[05:20] hello :) [07:35] Good morning [07:36] infinity: FWIW, I've always asked for sending locale patches upstream, and many of them should already have Bug: links [07:37] good morning pitti [07:39] bonjour didrocks, ça va ? [07:40] pitti: bon week-end, oui, et toi ? [07:41] didrocks: le week-end était magnifique -- Annet et moi étaions à Londres, pour le concert d'Adele [07:41] et un peu de tourisme :) [07:42] ah, d'où le rond point avec les feux rouges :) [07:42] oui :) [07:42] comme un arbre de Noël [07:43] héhé === Drac0 is now known as Guest96193 [07:55] moin! [08:55] morning desktoppers [09:02] morning! [09:03] what up Laney [09:05] willcooke: so good news bad news, cyphermox is making progress good news, bad news looks like it is most of the unity graphics stack that takes out the system on upgrade :( [09:05] davmor2, yeah I asked hikiko to take a deeper look in to that. [09:05] hikiko, did you find anything out yet? [09:06] davmor2, I did some rudimentary debugging with Compiz on Friday and it looks like the X server is going away [09:06] but needs hikiko to dive in to it in more dept [09:06] h [09:07] willcooke, you mean the crash during the upgrade? [09:08] hikiko, yeah [09:08] I couldn't reproduce it could you tell me what you did exactly and in which system? [09:08] hrm. It's very reproducible here. Did you see my email, all the steps are laid out there [09:09] Basically: install 14.04. upgrade to 16.04. Crash. [09:09] sorry I didn't see this one ok looking at it now [09:11] hi willcooke, how's it going? [09:12] oh I saw that I have marked the email as important but I totally forgot sorry :/ [09:12] thanks hikiko, this is a priority for now, even more than e zoom. Until we know what's going on at least. [09:12] ok [09:15] willcooke, do you experience this bug in vbox only> [09:15] ? [09:16] hikiko, no, davmor2 has it on real hardware as well: https://bugs.launchpad.net/ubuntu/+source/ubuntu-release-upgrader/+bug/1555237 [09:16] Launchpad bug 1555237 in ubuntu-release-upgrader (Ubuntu) "Upgrade from 14.04.4→ 16.04 dies midway taking out the session." [Critical,In progress] [09:17] mm ok, because here they say that it was a low memory issue: https://bbs.archlinux.org/viewtopic.php?id=183376 [09:17] (similar bug) [09:18] and I was wondering if it's because the vm graphics memory is very low and we use it all during the upgrade [09:18] ok I'll look further [09:18] ah, interesting [09:18] yeah, please do some more investigation [09:19] I'll also boost the RAM on my VM [09:19] hikiko: my vm has 2Gb of memory that should be enough for an upgrade and the hdd is 20Gb, and hardware wise it is hybrid intel nvidia prime, 8Gb ram and I can't remember how much memory the gfx card has [09:20] davmor2 I mean the graphics memory [09:20] which is 128 in vbox [09:20] oh you said you have 8GB [09:21] ok ok I'll look at it, I just wanted to make sure it's not something that simple :) (too good to be true) [09:25] hikiko: vm has 64Mb, looking on hardware now [09:26] hikiko: 256Mb on hardware [09:27] davmor2, you tried the virtualbox too? [09:27] I mean [09:27] you only tested this in vbox? [09:27] or in real hardware too? [09:27] hikiko: both and I use kvm not vbox [09:27] good day everyone [09:28] alright, I am downloading the 14.04 [09:28] hikiko: I check daily on vm that it is still happening and then once a week on spare hardware to ensure it is still happening there too [09:30] hikiko: I can run another test on hardware in a bit if you want and open up and ssh port to to it if that helps [09:31] I want to check if: 1- if it's possible that the program that runs during the upgrade has a bug that consumes the memory, 2- if something else tries to make use of the :0 while compiz is running [09:32] davmor2, I dont know if that helps I'll set up it here in a few minutes [09:32] but thanks :) [09:34] Trevinho: No Unity upload I see [09:37] Laney: it's in https://requests.ci-train.ubuntu.com/#/ticket/1146 but I noticed a small issue, that I've fixed, but I was waiting andyrock to