompatible ways [21:24] +1 [21:24] ask didrocks to tell you about what google does next time you see him ;-) [21:24] other platforms have an order of magnitude or 2 more developers.... [21:25] right, I'm not saying it's easy or doable [21:25] just that other platforms do it [21:25] :> [21:25] and we are not going to win devs if we screw them [21:25] having good reasons or not [21:25] at worst, this policy just aligns us with what Qt does [21:26] new incompatable (but parallel installable) major releases on a timeline measured in years [21:27] yeah [21:27] though 5 to 10 years is a better timeframe to break things [21:27] more frequently than that is not going to get you appdev sympathy [21:27] ...and then you get people complaining that gtk2 is stable, but too damn old [21:28] and it's not like we really break you once per two years. you can stay with the old one forever. [21:28] yeah, but nobody is going to fix bugs on those series [21:28] gtk is a free software project in which people do work that they think is valuable [21:29] if people want to step up to fix bugs on stable releases, because they think it is valuable work, we would welcome them with open arms [21:29] nobody has really done that, though [21:29] so apparently nobody really thinks that it's valuable enough [21:30] somewhat true [21:30] though in practice some people do [21:30] but their patches never get reviewed [21:30] and they go away [21:30] we should get better at that [21:30] i'll admit [21:32] (we're talking at the hackfest now about how to fix this) [21:34] (great) [21:35] desrt: i appreciate all your efforts and that you're a small team - and sorry to get back to "my problems" :) - but what would your advice be now for a small desktop like xfce? [21:36] which version can we meaningfully target with a stable release - obviously maximizing the hope that distros can ship it with "some gtk3 version" without major breakage? [21:36] i'm still focused on the present situation ("road to 3.26" or so), not the overall long term strategy [21:37] would you recommend to just target 3.20 (and exclude all distros that ship something below) or even try to target 3.22 already (knowing that it'll take a few more months until we're release-ready)? [21:38] we're talking about your situation in the room [21:38] (atm we're caught in <=3.18 | >=3.20 ifdef hell) [21:38] opinion: start porting now. release in a year. [21:38] and target whatever is around until then? :) [21:38] we're not going to be able to commit to a specific gtk version being the "last 3 stable" until the gnome release team gets involved [21:38] more or less, ya [21:39] ok [21:39] thanks for discussing this [21:39] (honestly) [21:41] seb128: how invested is ubuntu desktop in gtk anyway atm? i thought you guys were moving towards qt with convergence? [21:42] ochosi, we try to keep up with upgrades as best effort thing [21:42] Laney stepped up to redo the them and update to 3.20 [21:43] if he didn't do that I think we would probably have stayed on 3.18 for some cycles [21:43] what, so 16.10 is definitely going to go 3.20? [21:43] Laney: did you really think that through? :] [21:44] ochosi, don't make him think too much [21:45] I think the guy is slightly crazy [21:45] but let him run for it, he might get us there ;-) [21:45] (sry, don't have time to follow your discussions during the day because of $dayjob and the last i heard was you had put off 3.20 for at least another release) [21:45] heh [21:45] well i've looked at what's to do there (and i likely will have to do it myself for xubuntu), it's definitely not nothing [21:46] i think rebasing on top of adwaita is the only feasible way [21:49] I don't know [21:49] but Laney has been head down full time on the theme for like 10 days [21:49] we don't see him much on IRC anymore [21:49] he seems to have it mostly done now though [21:49] wow, nice [21:49] unsure how much impact it has on other things and other flavors [21:49] well, same impact it had on you [21:49] we can't really land screwing other flavors [21:49] and I doubt he's going to want to port your theme as well [21:50] so dunno what happens to the landing then [21:50] haha, meh ;) [21:50] tbh it's not just that [21:50] all applications that have some css embedded (and quite a few do) will need updates as well [21:51] right [21:51] most are GNOME code though [21:51] and they fixed their stuff [21:51] hah, you'd think [21:51] well, most of non GNOME don't use gtk3 :p [21:51] i know of at least two apps off the top of my head which are not [21:51] yeah yeah [21:51] :) [21:51] i helped port lightdm-gtk-greeter to gtk3 and it has custom css [21:52] then there are some xfce apps that i worked on where i tried to be "modern" and keep up with gnome [21:52] anyway, if we know in advance, maybe we can get most of that fixed / ifdefd [21:52] but still, it'll boil down to how many ppl want to sit and code in their free time during the summer [21:54] i gotta get some sleep now [21:54] thanks for the discussion though [21:54] night everyone! [21:55] and thanks for the advice, desrt, we'll see what model we'll end up with... [21:55] ochosi: thanks for hanging in there ;) [21:55] we're trying to get better. this new system is the first big step. [21:56] well what i can already tell you as a positive feedback is that now at least there seems to be an announced longterm roadmap [21:57] so that is helpful either way [21:57] thanks for the feedback =) [21:58] i'll try to be more communicative about all this, just don't have enough free time these days [21:58] anyway, off to bed now [22:13] jbicha, did you file bug 1592001 upstream? [22:13] bug 1592001 in gnome-software (Ubuntu) ""OS Updates is now installed" is bad English grammar" [Undecided,New] https://launchpad.net/bugs/1592001