[03:07] duflu: hi [03:08] I was pretty sure Caribou suffered from LP: #1698515 but I hadn't spent the time to verify that [03:08] Launchpad bug 1698515 in gnome-shell (Ubuntu) "On screen keyboard stays English even after changing language to Finnish" [Undecided,New] https://launchpad.net/bugs/1698515 [03:09] I think we'd be a lot better off with Onboard, but note that 1) Onboard doesn't work in Wayland yet and 2) is dependant on caribou in some parts of the UI [03:09] jbicha: Yeah but Caribou and the Shell OSK are separate, mostly, right? [03:09] you can try caribou by install gnoem-shell-extension-caribou and enabling the extension [03:09] I have tried it. It's not the same thing though [03:10] duflu: The Shell screen keyboard *is* caribou [03:10] Really? I thought Caribou was the little popup keyboard only [03:10] caribou has been undermaintained for years in GNOME [03:11] Caribou is the popup Shell screen keyboard [03:12] I think GNOME would be better off switching to Onboard but it would take a bunch of work to integrate it well [03:12] I swear when I tested it, Caribou was a different keyboard (tiny) and wasn't the shell keyboard (full screen width) [03:12] Caribou is just too minimal [03:12] there's 2 versions of Caribou: the one built into Shell and one that can run standalone in a different desktop [03:13] jbicha: Regardless, I think working with upstream Gnome would be best. I don't have confidence in onboard even after a few years [03:13] I think caribou-antler is the standlone one? [03:13] part of Onboard's problem is that they are not GNOME [03:14] if GNOME were to decide to work with Onboard then I don't think it would take long for Onboard to work in Wayland [03:14] Onboard is not reliable at rendering the correct size or correct proportions. So I would not endorse it without upstream integration work [03:14] and I think Onboard could handle the system dialogs (like Activities Overview, unlock prompts) [03:15] I think Caribou's keyboard doesn't support more than a few languages [03:16] and I think Caribou will be worse for accessibility [03:32] jbicha: Despite the bugs, I prefer what Gnome Shell is doing right now. :) [03:40] ok [05:43] good morning desktoppers! [06:51] good morning [06:56] morning [07:15] Hello, oSoMoN, didrocks, jibel [07:41] morning all [07:43] hey willcooke [08:00] willcooke: morning [08:00] afternoon duflu [08:03] willcooke, libinput 1.7 sounds interesting. I think we should get it in artful. Although 1.8 is coming soon too [08:04] AH MAN [08:05] HI LANEY [08:07] looks like my VPS done got rebooted [08:07] good morning desktopers [08:07] hey duflu willcooke Laney [08:07] Hey seb128 [08:07] hi duflu hi seb128 [08:07] you good? [08:07] duflu, what's new in libinput 1.7? [08:07] re/hey guys ;) [08:07] haha [08:07] down to business [08:07] seb128, https://lists.freedesktop.org/archives/wayland-devel/2017-March/033531.html [08:08] Laney, a bit tired but good! w.e was busy here, friends over for dinner on friday and sunday evenings and some other friends came for coffee on saturday [08:08] and election day yesterday [08:08] coffee [08:08] you dutch! [08:08] and almost 30°C which is too much [08:08] lol [08:08] tea, right? :p [08:08] Laney, how was your w.e? [08:09] good, went down to see family for my dad's birthday & fathers day [08:09] was over 30 at some points [08:09] and hey didrocks, how's it going? [08:09] nice weather for a bbq! [08:10] going to be less fun sitting at the computer today [08:10] indeed [08:10] duflu, indeed, nice update, I'm pretty sure we are going to get it in artful, it's still early in the cycle [08:11] seb128: Assuming the ABI is well maintained. Should be easy [08:11] Laney: very good, thanks! deep into session swapping ;) Weather was great, still at my parent's to a couple of days. Great to be in slightly fresher moutains environment than warm city. [08:12] * didrocks sees http://www.meteofrance.com/previsions-meteo-france/lyon/69000 and is not eager to go back… [08:12] 38 on Thursday… [08:14] ._. [08:19] shit [08:19] I forgot to run irssi in screen /o\ [08:46] oh noes [08:49] * Laney rips off the plaster [08:49] ze hive mind [09:20] morning! [09:21] andyrock, morning [09:21] morning andyrock [10:00] andyrock: waow, seems the unity cmake file changed quite a lot since I last touched it :) any idea about this build failure: https://launchpadlibrarian.net/324564067/buildlog_ubuntu-artful-amd64.unity_7.5.0+17.10.20170619-0ubuntu1~ppa1_BUILDING.txt.gz ? [10:00] I don't see anything hurting my eyes in the output [10:00] (it failed on all arhcs) [10:00] archs* [10:09] didrocks: i'm checking [10:09] this is in artful? [10:09] That's weird. I think I just resolved the #1 bluez bug as Invalid [10:09] * duflu runs away before proven otherwise [10:11] andyrock: yes, artful [10:13] didrocks: are you sure it's not just a random failures [10:14] andyrock: it failed on all archs [10:14] https://launchpad.net/~ubuntu-desktop/+archive/ubuntu/transitions/+packages [10:14] that would be quite a concidence, dont' you think? [10:14] don't* [10:14] I can try a rerun on like amd64 if you want [10:16] didrocks, it's weird, we got an upload which built in artful 10 days ago [10:17] neither compiz nor cmake changed since [10:17] yeah, that's why I'm puzzled [10:18] the diff is just metadata, so not related to the changes themselves [10:18] failed again: https://launchpadlibrarian.net/324566872/buildlog_ubuntu-artful-amd64.unity_7.5.0+17.10.20170619-0ubuntu1~ppa1_BUILDING.txt.gz [10:18] seb128: this is why I'm asking andyrock :) [10:18] did we push a new cmake or something like that? [10:19] oh, version in proposed is != from version in release [10:20] (for cmake) [10:20] let me try to enable proposed in the ppa [10:22] but that's quite old [10:24] ok, build starts with -proposed enabled in the ppa [10:24] tjaalton, hey, is libinput 1.7 an upload you are looking at? [10:24] didrocks: did you push to trunk directly? [10:24] * didrocks gives back on all archs [10:24] andyrock: yes, it's in trunk, but not released (as it's in the ppa) [10:25] ok, remaining archs building in https://launchpad.net/~ubuntu-desktop/+archive/ubuntu/transitions/ [10:26] didrocks, seems to work now? [10:27] seb128: see backlog ^ [10:27] didrocks, ? [10:27] I saw the backlog [10:27] maybe was unclear "build starts" was opposed to "fails at configure" [10:27] I was just saying that the build-with-proposed seems to go past the previous issue [10:27] oh ok [10:27] :) [10:27] I though you were saying that launchpad was picking up builds [10:28] yeah, got it after your "?" :) [10:28] I bet it's this cmake update blocked in proposed [10:30] didrocks: but why it did not fail one week ago? [10:30] is bileto building with proposed? [10:30] andyrock: if it didn't change since the train, yeah, it's supposed to mirror the archive config by building with proposed [10:40] the build log at least shows that unity was built using the cmake version in proposed [10:47] hrm ... since the latest firefox upgrade my FF hangs pretty frequently [10:47] (on 16.04 that is) [11:14] seb128: yes, I'll update it soon [11:15] tjaalton, great, thanks === pisi is now known as Guest85144 [12:48] jibel: Ubuntu's not going to participate in next week's Alpha 1 then? [13:04] any desktop dev... when this is britney approved, please publish it https://bileto.ubuntu.com/#/ticket/2822 [13:08] good morning Trevinho [13:08] hi seb128 [13:08] jbicha, no [13:10] jibel: what do you think of checkbox still being installed by default? it pulls in a lot of QML deps [13:11] maybe 50+ MB installed? [13:14] jbicha, it's used for certification I think, I'll ask them. [13:15] thanks [13:17] seb128: I've been saving up some questions for you :) [13:17] hey jbicha :-) [13:18] some people were complaining on the forums about gnome-shell suddently becoming untranslated (its first build after being promoted to main) [13:19] what do you think about doing an artful langage pack earlier this cycle because of that? [13:20] +1 for me [13:21] Do LP language coordinators