check it. Then we can go. By today. [09:38] k [09:39] hi Trevinho and hikiko [09:39] Hi, Laney, yeah 😃 === rhuddie_ is now known as rhuddie [10:13] wormery just arrived \o/ [10:14] Laney: does that have anything to do with worms? [10:14] hey pitti! [10:14] hey Laney, how are you? [10:14] yeah, food in, compost out [10:15] via worm digestion [10:15] pitti: great, thanks! [10:15] and you? [10:16] I saw your pictures, looked like a fun weekend [10:16] Laney: wow; is this available for the other way too? [10:16] Laney: indeed -- the concert was absolutely amazing [10:17] pitti: you mean compost -> food? [10:17] and we managed to visit Greenwich Observatory/Prime Meridian, the aquarium, Camden Town, the Regent canal (I think that one), a walk through Westminster, and some nice cafés/pubs too [10:17] Laney: yeah :) [10:17] that's the circle of life :P [10:17] oh wow [10:17] packed weekend [10:17] Laney: well, we started Friday morning already [10:18] so Greenwich and the concert was Fri afternoon/evening [10:19] We visited the Observatory in the summer last year [10:20] quite a nice park it's in too [10:20] and there was a market nearby that we visited, bit more crafty than the Borough one [10:20] * Laney blinks [10:20] someone just retried glib/s390x at literally the same time as me [10:42] Laney: race conditions ftw! [10:43] more impressed that someone else was looking at this :) [10:43] although... [10:43] ...has it hung again? [11:03] ffs [11:03] works on the autopkgtest machine of course [11:09] Laney: Saw your comment at bug #1559070. The plan is to wait with uploading ubuntu-docs again until the tranlators are done. Is there a reason to do it before? [11:09] bug 1559070 in Yelp "SVG icons not shown / "yelp " fails" [High,Confirmed] https://launchpad.net/bugs/1559070 [11:10] GunnarHj: So that it stops being broken [11:10] but, up to you [11:10] GunnarHj: willcooke is preparing some new text/screenshots for the slideshow atm FYI [11:10] don't think there has been UIFe for that yet [11:11] not yet - just got the OK from design [11:11] still need to finish doing the screenshots [11:11] there is a very small change to the text on one slide [11:12] the one about Ubuntu Software Centre [11:12] Laney: Then I'll wait with u-c. [11:13] 'k [11:13] willcooke: and accompanying screenshot which is still USC right [11:14] davmor2, yeah, changing the screenshot, plus a few others [11:14] actually, I'll just redo them all [11:14] so the wallpaper matches [11:14] willcooke, Laney: Suppose we should notify the translators via the mailing list when the text string is available in Rosetta. The screen shots is not a problem, since we don't include screenshots from the installer in the docs. [11:15] GunnarHj, sure thing - I will email the translation team once I have an MP up [11:15] willcooke: Great. [11:34] * Sweet5hark never uploaded a libreoffice himself, but just triggered/unblocked the migration from proposed for the first time. [11:36] THE POWER [11:38] Laney: https://www.youtube.com/watch?v=_BRv9wGf5pk === Drac0 is now known as Guest71064 [12:08] Sweet5hark: I was hoping it would be that [12:08] hey that band is German [12:08] who knew? [12:38] Laney: most of that embarrassing 1990ies eurodance trash is german ... [12:55] happy monday [12:56] DDDDDDDDDDDDDDDDDEEEEEEEEEEEEEEEEEEEEEEESSSSSSSSSSSSSSSSSSSSSSSSSSRRRRRRRRRRRRRRRRRRRRRRRRRRRTTTTTTTTTTTTTTTTTTTTTTTTT [12:59] what did i do now? :( [12:59] it was a hello [12:59] er.. i mean [12:59] hi laney :) [12:59] imagine me running towards you [12:59] for a glomp [13:00] is a glomp something good? [13:00] it sounds like it might be painful [13:00] in a good way [13:00] okay. good :) [13:00] glomp away [13:00] how's it going? [13:00] okay [13:01] chatting security 'bugs' again === alan_g is now known as alan_g|lunch [13:01] fun [13:01] and it makes me wonder [13:01] why can't we all just be nice to each other? [13:01] why all of this "security" anyway? [13:01] not that I was prepping you [13:01] ...but... [13:01] https://launchpad.net/ubuntu/+source/glib2.0/2.47.92-1/+build/9368499 <- this build keeps hanging [13:02] i mean... if i put a key on something, isn't it clear that i want you to stay out? why are we even talking about how to crack the key? [13:02] on https://git.gnome.org/browse/glib/tree/glib/tests/642026.c this test [13:02] and I can't make it happen on a test machineeeeeeeeeeeeeeeee whyyyyyyy [13:03] this case looks pretty simple [13:03] does x390 have an exotic memory architecture or something? [13:04] would expect to be able to reproduce it if so [13:04] * Laney checks kernel versions [13:05] a hang points to a problem with condition variables [13:05] which could very well be a problem with the kernel's futex implementation on x390 [13:05] our implementation of gcond is pretty straight forward [13:05] Linux laney 4.4.0-14-generic #30-Ubuntu SMP Tue Mar 15 13:01:11 UTC 2016 s390x s390x s390x GNU/Linux [13:05] could also be the mutexes of course... [13:05] this works [13:05] I think Laney is drunk with springtime. [13:05] Kernel version: Linux z13-017 4.4.0-10-generic #25-Ubuntu SMP Wed Mar 2 14:56:49 UTC 2016 s390x [13:06] this does not [13:06] check if anything changed with futexs in there [13:06] that would be my first guess [13:06] * Laney wonders how [13:06] git diff? [13:06] http://kernel.ubuntu.com/git/?s=idle [13:06] don't look at me :p [13:07] I guess http://kernel.ubuntu.com/git/ubuntu/ubuntu-xenial.git/ [13:09] spam has stepped up on my personal account in the last week or so [13:09] forget hosting this myself [13:15] i keep getting a string of very similar spam messages that the filter always fails to flag [13:16] i'm starting to wonder if the fastmail 'mark spam' button does anything at all [13:16] sender is always Sevgisandali and it's always the same message in turkish [13:22] yeah, I have dspam and it fails to catch duplicate things that I've already marked quite often [13:23] willcooke: don't know if you saw, but the change to make click scope depend on packagekit has landed in xenial. [13:23] oki, thanks dobey [13:24] davmor2, per our conversation on Friday, could you add the tests for codec downloading etc to your list?> [13:24] desrt: what does status 'reviewed' mean on bgo? [13:25] typically "i took a look at this but i am defering to someone else's judgement... see details in what i wrote" [13:25] https://bugzilla.gnome.org/show_bug.cgi?id=763617 [13:25] can be used if the person doing the review is not a committer, or if some changes were suggested but not mandated (ie: patch submitters choice about if to do it or not) [13:25] Gnome bug 763617 in general "giotypefuncs.c: Sort _get_type functions in the 'C' locale" [Normal,New] [13:25] * Laney gets pinged by reproducible people about this [13:25] * Laney giggles at "reproducible people" [13:28] Laney: a-c_n [13:28] ty [13:28] that rules is indeed hilarious [13:28] rule [13:29] * Laney rips own brain out [13:29] shut up about IDS [13:31] hey all [13:31] * Sweet5hark serves Laney a cookie and backs up slowly. [13:32] andyrock: good morning [13:32] howdy andyrock [13:37] Laney: can I forward you an email from an upstream asking for help with their AppStream icon? I'm not entirely sure what they should be doing [13:39] mhall119: ok [13:42] thanks Laney [13:50] replied [13:50] * Laney lunch [14:02] Laney: do we still actually need app-install-data? [14:02] mvo: ^ [14:03] (15 MB deb, 45 MB installed size, so nothing to sneeze at) [14:03] ATM sessioninstaller depends on it, but I wonder why? I thought sessioninstaller would merely install an already given pacakge name [14:12] pitti: not on the image for sure, maybe some other tools like the kde software center use it, but even if they do the data is now getting sstale, so lets remove it [14:12] pitti: it may use it to show fancy icons [14:35] Hey all, I noticed unity-scope-click was removed from my system in this morning's updates for some reason. When I try ti reinstall it, this is what I get: http://pastebin.ubuntu.com/15463069/ [14:35] Is there something special with