/srv/irclogs.ubuntu.com/2015/09/22/#ubuntu-desktop.txt

cyphermoxdupondje: hey01:50
=== maclin1 is now known as maclin
pittiGood morning03:51
duflupitty: afternoon :)04:03
duflupitti: afternoon :)04:03
dufluWow, dyslexia corrects my spelling04:03
dufluor not04:03
duflutoo many ts04:04
sarnoldit took me entirely too long to spot the difference :)04:05
pittiheh, it's a pitty, isn't it :)04:16
didrocksgood morning06:13
pittibonjour didrocks, ça va ?06:18
didrocksça va pitti, et toi ?06:20
pittididrocks: je vais bien, merci !06:22
didrockshum, on the phone ML: "06:26
didrocksI've now opened up publication permissions to everybody … This means that if your silo does not touch any files under debian/, you can publish your own silos totally by yourself, no trainguards required."06:26
didrocksI wonder if the tech board/release team is aware about this06:26
didrocksit means that publications are not handled by people with upload rights anymore06:27
didrocksLaney: thought? did you know about it? ^06:27
robrudidrocks: yes slangasek told me this was agreed with tech board.06:27
didrocksah nice!06:28
robruthe rule is as long as debian/ isn't touched06:28
robruand only for canonical-owned projects06:28
didrocksrobru: do you have anything to guard when debian/ is touched, just for the curiousity?06:28
didrocksor it's just the existing warning?06:28
thumpero/06:28
robrudidrocks: yes, the same "packaging diff" is in place, ACK requires proper checkUpload rights06:28
* thumper is done for the day06:29
thumperI miss you folks06:29
didrocksrobru: ok, as long as people are not too eager clicking checks, that's a nice one! :)06:29
didrockshey thumper!06:29
didrocksthumper: have a nice evening, long time not seeing you!06:29
thumperyeah.. one day... one day we might have another all hands06:29
didrocksheh, yeah, one day ;)06:29
thumperit'd be epic06:29
thumpermaybe when we actually make money06:30
thumper:)06:30
didrocks;)06:30
thumperlaters06:30
=== maclin1 is now known as maclin
pittididrocks, robru: err, I certainly didn't agree to this; you can break stuff with the upstream parts just as well as with the packaging bits; and the silo is not limited to touch-y packages, you can upload kernels or glibc or whatever to it06:31
robrupitti: uh, our boss told me to do this, sorry, wasn't my decision06:31
pittiwell, "my boss told me" != "TB agrees" :)06:32
robrupitti: he told me it was agreed with the TB.06:32
robrupitti: I'll dig up the emails and forward them06:33
pittihm, so we now have folks who never signed the CoC, never received any ubuntu dev training, ignore test results, and have unrestricted access to the ubuntu archive06:34
didrocksthat's the difference with the daily release, indeed, when we had a filtering list (in addition to the bot which only uploaded some components) on components to autopublish06:34
robrupitti: didrocks: some discussion here: https://bugs.launchpad.net/cupstream2distro/+bug/145918606:35
ubot5Ubuntu bug 1459186 in CI Train [cu2d] "Implementing proper permission checks during silo publish" [High,Fix released]06:35
pittididrocks: was that a public ML, or internal one? I. e. anything to refer to on the TB list?06:35
robrupitti: if there are train users you know of who didn't sign CoC please let me know and I'll hound them06:35
didrockspitti: I was referring to the announcement on the phone ML06:37
pittirobru: hm, wasn't that bug just the opposite? restricting silo uploads (or at least publication) to people who can upload that package directly to ubuntu too?06:37
didrocksyeah06:37
didrocksI understand the bug the same06:37
didrocksespecially with: "2) Signoff permissions should track the archive permissions for a package: respecting archive components (ubuntu-core-dev vs. motu) and per-package uploaders."06:37
pittididrocks: I meant, was it the public phone ML?06:37
robrupitti: in the comments slangasek explains the exception for canonical-owned projects that have no debian/ changes06:37