have to manually approve the upstream translations? Or do we get them for free? [13:27] what do you mean by upstream translations? [13:27] the ones coming from the .po shipped with the source are autoimported [13:27] no approval needed [13:28] I feel like there's a lot I still don't understand about LP translations so that's why I was asking :) [13:28] desktoppers - duplicate posts to the desktop mailing list are being looked at. I think I know where they are coming from, but not why yet. If I dont get a response I'll block the email address and see if that helps. [13:29] Should ~ubuntu-core-dev be part of ~indicator-applet-developers and ~unity-team ? [13:33] jbicha, that would make sense I guess? unsure if that hadn't been done for a reason, like email spam or something [13:34] probably the upstream teams wanted to review changes going in [13:34] but that's not so relevant now :( [13:34] yeah, IIRC, that was the discussions [13:34] separations of concerns [13:34] but that was before debian/ dir was in upstream tree [13:35] I emailed charles late Friday about ~indicator-applet-developers [13:36] core-dev has been invited [13:36] https://launchpad.net/~indicator-applet-developers according to this page [13:36] someone on the DMB could accept that [13:37] sil2100: mind doing it ^ [13:37] Looking [13:37] ♥ [13:39] Laney, didrocks: done! [13:40] thanks sil2100 :) [13:40] yw ;) [13:40] didrocks: are you going to have unity-session as a binary package from the Unity source? [13:40] I wonder if the pspmteam for product strategy is still a 'thing'? [13:41] sil2100, probably not, but not sure [13:41] jbicha: no, it's going to be from gnome-session package for now (as it was there before and all our patches are there) [13:41] didrocks: I'd rather that not be in maintained inside gnome-session [13:41] we can see and move it to unity source later on [13:41] but for now, let's keep it simple, there are a lot of moving pieces already [13:41] but yeah, nothing against moving it later on :) [13:43] I think it makes gnome-session *more* complicated to patch it instead of just shipping plain session files in the right package, but maybe later :) [13:44] it doesn't: look at the diff, but yeah, later will makes sense and it's easy enough once all the transition is done [13:45] seb128: can you look into the desktop version tracker? it stopped updating a few weeks ago [13:45] sil2100: sorry again, I invited ~ubuntu-core-dev to ~unity-team as well. [13:45] didrocks: on it! [13:45] thx man :) [13:46] didrocks: done again ;) Feel free to poke me if any other places need that, it's just a button press so I can do that anytime [13:46] * Laney breaks all things all the time [13:47] sil2100: there probably will be, thanks! :) [13:47] Laney: like what? [13:47] hmm, let me think [13:47] not enough easter eggs in unity [13:48] don't know if G-S has any btw, if there are less than unity, it's a critical bug! [13:49] jbicha, let me have a look to versions [13:51] didrocks: I think we're missing a dependency… https://bugs.archlinux.org/task/28957 [13:51] I'll let you handle the MIR ;) [14:01] jbicha: ahah, I'll surely autoapprove all the things for the easter egg :p [14:01] especially for wanda [14:21] I just got a weird sense that I was looking down a hole into my computer [14:22] something to do with the black top bar and the colours of the website I was on [14:22] ...or a brain injury... [14:24] sounds like an interesting website! [14:25] bugzilla.gnome.org! [14:25] actually that is quite hole-like [14:26] haha === ricotz_ is now known as ricotz [14:59] Laney: could you publish https://bileto.ubuntu.com/#/ticket/2822 ? [14:59] sure [15:25] ta === pstolowski is now known as pstolowski|erran [15:41] So, we at System76 are looking at using GNOME Initial Setup instead of Ubiquity OEM mode for OEM installs [15:41] We feel it has better integration with the rest of the GNOME experience [15:42] I am implementing encrypted home support in accountsservice, and I would like to allow the