[00:49] <jbicha> robert_ancell: good morning
[00:50] <jbicha> I tried running gnome-software master; it fixes some Ubuntu issues but has new ones
[00:50] <jbicha> it's good though, I was getting bored of the old bugs ;)
[01:46] <robert_ancell> jbicha, I'm trying to merge with 3.20 right now...
[01:47] <jbicha> would 3.22 be any easier?
[01:48] <jbicha> are you planning to use the packagekit plugin now?
[01:50] <robert_ancell> jbicha, this is for SRuing
[01:50] <duflu> robert_ancell: Is it a known fixable issue that plymouth starts too late, and/or ends too soon? I suspect for many of us most of the boot time is pre-plymouth... so you just never see it any more
[01:52] <robert_ancell> duflu, I'm not sure if anyone is managing Plymouth (it used to be foundations, I suspect it's pretty much unmaintained). In terms of LightDM, we stop it when we're ready to start. So anything else is managed in Plymouth/systemd
[01:52]  * duflu suspects we need to display something nicer from grub till plymouth starts
[01:53] <robert_ancell> duflu, we just need u-s-c :)
[01:53] <duflu> robert_ancell: Getting working graphics in to initrd is the thing :)
[01:53] <jbicha> yay! Ubuntu Software Center ;)
[01:54] <robert_ancell> yep, that's the hard part
[01:56] <duflu> robert_ancell: The login screen is not Xorg right?
[01:56] <duflu> It's DRI?
[01:56] <robert_ancell> duflu, yeah, it's something direct like that. Definitely not Xorg
[01:56] <robert_ancell> Plymouth is pluggable
[01:57] <duflu> OK then. This bootchart is wrong
[02:03] <duflu> robert_ancell: So we should never see Xorg starting at all if you log in to Unity8 first right?
[02:04] <robert_ancell> duflu, unless you have unity-greeter running
[02:04] <duflu> robert_ancell: Oh, yes that's it.
[02:04] <duflu> robert_ancell: Ummkay, so how do I make an Ubuntu system not start the unity-greeter?
[02:05] <robert_ancell> duflu, set autologin
[02:05] <duflu> OK, thanks
[02:06] <jbicha> robert_ancell: you should try appstream-glib 0.6.1, it magically makes gnome-software support plugins
[02:10] <jbicha> although the Installed tab doesn't handle them right (only 1 plugin from each package is shown)
[02:23] <duflu> robert_ancell: lightdm supports Mir system compositors too doesn't it?
[02:23] <robert_ancell> duflu, yes
[02:23] <duflu> OK. We're just prioritizing visual quality (changeovers) for Unity7 for now
[02:24] <duflu> robert_ancell: Oh. I mean running on Mir, not starting Mir
[02:24] <robert_ancell> duflu, running what? sessions?
[02:25] <duflu> robert_ancell: Displaying a greeter... that's not Xorg based
[02:25] <robert_ancell> duflu, yes
[03:11] <robert_ancell> attente, around?
[03:13] <attente> robert_ancell: hey
[03:13] <robert_ancell> attente, I'm trying to merge GNOME Software, but ca63436b045f3928538c5d83b745022498f7d641 is causing difficulty
[03:14] <robert_ancell> In particular the change in src/plugins/gs-plugin-appstream.c where it does:
[03:14] <robert_ancell> +               plugin->priv->done_init = FALSE;
[03:14] <robert_ancell> +               gs_plugin_appstream_startup (plugin, NULL);
[03:14] <robert_ancell> They no longer exist so not sure what to replace them with
[03:16] <attente> robert_ancell: is this on master? i think there are two other flags that kind of replace this functionality. one is '_METADATA' and the other is '_PAYLOAD'
[03:17] <robert_ancell> attente, no, this is trying to merge gnome-3-20 into wip/ubuntu-xenial
[03:17] <robert_ancell> but yeah, master has entirely different code so I'm not sure what to do
[03:22] <attente> robert_ancell: part of me thinks that code should just call gs_plugin_setup () directly, but there might be some teardown involved
[04:48] <robert_ancell> attente, what does GS_PLUGIN_REFRESH_FLAGS_UI actually indicate?