didrockspitti: yes "IMPORTANT ANNOUNCEMENT: CI Train changes"06:38
didrocksrobru: yeah, but he seems to intend that "must enforce checkupload for merges with diffs, all manual sources, and NEW.06:38
didrocks"06:38
didrocksas you wrote06:38
pittirobru: right, but that certainly doesn't cover things like glibc, qt, systemd, gcc, etc.06:38
didrocksdid you add those checks?06:38
didrocks(before opening publication for the other case to everyone)06:39
robrupitti: maybe it wasn't clear, not just anybody can upload anything. train is not allowing kernels or glibc published by anybody. only canonical-owned stuff.06:39
pittirobru: i. e. you can upload anything directly to the silo06:39
robrupitti: you can upload anything to the silo but the only time checkUpload is not enforced is if the package is owned by canonical and has no debian/ changes.06:39
pittirobru: ah, which creds do you need to upload directly to a silo?06:39
didrocksah, so you only enable people to publish if there is no change in debian/, (and so, not a new packages) + only canonical project?06:39
robrupitti: also actually not just anybody can upload to the silos.06:40
pittirobru: ok, so for publishing the silo it checks the packages against a whitelist?06:40
didrocksok, the announcement wasn't clear about it :)06:40
robrupitti: you need ~ci-train-ppa-service membership to upload sources to silos.06:40
robrupitti: there's no whitelist per se, currently the train assumes "MP = canonical owns this", which is reasonably true at this point as there are rather narrow packaging requirements that the train enforces that other random source packages don't have.06:41
robrupitti: eg if you were a bad actor and you tried to fake an MP to get around the check, your MP would necessarily have a ton of packaging changes that would trigger the check.06:41
robruor it wouldn't even build06:42
pittirobru: thanks for clarifying06:45
robrupitti: you're welcome! Feel free to raise this with slangasek when he's back next week, I'm happy to revisit this (actually I wasn't even involved in the original decision anyway)06:46
pittirobru: well, it sounds like it's certainly far from "if your silo does not touch any files under debian/, you can publish your own silos totally by yourself"06:46
robrupitti: right, it's for MPs only, i forgot to mention that.06:47
Mirvdidrocks: the ACK:s are now not possible by anyone who does not have the actual upload rights to the particular package07:02
Mirvso, all main packages with debian/ changes always require a core-dev to run the publishing07:02
Mirvrobru: pitti: manually uploaded packages can't be published regardless of which changes they had, the publisher needs to have MOTU or core-dev rights accordingly07:03
Mirvoh, as discussed07:03
robruheh, yeah07:04
didrocks;)07:04
didrocksthanks for confirming, I guess he main misunderstanding is coming from the email (a little bit too vague)07:04
MirvI just started from where I was higlighted with 'trainguards'07:04
didrocks:p07:04
robruI'll clarify the email07:05
didrocksthanks robru!07:05
robrudidrocks: you're welcome07:05
larsugood morning!07:16
seb128good morning desktopers ;-)07:18
* larsu gets some Schrippen07:19
seb128larsu, bread?07:20
didrocksgood morning larsu, seb128!07:21
seb128lut didrocks07:22
larsuseb128: https://en.wikipedia.org/wiki/Bread_roll07:22
larsuit even contains that word :)07:22
pittibonjour seb128 !07:22
pittihey larsu, guten Appetit!07:22
larsusalut didrocks and pitti !07:22
seb128pitti, salut, ça va ?07:22
larsudanke :)07:22
pittiseb128: ça va bien, et toi ?07:22
seb128pitti, ça va bien aussi merci ;-)07:27
larsupitti: Laney suggested last night that there might be some dh_ magic for my rules file to get language-specific things installed07:33
larsuor would I need to hack langpacks for that?07:33
pittilarsu: what are "things"?07:34
pittilarsu: pkgstriptranslations currently strips out GNOME help files07:35
pitti(both old and mallard)07:35