python3-aptdaemon.pkcompat since it is trying to be removed? [14:36] BTW, when unity-scope-click is removed, then there is no App Scope in the Unity 8 desktop. === alan_g|lunch is now known as alan_g|afk [14:41] hi ChrisTownsend [14:41] willcooke: Hey:) [14:41] ChrisTownsend, yes - we're trying something to fix installing of Clicks in to the U8 session in 16.04 [14:42] we're going to drop the python compat and replace with with PK0.8# [14:42] and see what breaks in U7 [14:42] Click needs PK0.8 in order to work [14:42] willcooke: Ok. I wouldn't think removing the scope that is needed to actually launch apps in Unity 8 desktop is intended though, right? [14:43] no, that's a strange one [14:43] dobey, FYI ^ [14:45] willcooke: I'm not sure as to what actually wanted to remove it as apt was perfectly happy removing it and updating other packages and of course I just just 'Y' willy nilly:) [14:45] ChrisTownsend: i guess it might have gotten removed if there was a aptdaemon update at the same time [14:45] but otherwise, no, i can't tell you why it would have been removed. it probably shouldn't have [14:46] if you don't have the apt-get summary that you pressed 'Y' to, i can't make a more educated guess either really [14:47] dobey: Yeah, it scrolled off unfortunately. I have /var/log/apt/history.log, but that doesn't really reveal anything. [14:48] right. history log just shows it getting removed i guess, and not why it got removed [14:49] dobey: But if I try reinstalling unity-scope-click, it wants to remove python3-aptdaemon.pkcompat. [14:50] hello guys, is Ricardo Sousa present here by any chance? [14:52] ric_s: ^ maybe it's you? :) [14:54] yes mikedld|w === Drac0 is now known as Guest68692 [15:06] ChrisTownsend, yeah, we can't have PK0.8 && the Py compat layer, They clash on dbus namespace [15:12] willcooke: Ok, I see the issue. python3-aptdaemon.pkcompat conflicts with packagekit. And unity-scope-click depends on packagekit. [15:14] and I'm back on a system with a gui [15:14] willcooke: No idea how this would be fixed, but it's not good. [15:15] ChrisTownsend, well, we're dropping python3-aptdaemon.pkcompat [15:16] (at least, we're trying it out) [15:16] willcooke: Well, that's good. Was there a recent change here that broke this though? [15:17] ChrisTownsend, yeah, we changed it on Friday (IKR) [15:17] dropped pkcompat [15:18] so we can see if it a) fixes click installs on the U8 session and b) see what horrors emerge in U7 [15:18] Seb's theory is that not very much will break, and davmor2 is going to be testing that hypothesis this week [15:19] willcooke: Hmm, ok. I'm just confused about how this chain of conflicts recently showed up in the archive to where unity-scope-click was removed. [15:20] willcooke: I mean, I'm probably one of the few who actually does stuff in the Unity 8 desktop, but it's perplexing nonetheless. [15:21] willcooke: And if I reinstall unity-scope-click and it removes python3-aptdaemon.pkcompat, that will be OK? [15:24] ChrisTownsend, it /should/ be OK. But I'd like to know if it isnt. Not sure why u-s-click got removed. But we can look in to that as soon as Seb is back. [15:24] And yeah, it will only effect u8 session users [15:25] willcooke: Ok, thanks [15:26] ChrisTownsend, I think that particular issue will "only" affect people who already have u8 session installed on a 16.04 machine and upgrade. [15:26] but we'll take a look [15:27] willcooke: Ok. I took the plunge to see what happens. The Apps scope is back and I'll keep an eye out for other things. But I admit, it's just a test machine, so some things may never be observed that a regular user might see. [15:46] Trevinho, Not sure if this is in the queue? https://code.launchpad.net/~ubuntu-mate-dev/compiz/ubuntu-mate-wallpaper-common/+merge/288777 [15:46] flexiondotorg: mh, oh... I forgot to copy that -_- [15:46] flexiondotorg: next landing will be there. [15:46] Trevinho, Thanks :-) [15:47] flexiondotorg: thank you [16:01] aney: is everything on https://bugs.launchpad.net/ubuntu/+source/kdocker/+bug/1558713 