[04:58] <attente> robert_ancell: it's there to tell the appstream plugin to re-initialize the AsStore object so that the directories associated with the appstream data can be watched again
[04:58] <attente> robert_ancell: related to https://bugs.launchpad.net/ubuntu/+source/gnome-software/+bug/1554023
[04:59] <robert_ancell> attente, so is it the same as GS_PLUGIN_REFRESH_FLAGS_METADATA?
[05:00] <robert_ancell> Because there was GS_PLUGIN_REFRESH_FLAGS_UPDATES which is the same as GS_PLUGIN_REFRESH_FLAGS_METADATA I think.
[05:00] <robert_ancell> But in _UI was treated as something different
[05:01] <attente> robert_ancell: i don't know, it sounds like they serve the same purpose, but i didn't confirm it
[05:05] <attente> robert_ancell: from what i remember, the original reason for adding that in the first place is because the AsStore object is monitoring some set of directories by adding inotify watches. but the problem is the directories created by the appstream refresh don't exist so the AsStore can't watch for them, and when those directories are created eventually,
[05:05] <attente> there's nothing to tell it that they were created
[05:05] <robert_ancell> attente, ok, thanks
[05:06] <attente> so that flag is basically the second step of a two-stage update
[05:06] <attente> where the second stage is to re-create the AsStore so that it can start adding watches on the new directories that were created by the appstream refresh
[07:25] <Trevinho> Morning... Although it's national holiday today, so... Have a nice day! :-P
[08:00] <Laney> HEY!
[08:02] <xnox> after a cloudy weekend, facebook is trolling me:
[08:02] <willcooke> morning Laney, all
[08:02] <xnox> "Clouds are cleaning up, enjoy sunshine today!" - On a Monday morning.....
[08:05] <Laney> xnox: was nice up in yorkshire :-)
[08:05]  * Laney got a bit sunburned
[09:14] <Laney> brb
[10:33] <Laney> vlc is back in appstream after a re-run for those who care(d)
[10:33]  * Laney stabs that problem
[10:34] <Laney> coffee or tea?
[10:35] <RAOF> (time < 10am).and_then(|| { tea() });
[10:39]  * xnox russian airlines: "tea or coffee?"; "coffee"; "you guessed wrong, this is tea"
[10:39]  * Laney learned some rust today
[10:45] <willcooke> thx Laney
[10:54] <RAOF> Laney: You look a likely victim - how do we get new CI'd packages into the archive these days?
[10:55] <Laney> RAOF: Can't the train just dump it into NEW?
[10:56] <RAOF> If that's the process, then I guess?
[10:56] <Laney> I don't know of a special process
[10:56] <Laney> there's some bug in Launchpad that package skip binNEW I think
[10:57] <Laney> but AFAIK sourceNEW is just fine
[10:57] <Laney> s/package/packages/
[11:02] <seb128> the citrain team usually requires for preNEW reviews before landing though, unsure why
[11:03] <seb128> but yeah, what L_aney said otherwise
[11:03] <seb128> source NEW should end up in the queue
[11:03] <willcooke> hey seb128 :)
[11:04] <seb128> hey willcooke ;-) (I've been around for some hours, just quiet on IRC)
[11:04] <seb128> had a good w.e?
[11:04] <willcooke> yes thanks
[11:04] <willcooke> you?
[11:06] <seb128> quite good, mostly relaxing which is nice ;-)
[11:08] <willcooke> :)
[12:25] <flexiondotorg> Laney, I was stuck down with a virus last week.
[12:26] <Laney> hey flexiondotorg
[12:26] <flexiondotorg> So the rush is on for GTK 3.20 in Ubuntu MATE for Beta on Thurs.
[12:26] <Laney> sorry to hear that, hope that you managed to get some sufficient antivirus software installed
[12:26] <jbicha> flexiondotorg: Beta isn't until next week
[12:27] <flexiondotorg> Laney, I cat'd EICAR.COM.TXT through my system this morning. All checks passed.
[12:28] <flexiondotorg> I must be recovering, right? ;-)
[12:29] <flexiondotorg> jbicha, Yes. But freeze is Thursday. I'd like to get "everything" (or close to it) done by then.
[12:52] <bregma> seb128, willcooke, have all the unity8-MIR tasks been assigned?  I have a little spare bandwidth to fill in holes if not....