larsupitti: translations of place names07:35
larsufor the list in system settings07:35
larsu(when autocompleting time zone)07:35
pittilarsu: are these po files?07:35
pittiI really suggest that this should use standard gettext with po/mo files, not some custom XML lists or what not07:36
larsupitti: it would be more efficient if they weren't07:36
larsuI was kind of expecting this answer :)07:36
pittilarsu: so, we have a concept of "static" translations, like translated help files or screenshots; pkgbinarymangler can pick those up and put them into the _static.translations.tar.gz tarball07:37
pittithey will then end up in langpacks07:37
pittibut these are hard to translate obviously07:37
pittiso if you actually care for translations, do use gettext -- what makes this hard?07:37
larsuI don't07:37
larsuI already have the translations07:37
larsuthey're part of geonames07:38
larsupitti: how does this work exactly?07:38
dupondjenobody with wily that uses L2TP ? ;) might be cool if somebody could test https://bugs.launchpad.net/ubuntu/+source/network-manager-strongswan/+bug/1487183 on wily.07:39
ubot5Ubuntu bug 1487183 in network-manager-strongswan (Ubuntu) "Upgrade nm-strongswan to latest 1.3.1 with psk support" [Undecided,Incomplete]07:39
pittilarsu: so, if you e. g. have a file "geonames_de.xml", you can put that into the static tarball; and then langpack-o-matic needs a corresponding patch to recognize those, detect the locale from the path or file name, and put that into the corresponding langpack07:41
pittibut that's a really inflexible hack07:41
pittilarsu: again -- please do use gettext at runtime, static translations are hideously hard to translate, process, and update; we also don't support translating them on LP07:42
pittilarsu: if you need some fast typeahead search, build a list of all translated items (for your current locale) in memory07:42
pittibut I figure adding a dgettext() call to the iteration should be okay07:42
pittidgettext is fast07:43
seb128dupondje, you might want to try on #ubuntu-devel unsure anyone on desktop use strongswan07:44
larsupitti: if I use it at runtime I might as well use .po files. Not sure that we'll ever need translations on launchpad for this, but I guess you're right: let's go with the known workflow first and change it when if we notice problems07:46
larsupitti: thanks!07:46
pittilarsu: asked the other way around, what makes calling dgettext at runtime harder than using some pre-translated separate files?07:46
larsupitti: it's not harder, but doing a string lookup (with context, as there are different cities with the same name in different countries) might be less efficient07:48
larsu(not in speed, but memory)07:48
larsubut lets see how big this will get07:48
pittilarsu: you have the .mo files on the system anyway, no?07:49
pittilarsu: I can't see how loading a pre-translated set of names into memory would be much more efficient07:49
larsupitti: hm, indeed. I got most of this from a short discussion with desrt yesterday. I'm not yet sure myself :)07:51
Sweet5harkmoin07:56
seb128hey Sweet5hark07:58
willcookegoodly morlode07:59
didrockshikiko: hey, when you get a chance, do you mind having a look at https://code.launchpad.net/~didrocks/unity/remove-unity2d/+merge/271916?08:02
seb128hey willcooke08:02
seb128is that command working for others?08:02
seb128$ /usr/share/software-center/piston_generic_helper.py --datadir /usr/share/software-center/ --needs-auth --no-relogin SoftwareCenterAgentAPI subscriptions_for_me08:02
seb128I've errors about it in my syslog08:03
seb128mvo_, dobey, ^ do you know about that?08:03
seb128"ERROR: can not obtain a oauth token"08:03
didrocksseb128: opening a window, want me to connect?08:04
hikikohey didrocks08:04
hikikosure08:04
didrocksthx :)08:04
seb128didrocks, if you can, just to see if it works08:04
mvo_seb128: same failure, without further looking at this I wonder if the sso api has changed maybe?08:04