done the way you wanted? [16:02] Launchpad bug 1558713 in kdocker (Ubuntu) "AppStream icon for kdocker.desktop" [Undecided,New] [16:02] will look soon [16:02] fixing upstream -> win [16:03] pitti: laney@nightingale> for r in trusty xenial; do echo -n "${r}: "; GET http://ddebs.ubuntu.com/dists/${r}/main/binary-amd64/Packages | grep-dctrl -c .; done ~ [16:03] trusty: 121 [16:03] xenial: 3021 [16:03] something borked with ddebs/trusty? [16:06] Laney: urgh, obviously :/ [16:07] was looking for compiz ones [16:07] they seem to be gone from the pool too [16:07] seems like a bad sign [16:19] hikiko: hey [16:20] hikiko: i'm trying to modify some behavior of ezoom in compiz. [16:20] i'd like to have the mouse always centered inside the zoomed area [16:20] i can't seem to find the right maths in the code. [16:20] hikiko: would you have any idea about that? [16:34] Laney, desktoppers: ignore the buttons on the righthand side for now, but thoughts on: http://imgur.com/JJrJ8b3 [16:36] willcooke: you hacked on what precisely? tabs? [16:38] yeah, tabs have different colour depending on current tab or not, plus I had to faff around with the borders between the tabs because the old one which was a light colour gradient clashed with the different colour tabs and gave a 1px white border between the tabs and the menu bar [16:38] Laney: it looks like trusty now only has the packages which have the same version in xenial [16:38] how on earth did that happen [16:39] willcooke: for terminal only, maybe it would be nice to have the same idea of firefox: so selected tab is white, others are matching the toolbar... what you think? [16:39] I'll start with gtk3 notebook [16:39] lemme get that done first :) [16:39] :) [16:40] so gedit will have something similar (not identical because GNOME Terminal seems to do things a bit differently) [16:40] actually, it should be identical [16:40] what do you think about the thickness of the border between tabs? [16:40] Look ok to you: [16:40] ? [16:41] ok as for the notebook itself, maybe the unselected color is still to be tuned imho... I can't explain how :-D, but that reminds me too much the windows95 gray :P [16:41] I think that's ok. [16:42] oki, cool. I'll tweak the colours and get design to take a look [16:42] oh I should fix those buttons on the right first [16:42] I hate those buttons [16:45] In other design related news..... http://imgur.com/JWjBRBl [16:45] Mh, [16:46] willcooke: that's something [16:46] hmm [16:47] I have no eye for this stuff [16:47] willcooke: did you tell GunnarHj & friends that you're working on that? [16:47] also [16:48] I attached gdb to compiz and did a release upgrade to xenial [16:48] https://paste.ubuntu.com/15465006/ [16:48] Laney, willcooke: I heard that. :) [16:49] compiz got multiarched in the meantime [16:49] could that break this? [16:49] hikiko: ^? === alan_g|afk is now known as alan_g [16:52] Laney: mh it could be that [16:52] Laney: however that trace seems suspcious to me... like some memory issue [16:52] yes [16:52] in fact that signature is the one used by the unity panel service iirc [16:52] whole chunks of the system got swapped out from under the running process [16:52] I only got one frame on the bt before gdb exited [16:53] #define ENTRY_SIGNATURE "(sssusbbusbbi)" [16:53] it mismatches because this is the one, but the old one was like the one in the trace [16:54] but, I don't see why the file should be unloaded... [16:54] Laney: almost done with implementing the icon-loading code in the new generator [16:54] it's probably likely that removing that file causes something problematic [16:54] but... [16:55] willcooke: dobey: I filed https://bugs.launchpad.net/ubuntu/+source/unity-scope-click/+bug/1560129 for tracking the issue I saw. [16:55] Launchpad bug 1560129 in unity-scope-click (Ubuntu) "unity-scope-click is either held back or removed depending on apt-get command" [Undecided,New] [16:55] Laney: unfortunately, the loading and processing of the Contents file slows the thing down massively [16:55] ChrisTownsend: well, "held back" makes sense [16:55] ChrisTownsend: it added a new binary depends; when that happens, apt generally holds it back until you dist-upgrade to pull in the new package [16:56] dobey: Yeah, which is probably what most users will see. But I'm not most users and I willy nilly say 'Y' on dist-upgrade:) [16:56] dobey: Well, dist-upgrade removes unity-scope-click. [16:58] dobey: As I tried to explain in the bug, packagekit conflicts with python3-aptdaemon.pkcompat which is already installed, so bye-bye unity-scope-click. [16:58] ChrisTownsend: well, python3-aptdaemon.pkcompat should get removed, generally [16:59] ximion: Thought you were making that yourself [16:59] ChrisTownsend: it would only be an issue if both are being upgraded at the exact same time [16:59] dobey: As willcooke explained, not when U7 is installed. And U8 and U7 co-existing is the norm. [17:00] ChrisTownsend: huh? [17:00] Trevinho: hm I removed that whole directory and it didn't break [17:01] Laney: not yet, because I run into an henn-and-egg problem there: To process metadata, I need to know about all icons in all packages. But that information is not yet available when going through the packages, because some might be processed later in the queue. So I would need two passes at the packages, one for building the contents information and marking interesting packages, and one for actually extracting the data [17:01] that is much slower than using the Contents file, so I focus on the traditional method now, and the "we-build-contents-file" method will be implemented later [17:01] ChrisTownsend: this is an issue with how apt resolves dependences, and just bad timing with the update [17:02] I also want to reach feature parity with the dep11-generator ASAP, so that's the focus at time [17:02] dobey: Yeah, most likely the case. I guess everyone will just have to work through it then. [17:03] hrmm, but aptdaemon wasn't updated [17:03] so dist-upgrade should have pulled in packagekit, and removed python3-aptdaemon.pkcompat [17:04] dobey: Maybe this? [17:04] $ apt-cache depends ubuntu-desktop | grep python3-aptdaemon.pkcompat [17:04] Recommends: python3-aptdaemon.pkcompat [17:05] Although it's not a specific depends, so I may be grasping at straws. [17:05] ximion: The first pass is to look at the contents, not the data itself [17:05] ChrisTownsend: that's just a Recommends, so it should have just been removed as a result of that [17:05] And only has to be done once per .deb [17:05] and as you said, if you just install unity-scope-click again after, it works; so it should have worked during the upgrade [17:05] ximion: did you apply for that funding yet? [17:06] ChrisTownsend: was "ubuntu-desktop" in the updates list? [17:06] dobey: Yeah, I really don't know why apt wanted to hold on to it. It was happy to remove it if I manually (re)installed unity-scope-click. [17:06] ChrisTownsend: can you attach the bit of the apt history log from the time when you ran dist-upgrade to the ubg? [17:06] bug even [17:06] Laney: will do that in an hour (funding) [17:06] 'k [17:07] dobey: Yes, ubuntu-desktop was in the updates list. [17:07] dobey: Yes, I'll attach the log. [17:07] Laney: even for the contents, we need to unpack the ar archive, and open up the data.tar.* tarball - this is much faster than in the dep11-generator, but still not very cheap [17:08] regardless of that, the "we generate contents data" approach will come [17:08] ximion: Doesn't it use the md5sums in control? [17:08] I believe you that it's slower [17:08] but in terms of development effort I would avoid doing something that you know you're going to throw away [17:08] Laney: no - but that's something we could do, haven't thought about that at all yet - that could be a massive optimization [17:09] especially on a rewrite [17:09] ChrisTownsend, dobey - I'm a bit lost I think, but just wanted to state that the *default* situation should be that in 16.04 python3-aptdaemon-pkcompat should be as it is today (i.e. installed). Only