[13:08] <jbicha> mterry: good morning, I can upload gedit now?
[13:12] <seb128> bregma, check on the trello board, we added comments to the cards where a maintainer got emailed (and maybe ask the most active contributor before starting just in case)
[13:12] <seb128> there are probably a few that are still unclaimed
[13:12] <seb128> especially in the recommends land
[13:12] <seb128> we worked from the list without recommends to start
[13:12] <seb128> since that's what is needed to install
[13:13] <bregma> I still don;t understand how you can generate an ISO without recommends
[13:14] <mterry> jbicha: I just approved gspell, so probably?  Was there anything else waiting on MIR?
[13:15] <mterry> jbicha: and good morning  :)
[13:15] <seb128> recommends are not depends
[13:15] <seb128> if things are not in main they don't get installed and are missing on the iso
[13:15] <seb128> but no error
[13:18] <jcastro> Sweet5hark: I'm using rev4 of your LO 5.2.0.4 snap, but I appear to be missing .desktop files, CLI launches just fine, known bug or should I report it?
[13:25] <jbicha> mterry: thanks, I believe the only other GNOME-related MIR is bug 1547395 which is blocking totem-pl-parser (auto-synced from Debian)
[13:27] <mterry> jbicha: ah yes.  doko assigned that to the ubuntu-desktop team; I think he wanted you (welcome back on the team :)) to fill out the MIR boilerplate
[13:28] <seb128> jbicha, can that one be avoided? seems like quite a stack to maintain there
[13:29] <jbicha> I don't know much about it
[13:53] <willcooke> jcastro, known bug - there's no easy way to differentiate in the dash so users wouldn't know which version they're running.  Sweet5hark is thinking of ways to work around that - like maybe a different icon or something
[13:54] <willcooke> bregma, seb128 - no harm is filing MIR bugs for anything though right?
[13:55] <bregma> don't want to overload the MIR team
[13:55] <willcooke> bregma, would you also consider joining the MIR team?
[13:55] <willcooke> bregma, yeah, fair
[13:56] <bregma> I would consider it, but I assume it requres being a core dev
[13:57]  * willcooke hand-waves 
[13:57] <willcooke> probably
[13:57] <willcooke> I've got a couple of other folk who are being press-ganged in to it, so if we get stuck I'll let you know.
[13:58] <seb128> willcooke, that sounds like a problem that shouldn't be solved on Sweet5hark's side but on the unity side
[13:59] <seb128> willcooke, right, not harm, but again priority matters, if we dump optional and mandatory ones they might review in the wrong order and delay
[13:59] <willcooke> seb128, I wonder if we could borrow the idea from g-s and have the badge icon overlay?
[14:00] <qengho> Hi hi!
[14:00] <willcooke> morning qengho
[14:00] <seb128> hey qengho
[14:02] <Sweet5hark> seb128, jcastro, willcooke: yes, adding desktop files is still an issue -- as is having dpkg and snap installed in parallel. .desktop files arent hard per se from the libreoffice side. harder: how to do that with parallel installs.
[14:02] <jcastro> I didn't really see a need to keep the old installation around so I removed it
[14:03] <qengho> cking: have you heard of or had trouble with zfs-initramfs being un-installed in X or Y? I am quite sure I didn't touch it, but it was gone from my machine just some time over the summer, and this first reboot was a disaster.
[14:03] <Sweet5hark> FWIW, I thought about having an libreoffice-snap command in addition to plain libreoffice, to prevent the need for running /snap/bin/libreoffice
[14:03] <qengho> cking: Still in archive, of course. I'm looking through APT logs.
[14:03] <cking> qengho, nope, that's new to me
[14:04] <Sweet5hark> jcastro: there will be more then enough, trust me ;)
[14:05] <seb128> Sweet5hark, well if you install 2 versions of libreoffice it's normal you get 2 icons
[14:05] <seb128> if they have the same name and icon it's a bit unfortunate
[14:06] <seb128> but you could decide to tweak the name and/or icon
[14:08] <qengho> cking: Looks like some attempt to install zfs-zed, and then "apt -f install". Hard to read what happened. :\ OH well.
[14:08] <cking> hrm
[14:09] <qengho> I should pay more attention.