seb128mvo_, yeah, unsure ... what would be the component to report that on?08:04
mvo_seb128: its part of software-center08:05
Laneyhullo08:05
didrocksseb128: mvo_: interesting, wfm though (and I see my previous purchases)…08:05
mvo_seb128: or did you maybe revoke your token at some point?08:05
didrockshey Laney08:05
hikikodidrocks, give me 10 minutes to get rid of my built unity and I ll compile it08:05
seb128mvo_, I didn't but maybe it had been revoked for some external factor, shouldn't it reprompt me then?08:06
seb128didrocks, thanks08:06
mvo_seb128: iirc it will ask via dbus for a token from the sso-client and then verify the client against the server. for me it seems to fail in this step, what backtrace do you get?08:06
mvo_seb128: and yes, it should re-prompt you, not sure why its not doing this08:06
mvo_seb128: aha, because of "--no-relogin"08:06
seb128File "/usr/lib/python2.7/dist-packages/piston_mini_client/__init__.py", line 491, in get08:07
seb128return self.request_url(url, method='GET', headers=headers)08:07
seb128File "/usr/lib/python2.7/dist-packages/piston_mini_client/__init__.py", line 435, in request_url08:07
seb128body = handler.handle(response, response_body)08:07
seb128File "/usr/lib/python2.7/dist-packages/piston_mini_client/failhandlers.py", line 116, in handle08:07
seb128data=self.data)08:07
mvo_seb128: is that the full bt? if not, could you pastebin the full one? no api error or something like this?08:08
seb128mvo_, http://paste.ubuntu.com/12519695/08:10
mvo_seb128: thanks, thats similar to what I see - looks like a genuine bug :/08:13
seb128mvo_, is that something dobey or nessite would know about if it's ubuntusso?08:18
mvo_seb128: so if you use seahores and delete the "Ubuntu one" token, I suspect it will ask you to relogin?08:18
seb128mvo_, I guess I could try but then I loose the way to reproduce and test a fix/get more info08:18
mvo_seb128: not sure if they will know, maybe re-prompt expired tokens never got implemented in s-c, I'm not sure08:18
mvo_seb128: ok, then keep it this way08:18
Laneyhttps://bazaar.launchpad.net/~ubuntu-art-pkg/ubuntu-wallpapers/ubuntu/view/head:/warty-final-ubuntu.png09:05
LaneyO_O09:06
Trevinhomhmh09:06
willcookewat09:07
TrevinhoI guess that nautilus desktop is smart enough to convert that log in some kind of art...09:08
Laneyturns out that dbus logs translate exactly to the Ubuntu default wallpaper09:08
LaneyWHO KNEW09:08
TrevinhoNow that we know the trick, we can just do the new art design by ourselves at every release :)09:09
seb128lol09:10
didrocksseeing the number of "Laney" occurence in the calendar events, we don't know at all who committed that for sure :p09:13
Laneyhttps://bazaar.launchpad.net/~ubuntu-art-pkg/ubuntu-wallpapers/ubuntu/changes?filter_file_id=elephantskin.jpg-20070925092320-wp63xkqaf9y3tbmp-109:13
Laneythis probably gives it away too :P09:13
didrocksLaney: quick, whistle and walk away :p09:13
Laneywillcooke: btw...09:13
Laneylooking at this reminds me that we have an "official" greyscale too09:14
Laneydo we get a new version of that?09:14
willcookeLaney, I'll speak to design now...09:14
Laney/usr/share/backgrounds/Suru_Wallpaper_Desktop_4096x2304_Gray.png09:14
Laneyit's based on the colour one09:15
didrocksthanks for the merge Trevinho :)09:15
Laneyso hopefully not much work09:15
willcookeyeah, should be easy09:16
* didrocks was cleaning his desktop, hence the unity2d notice09:16
Trevinhoseb128: can you ack me this https://code.launchpad.net/~3v1n0/libunity/launcher-use-class-settings/+merge/271847 ?09:16
Trevinhodidrocks: np, I'll land that soon09:16
Trevinhoseb128: for bug #1470097 is there a way to get those ddebs on a ppa instead of waiting the SRU thing?09:21
ubot5bug 1470097 in unity (Ubuntu) "Unity does not support cpus without sse4" [High,Incomplete] https://launchpad.net/bugs/147009709:21
seb128Trevinho, you can do an upload to a ppa, citrain ones include ddebs I think09:22