password strength to be overridden by the user, if desired [15:42] How does this sound? [15:45] jackpot51: hi - to make sure that I understand, you're wanting to allow the user to enter a weak password? [15:46] tyhicks: I think ubiquity already does that? [15:46] jbicha: I'm not certain about that [15:46] let me test today's daily and check :) [15:46] jbicha, the fact that ubiquity does it or not doesn't change the question? [15:47] I think that gnome-initial-setup's password check (should be the same as gnome-control-center's) is stricter than ubiquity's at least [15:47] Anaconda lets you click the next button twice [15:47] ubiquity hints about the password strength but doesn't prevent you using a weak one [15:47] and it would be nice for them to match anyway [15:47] Yeah, gnome is super duper strict [15:47] g-c-c/u-c-c do prevent you using a weak one [15:48] I remember that, I had to use passwd for a temporary user [15:48] We often set up lab machines with really simple passwords, so it is a nuisance [15:48] jackpot51, do you plan to work with us on that or just do it as a system76 specific change? [15:48] that = using gnome-initial-setup [15:48] Absolutely, I plan to work with you. I am building a patch for accountsservice that will add encrypted home [15:49] great [15:49] do you have user testing of both? [15:49] I made ubiquity skip the account setup by removing the ubi-usersetup plugin, then it booted straight into gnome initial setup [15:49] for context, I have some instructions to try out Initial Setup's new user mode at LP: #1673453 [15:49] Launchpad bug 1673453 in Ubuntu GNOME "Use gnome-initial-setup to replace 2nd half of ubiquity" [Wishlist,New] https://launchpad.net/bugs/1673453 [15:49] and how do they compare feature wise? [15:50] (that was before the big GNOME announcement and I had not thought of ubiquity oem-mode when I opened that bug) [15:50] They are nearly feature identical [15:50] jackpot51: from a security standpoint, protecting the mount key with a weak passphrase is really a waste of CPU cycles [15:50] The only missing thing is encrypted home [15:51] jackpot51: the wrapped-passphrase file, which contains the mount passphrase that's then wrapped with the login passphrase, is the weak link in home dir encryption [15:51] Timezone, keyboard setup, network setup, and user account setup from ubiquity are handled by Initial Setup [15:51] tyhicks, I would expect users selecting encrypted home to use a strong password, but it would be up to them how to set up their computer [15:51] jackpot51: if an attacker can easily brute-force the login passphrase, then you're wasting your time doing encryption [15:52] jackpot51: I agree - however, the weak password warning should be very clear about this because most users will not understand the consequences [15:53] jackpot51, so you plan to strip down ubiquity ? [15:53] I think those plans should be discussed on ubuntu-devel] [15:53] @ [15:53] with rational [15:54] they are probably something we might to want to look at for Ubuntu itself, but having people driving the discussion with us would be useful otherwise it feels like people are just too busy and nobody is going to drive forward [15:54] tyhicks, I don't think you understand. I want to make two *independent* changes to Initial Setup: [15:54] 1 - allow the user to use any password they like. Pop up a dialog forcing confirmation for weaker passwords [15:54] 2 - allow the selection of encrypted home on the account setup page [15:54] tyhicks: The real problem with encrypted home and weak/no passwords is poor security, not CPU cycles [15:54] jackpot51: do you intend to push all Ubuntu users through Initial Setup's new user mode, or just for OEMs? [15:54] Which can be mitigated with a good warning about weak passwords [15:54] Yes, jbicha, I think both can go through Initial Setup [15:55] jackpot51: have you tested an ISO with ubiquity changed to do that yet? [15:55] isantop: yes, I know that. I was saying that it is a waste of time to encrypt your data