[14:09] <jcastro> Sweet5hark: well ideally you'd want us enthusiasts banging on the snap, not the old 5.0.x version xenial comes with right?
[14:09] <qengho> Some zfs-doc version dependency problem, shared between them.
[14:10] <qengho> cking: It might be my problem, but perhaps that zfs-doc Dep could be looser.
[14:14] <Sweet5hark> jcastro: sure. enthusiasts mostly know what they are doing. the non-enthusiasts who dont know what they are running accidentally running into things like bug 1612226 is much more eek.
[14:18] <seb128> Sweet5hark, well it's a bug, you have bugs in deb updates as well
[14:20] <Sweet5hark> seb128: well, on dpkg for a software the size of LibreOffice it takes about a week with 50.000 downloads from the ppa to be reasonably sure its not doing something terrible. We dont have that yet for snaps.
[14:23] <Sweet5hark> sure the solution is get more folks on snaps. But its a balance: If your browser or media player crashes you restart, reopen the file/stream and on you go. OTOH, people actually have data and state in LibreOffice documents.
[14:23] <seb128> well, the format shouldn't matter and upstream is shipping it so they should have some confidence
[14:23] <seb128> or libreoffice upstream is acting reckless with their users
[14:34] <seb128> happyaron, how is that test deb for the 802.1x bug going?
[14:34] <seb128> happyaron, oh, and hey ;-)
[14:49] <seb128> TheMuso, hey, did you see that bug #1582213 and bug #1574324 got the info your asked for?
[14:51] <Sweet5hark> meh, a local rebuild of libreoffice against yakkety-proposed .... succeeded.
[14:51] <Sweet5hark> To be, or not to be, that is the question: Whether 'tis nobler in the mind to ignore the slings and arrows of a hiding Heisenbug, or to take arms against a sea of troubles and by opposing end them.
[14:59] <jbicha> seb128: the new evolution wants webkitgtk 2.13 :( https://git.gnome.org/browse/evolution/commit/configure.ac?id=332789ff
[14:59] <seb128> :-/
[14:59] <seb128> is that really a requirement or did they just pick the current version?
[15:00] <jbicha> let's find out…
[15:04] <mhall119> willcooke: ping, have you found any solution for live-session or vm testing unity 8?
[15:04] <mhall119> jouni was trying to put together something last week too
[15:05] <willcooke> mhall119, not yet, won't get looked at this cycle
[15:05] <willcooke> beside, once u8 lands in the iso that is probably a solution
[15:09] <mhall119> any estimate on when it will get on the ISO? Is it still going through MIR (confusion ensues) process?
[15:12] <willcooke> mhall119, yeah we're starting the MIRs, so hard to say atm
[15:14] <davmor2> mhall119: unity8 will run in kvm with qxl gfx using unity8-desktop-session-mir
[15:15] <mhall119> I thought the -mir package was deprecated
[15:16] <davmor2> mhall119: what other way is there to install it
[15:16] <mterry> jbicha: for bug 1547395...  is that still needed?  libquvi-scripts doesn't seem to depend on anything, and I don't see those packages in the component mismatch svg
[15:17] <davmor2> mhall119: -mir is turned into a transitional package for unity8-desktop-session
[15:17] <davmor2> mhall119: so they both install the same thing
[15:18] <seb128> hum
[15:18] <seb128> is "get your position" working for anyone on https://www.openstreetmap.org ?
[15:18] <seb128> I get errors in firefox and chromium on xenial
[15:19] <jbicha> mterry: it's on update_excuses.html totem-pl-parser > libquvi > libquvi-scripts > lua-bitop,-expat,-json,-socket
[15:20] <mterry> jbicha: oh... huh...  I don't know why apt doesn't see that proposed version for me.  I'm on yakkety-proposed...
[15:20] <mterry> jbicha: but ok, thanks
[15:21] <jbicha> mterry: living dangerously?
[15:21] <mterry> jbicha: can't compile unity8 trunk without it for now
[15:21] <Laney> seb128: doesn't work for me either
[15:21] <Laney> something about a 403 error from google
[15:21] <mterry> jbicha: and I don't believe in chroots  :-P
[15:21] <seb128> Laney, thanks
[15:21] <jbicha> lol, ok
[15:22] <seb128> http://html5demos.com/geo doesn't work either
[15:22] <seb128> but chromium complains about it not being https which seems enough for it to not do geoloc
[15:22] <seb128> firefox just returns an error
[15:23]  * seb128 was trying to look if bug #1611037 is a real regression but I don't play with those often enough to say if that was working before
[15:23] <seb128> qengho, ^ do you know?