seb128Trevinho, let me review the settings one09:22
Trevinhoit shouldn't https://bugs.launchpad.net/ubuntu-lt/+bug/142018509:23
ubot5Ubuntu bug 1420185 in Ubuntu Landing Team "A way to provide ddebs to landing PPA:s?" [Wishlist,Fix released]09:23
Trevinhoah, last comment doesn't match the bug status, so maybe...09:23
hikikodidrocks, it looks good to me09:27
hikikoI can't see anything wrong09:27
hikikoI am going to approve it09:27
didrockshikiko: great! I think that Trevinho approved it already though09:28
hikikohahaha09:28
hikikolol and I didn't look at the diff, I built it first...09:28
hikiko\m/09:28
hikikothen I noticed09:28
hikikoit's just text09:28
hikiko:p09:28
hikikofail :p09:29
seb128Trevinho, libunity change  approved, thanks09:34
Trevinhoseb128: thank you!09:36
seb128Trevinho, yw! thanks for the fix ;-)09:37
Trevinhoseb128:  I was thanking you... Then you went out, so my tab autocompletion was actually about to mention s---abdfl09:38
Trevinhothank God i didn't press enter too quickly :D09:38
seb128lol09:38
seb128Trevinho, seems like xchat-gnome closed/segfaulted/something09:38
Trevinhoseb128: I've this one for you too btw :) https://code.launchpad.net/~3v1n0/compiz/vivid-rebuild-ddebs/+merge/27193009:38
seb128I saw the notify-osd bubble but xchat was closed09:38
seb128Trevinho, there is a merge conflict there09:39
Trevinhoah, yeah I was wondering what happened... That's weird that it was a segfault though as we got the "Remote host closed the connection", and not ping timeout09:39
Trevinhoseb128: sorry, I ddin't propose to vivid branch (lp-propose doesn't work that way -_-)09:39
Trevinhoseb128: this should work https://code.launchpad.net/~3v1n0/compiz/vivid-rebuild-ddebs/+merge/27193109:40
seb128Trevinho, right, approved09:42
Trevinhoseb128: thanks09:43
seb128yw!09:43
Trevinho... and silo'ed09:43
=== alan_g is now known as alan_g|lunch
=== alan_g|lunch is now known as alan_g
TrevinhoEuropeans, do the agency replied promptly to you for tickets request? I got no answer yet (since yesterday)13:30
pittiTrevinho: Omnia? yes, they usually do within a few hours13:30
Trevinhopitti: yeah, generally so it happens, but not this time13:30
larsuTrevinho: they replied to my email from last night an hour ago13:31
Trevinholarsu: mh, ok... maybe I'm on the queue then13:31
larsuTrevinho: ya, probably they're just a bit busy right now13:31
* larsu files dbus bug after not being able to hunt the problem down himself13:37
Laneyiz dbug bug?13:38
larsuyes13:38
larsuhttps://bugs.freedesktop.org/show_bug.cgi?id=9207413:38
ubot5Freedesktop bug 92074 in core "destination= rules don't match anything when using BecomeMonitor" [Normal,New]13:38
seb128Laney, btw, new glib stable is out, I guess it's on your list?13:38
larsuthankfully we have busctl monitor now, which works as expected because it doesn't yet use the new BecomeMonitor stuff13:38
seb128I'm asking because I'm chassing some of the warnings spam in logs and they turn off deprecation warnings13:38
seb128which I could use ;-)13:39
didrockshaha, they did as well, nice \o/13:39
larsuoh wow13:39
Laneyyes thanks13:39
larsuI guess they didn't consult desrt?13:39
seb128https://git.gnome.org/browse/glib/commit/?id=3bd1618ea955f950f87bc4e452029c5f0cea35aa13:39
seb128unsure13:39
seb128no bug report mention13:39
larsuthat commit message is gold13:40
LaneyI think it's just the done thing now13:40
larsu(and spot on)13:40
desrtalso only on the stable release, as we did before13:40
seb128I'm glad people agreed on that13:40
larsudesrt: you argued against doing that....13:40
seb128desrt, there was some pushback previous cycles to do it even in stable13:40
desrtlarsu: and yet, did it :)13:40
larsuindeed13:41
desrti don't like it, but i agree with the commit message13:41
didrocksseb128: +1+1+1 :)13:41
seb128didrocks, :-)13:41
Laneyglad you are all so excited13:41
* larsu learns about org.freedesktop.DBus.Debug.Stats13:41
* desrt yawns13:41
larsuhow is that not specified? This is awesome stuff13:41