if the user installs unity8-desktop-session-mir should PK0.8 get installed and py3-pkcompat get removed. [17:09] ximion: you're handling this manually? [17:09] :-o [17:09] s/today/last Monday [17:09] willcooke: that is the default behavior now [17:09] dobey, cool! thanks [17:10] Laney: jup, using libarchive directly turned out to be 1.5 -2 times faster spawning dpkg [17:10] http://pastebin.ubuntu.com/15465408/ <- that's what i get if i try to install unity-scope-click right now [17:11] dpkg itself also seems to run tar and xz, while libarchive has archive support built-in, which works well with the multithreading [17:11] for getting the contents you don't really want to use the tar file if you can help it [17:11] :) [17:11] tar sucks for that [17:11] dobey: Yeah, I had something similar to that when installing unity-scope-click manually after it was removed. [17:11] Trevinho, could you follow up with hikiko on that issue about compiz crashing on upgrade. She's been looking at it today [17:12] dobey: I also attached the history log to the bug. [17:12] ok [17:12] dobey: Thanks [17:12] willcooke: I've asked also andyrock to look at it, he's playing with VMs right now [17:12] nice one, thanks Trevinho, andyrock. hikiko FYI ^ [17:14] ChrisTownsend: i'll take a peak, but so far all i can tell is that it will work correctly for new users of 16.04. it's only potentially an issue on upgrade if unity-scope-click is already installed. so moderately low priority [17:15] Laney: we will still need to open the control tarball ^^ - but since that one is smaller, chances are we can process packages even faster (at time, the generator does roughly 4-5 packages per second on my machine) [17:15] it's not the opening [17:15] it's reading the entire huuuuuuuuuge thing to get the filenames [17:16] looking forward to hearing how much this helps [17:17] dobey: Ok, thanks. We just need to be prepared for those folks who do already have it installed and they complain that the Apps scope is gone (if they do dist-upgrade) or wonder why they still can't install clicks when we are saying they can (due to it being held back). Not complaining, just saying:) [17:19] ChrisTownsend: sure; this dependency might not stay there either. we added it so it can be more easily tested if packagekit being installed breaks other unity7 features [17:19] dobey: Ok. Thanks again. [17:20] Laney: I'll know that soonish - need food first. Btw, I reduced the dependencies of the generator a little, the only thing I now still want to get rid of is the gdlib for image rendering. I originally used it to avoid gdk-pixbuf and pulling in X11 stuff etc. and make the whole think more lightweight - but it turns out librsvg2 depends on gdk-pixbuf anyway, and if I want to render SVGs, there's no way around it [17:20] sure, np :) [17:21] Laney: so I could use gdk-pixbuf directly anyway ^^ - will do that later, when I feel motivated enough to dive into image rendering / scaling again [17:21] (huge pain - but Cairo + librsvg2 from D is again faster than using it from Python, which in this case makes less sense, since both should only be calling C stuff) === muktupavels_ is now known as muktupavels [17:32] Laney: britney is running on https://requests.ci-train.ubuntu.com/#/ticket/1146 for some time, once it's done, it's good to me. [17:33] Trevinho: ok, ta [17:35] Trevinho: how do you see the britney thing? [17:35] Laney: I can't... I only know it *should* run. [17:35] where is it normally linked from? [17:35] * Laney is a train newbie [17:43] Trevinho: whatever, forget britney [17:43] I published it [17:43] ooooooooooh no [17:43] fix that! [17:46] Trevinho willcooke hikiko I'm building unity with debug info on a vm right now [17:46] hopefully i can reproduce the issue [17:46] andyrock: you should do compiz too [17:46] Laney: done [17:46] ok [17:46] Laney: i already did that [17:47] nux and compiz [17:47] okey [17:47] and unity is on his way right now [17:47] thanks andyrock - just fwd'd you a mail with some steps to reproduce [17:48]