[15:23] <Laney> pretty sure I used that before
[15:23] <qengho> seb128: Yes! The Google API allocation for geolocation is pegged. I opened a ticket with GOOG.
[15:24] <seb128> qengho, can you comment on the bug ^ with maybe the reference to your ticket?
[15:24] <seb128> qengho, Laney, thanks
[15:30] <seb128> qengho, I'm not familiar with that but you mentioned a quota on your comment, who is being given a quota? is that a webbrowse thing? website one?
[15:33] <qengho> seb128: All of Ubuntu as an organization has a gratis API key that Google gave us, to give our browsers access to Google API. Firefox and Chromium shares a "secret" key, and all of our users together get N API calls per month. We just hit the natural default ceiling that they set when GOOG gave us the key.
[15:33] <seb128> ah ok
[15:34] <seb128> do we have stats of the websites our user hit the api through?
[15:34] <seb128> also is that a monthly quota? when do we hit the ceiling?
[15:34] <qengho> seb128: No, no web-site specifics. Only a running sum over each month.
[15:34] <seb128> hum, k, annoying
[15:35] <seb128> qengho, thanks
[15:35] <qengho> seb128: I think July was the first time we hit the limit, at day 26 or something.
[15:35] <seb128> well that bug was opened a week ago
[15:35] <seb128> so august already hit the limit?
[15:36] <seb128> chrisccoulson, oh, hey, thanks for the email reply ;-)
[15:37] <chrisccoulson> HI seb128
[15:39] <qengho> seb128: I mis-spoke. It's N per day. And July 26 was the first day we filled our quota within the day.
[15:40] <seb128> qengho, ah ok, well the day is not over and it seems we already hit our quota ... let me try again tomorrow morning ;-)
[15:41] <chrisccoulson> seb128, yeah, we're over the quota for today already
[15:42] <seb128> it's annoying that we don't get stats/details about the usage
[15:42] <seb128> oh well, let's hope they raise the quota then
[15:43] <chrisccoulson> And the quota is reset at midnight pacific time. So it's not even been 9 hours
[15:43] <Laney> for
[15:43] <Laney> oops
[15:43] <seb128> urg
[15:44] <chrisccoulson> Right, I need to get back to packing :)
[15:44] <qengho> chrisccoulson: Some one might have "borrowed" our API key for their own project instead of paying for their own.
[15:45] <chrisccoulson> I'm in France for a few days. Well, not proper France. Disneyland. The kids are waaaay more excited than I am
[15:45] <chrisccoulson> qengho, I wish we could keep these secret
[15:45] <qengho> The Ennui-est Place on Earth!
[15:46] <seb128> lol
[15:46] <seb128> chrisccoulson, enjoy "France" ;-)
[15:46] <chrisccoulson> qengho, the Mozilla folk keep their key on their build infrastructure, and have the build load it off disk so that it's never exposed in a public repo
[15:46] <chrisccoulson> seb128, thanks ;)
[15:49] <seb128> qengho, that someone would be quite popular if that's the issue
[16:07] <xclaesse> seb128, fwiw, I'm admin of gnome's facebook API key, and I regularily get limit alerts
[16:07] <xclaesse> seb128, checking the history, there are some really huge peaks, I'm surpecting someone does something nasty
[16:07] <xclaesse> some DoS on the key maybe, dunno...
[16:08] <xclaesse> maybe ubuntu hits the same attack ?
[16:14] <seb128> xclaesse, hey, could be...
[16:17] <seb128> difficult to say/know exactly what's going on in those cases
[16:38] <seb128> Sweet5hark, could you look at bug #1577093 if it's true/a known issue?
[16:38] <seb128> it claims that libreoffice struggle on big documents when -gtk is installed
[16:38] <seb128> not in older ubuntu series nor without gtk though
[17:26] <seb128> have a good evening desktopers!
[17:26] <willcooke> cya seb128
[18:03] <willcooke> night all