desrtgot an extra-large coffee today.13:41
desrt*bzz*13:42
Laneybzbzbz13:51
chrisccoulsonricotz, please let me know if you have any issues with the firefox-beta.* branches (I've not tried creating a Firefox 42 upload from it yet)14:12
ricotzchrisccoulson, I will notice soon then14:35
=== hikiko-lpt is now known as hikiko
qenghochrisccoulson: Will you please give me a tutorial of your firefox workflow one day?15:16
willcookeit's coming15:26
willcooke#startmeeting Desktop Team Weekly Meeting - 2015-09-2215:30
meetingologyMeeting started Tue Sep 22 15:30:53 2015 UTC.  The chair is willcooke. Information about MeetBot at http://wiki.ubuntu.com/meetingology.15:30
meetingologyAvailable commands: action commands idea info link nick15:30
=== meetingology changed the topic of #ubuntu-desktop to: Home of the Desktop Team, https://wiki.ubuntu.com/DesktopTeam | For help or questions, try #ubuntu | Breaking your systems: GTK 3.16 - file bugs with tag "gtk316" | Desktop Team Weekly Meeting - 2015-09-22 | Current topic:
didrockslet's meet!15:30
willcookeRoll call: andyrock (out), attente, desrt,  dgadomski, didrocks, fjkong, happyaron (out), hikiko, laney, larsu, qengho, seb128, sweet5hark (probably out), themuso (out), tkamppeter, trevino, robert_ancell (out)15:31
* Trevinho with "h" is here :)15:31
Laneyrevino15:31
Laneyevino15:31
FJKonghi15:31
Laneyvino15:31
Laneyino15:31
Laneyno15:31
Laneyo15:31
seb128hey15:31
Laney 15:31
dgadomskihey15:32
TrevinhoVino... (means wine in Italian :p)15:32
qenghoThree wines, plz15:32
Trevinhothat's it :D15:32
larsu\o15:32
* willcooke fixes his list15:32
willcookeand has one of the wines15:33
Trevinho:D15:33
willcookeRight then...15:33
willcooke#topic andyrock15:33
=== meetingology changed the topic of #ubuntu-desktop to: Home of the Desktop Team, https://wiki.ubuntu.com/DesktopTeam | For help or questions, try #ubuntu | Breaking your systems: GTK 3.16 - file bugs with tag "gtk316" | Desktop Team Weekly Meeting - 2015-09-22 | Current topic: andyrock
willcooke* Bug scrubing15:33
willcooke* Investigating on why the scopes on 15.04+ are not reliable15:33
willcooke* Code reviews15:33
willcooke* Helping understanding the cause of #1497002 Recent Wily i386 live15:33
willcookeDVDs load with no launcher, top bar, desktop icons, etc. There is a15:33
willcookequick workaround15:33
willcooke* Updating some old branche15:33
willcooke#topic attente15:33
=== meetingology changed the topic of #ubuntu-desktop to: Home of the Desktop Team, https://wiki.ubuntu.com/DesktopTeam | For help or questions, try #ubuntu | Breaking your systems: GTK 3.16 - file bugs with tag "gtk316" | Desktop Team Weekly Meeting - 2015-09-22 | Current topic: attente
attentepopup menu placement continued, in the process of porting over several gtk widgets to it15:34
attentespent a day up at xdc with desrt and robert_ancell15:34
attente(eof)15:34
willcookethanks attente15:34
willcooke#topic desrt15:34
=== meetingology changed the topic of #ubuntu-desktop to: Home of the Desktop Team, https://wiki.ubuntu.com/DesktopTeam | For help or questions, try #ubuntu | Breaking your systems: GTK 3.16 - file bugs with tag "gtk316" | Desktop Team Weekly Meeting - 2015-09-22 | Current topic: desrt
willcookedesrt we will come back to you15:36
willcooke#topic dgadomski15:37
=== meetingology changed the topic of #ubuntu-desktop to: Home of the Desktop Team, https://wiki.ubuntu.com/DesktopTeam | For help or questions, try #ubuntu | Breaking your systems: GTK 3.16 - file bugs with tag "gtk316" | Desktop Team Weekly Meeting - 2015-09-22 | Current topic: dgadomski
dgadomskihey15:37
dgadomskionly 1 thing in the desktop area this week: I've been looking into the reason for problems with installing PPA version of octave on trusty with HWE updates - bug #1424059. Looks like the cause was incorrect resolving of versioned dependencies with 'Provides:'. A workaround for now is to prepare a osmesa-less build of octave (agreed with the ppa maintainer and tested).15:37