with a key that's persistently stored with a weak wrapping passphrase. [15:55] Yeah, I have one I have been testing [15:56] tyhicks: Not more of a waste than using a strong password [15:57] isantop: I don't understand what you mean by that but we seem to be arguing for the same thing [15:57] jbicha: Basically, take the ubuntu gnome ISO. `rm /usr/lib/ubiquity/plugins/ubi-usersetup.py` in the squashfs [15:57] Then it will take you straight from the installer to initial setup on the first reboot [15:57] I think the same thing would work with the 17.10 ISO [15:58] I did not have as much success skipping the timezone or consolesetup plugins [15:58] oh that sounds easier than my method :) [15:58] But I am sure I could figure it out ;) [15:58] jackpot51: note that I'm looking into moving home directory encryption from eCryptfs over to use ext4 encryption but I'm still in the early stages of that [15:58] tyhicks: That is good to know. Will it still use `adduser --encrypt-home` ? [15:59] jackpot51: yes, adduser will be updated to do the right thing [15:59] jackpot51: are you solely relying on adduser to handle the setup of the encrypted home? [16:00] Yes, gnome-initial-setup uses libaccountsservice, which uses adduser internally to do encrypted home. So my patch will add a boolean to add --encrypt-home to the adduser call [16:01] I'm off for a bit, see you later [16:01] Here it is, so far: https://pastebin.com/PKVF5DSp [16:02] And since I like gist better (no ads): https://gist.github.com/jackpot51/8e24f48ec97d26f9f49d0750a92bd8bc [16:04] jackpot51: great, it should be a transparent change from that standpoint [16:05] Sweet === pstolowski|erran is now known as pstolowski [16:43] k, time for dinner&co, I might read some backlog and deal with some emails later on [16:43] have a nice evening desktopers [16:43] you too seb128 [16:44] jackpot51, in any case please start a discussion on ubuntu-devel@ on what changes you think would be improvements and why you think Ubuntu should follow that road [16:44] jackpot51, that would be the right way to get what you want and help Ubuntu at the same time :-) [16:44] kenvandine, thanks === Guest24249 is now known as ahayzen === ahayzen is now known as Guest14311 === Guest14311 is now known as ahayzen [17:04] sarnold, sorry to bug you, any update on the gdm MIR? [17:07] kenvandine: no progress, last week was crazy. sorry. [17:08] sarnold, understood, i'll pester you again in a couple days :) [17:15] so its GDM then? [17:15] set in stone? [17:15] kenvandine: thanks! :D [17:15] immu, yup [17:37] night all [17:41] oke then, how is the distro shapping up and how can i move from 17.04 to daily builds [17:45] immu: 17.10 is still in early Alpha so you may want to try in a VM? [17:46] ok [17:46] daily builds are at http://cdimage.ubuntu.com/daily-live/current/ [17:47] and you can upgrade from an older release with update-manager -d [17:48] I use artful on my main machine, with gnome shell and X and it's quite stable. [17:49] but you are french :) [17:50] ogra, do you mean our definition of stability is less strict than the german one? :) [17:51] dunno, but you seem at least more brave than everyone else (seeing the election) [17:51] ;) [17:51] indeed :) [17:52] ogra, unlike you we don't keep a chancelor for 12 years ;) [17:52] well, we lack a proper replacement [17:53] at least we did ... schulz coould actually be a candidate for the first time in ages (though i dont think he will make it) [17:53] ogra, next target for Macron, chancelor of Europe ;P ... in a year [17:53] LOL === CrazyMelon is now known as CrazyLemon === Beret- is now known as Beret === meetingology` is now known as meetingology [19:23] Please look into https://bugs.launchpad.net/ubuntu/+source/indicator-datetime/+bug/1698638 [19:24] Ubuntu bug 1698638 in indicator-datetime (Ubuntu) "indicator-datetime list evolution calendar full-day event a day before actual date of event (Unity-7)" [Undecided,Confirmed] === TheMuso` is now known as TheMuso === JanC_ is now known as JanC