ubot5bug 1424059 in mesa (Ubuntu Trusty) "libosmesa6 conflicts with libglapi-mesa-lts-*" [High,In progress] https://launchpad.net/bugs/142405915:37
dgadomskiEOF15:37
willcookethanks a lot dgadomski15:37
willcooke#topic didrocks15:37
=== meetingology changed the topic of #ubuntu-desktop to: Home of the Desktop Team, https://wiki.ubuntu.com/DesktopTeam | For help or questions, try #ubuntu | Breaking your systems: GTK 3.16 - file bugs with tag "gtk316" | Desktop Team Weekly Meeting - 2015-09-22 | Current topic: didrocks
didrockso/15:37
didrocks!\ This reports include the last 2 weeks as was on a conference last week15:38
ubot5didrocks: I am only a bot, please don't think I'm intelligent :)15:38
didrocksUbuntu Make:15:38
didrocks- released 15.09.2 featuring Android SDK downlod only, exit code fixes (both of those, by external contributor)15:38
didrocks- added additional pep8 and tests fixes. Loving pep8 on trusty btw (no easy way to get something passing on trusty pep8 and wily pep8 in term of identation) :p15:38
didrocks- archived our daily jenkins test jobs to ensure we have an in-code backup15:38
didrocks- bugs management and PR reviews15:38
didrocksDeveloper experience:15:38
didrocks- experiment and play on our developer experience writing a Go service on snappy15:38
didrocks- write that experience on a (too long) documentation, one with and without snapcraft, on a VM and with a raspbery pi 215:38
didrocks- draft what our next step for developer experience on snappy should be and separation of tools concerns15:38
didrocks- attended to various meetings on this15:38
didrocksMisc:15:38
didrocks- attended to Polymer first ever conference. Back with some plastic ideas on how our HTML5 developer story should be15:38
didrocks- talked with dbarth about an eventual prototyping around our HTML5 developer story & polymer would be15:38
didrocks- archive admin and patch piloting duties15:38
didrocksEOF15:38
willcookethanks didrocks15:38
willcooke#topic FJKong15:38
=== meetingology changed the topic of #ubuntu-desktop to: Home of the Desktop Team, https://wiki.ubuntu.com/DesktopTeam | For help or questions, try #ubuntu | Breaking your systems: GTK 3.16 - file bugs with tag "gtk316" | Desktop Team Weekly Meeting - 2015-09-22 | Current topic: FJKong
FJKongpinyin search:15:39
FJKongimprove speed of indexing file to db15:39
FJKongresearch on index every first Chinese Character15:39
FJKongswitch to cmake15:39
FJKongsogou pinyin:15:39
FJKongtest right clicking pop menu15:39
FJKongeof15:39
willcookethanks FJKong15:40
willcookeFJKong, any visa issues or is that all good?15:40
desrtzomg.  sorry!15:41
FJKongno problem15:41
willcookeFJKong, great!15:41
willcooke#topic desrt15:41
=== meetingology changed the topic of #ubuntu-desktop to: Home of the Desktop Team, https://wiki.ubuntu.com/DesktopTeam | For help or questions, try #ubuntu | Breaking your systems: GTK 3.16 - file bugs with tag "gtk316" | Desktop Team Weekly Meeting - 2015-09-22 | Current topic: desrt
FJKongI have VISA for 2 years for UK15:41
willcookeFJKong, excellent, that's worked out nicely then15:41
willcookehey desrt15:42
desrthey.  had a minisprint last week for a few days, also went to xdc for a day and worked on the usual inotify stuff during friday.  did some patch reviewing and stuff this morning.15:42
Laneydoing better the Ai Weiwei15:42
desrtnot much else to report15:42
willcookethanks desrt15:42
willcooke#topic happyaron15:43
=== meetingology changed the topic of #ubuntu-desktop to: Home of the Desktop Team, https://wiki.ubuntu.com/DesktopTeam | For help or questions, try #ubuntu | Breaking your systems: GTK 3.16 - file bugs with tag "gtk316" | Desktop Team Weekly Meeting - 2015-09-22 | Current topic: happyaron
willcooke1. fcitx package update (but not a complete fix, thanks Laney)15:43
willcooke2. libxml2 NMU mess resolved for Debian15:43
willcooke    http://anonscm.debian.org/cgit/debian-xml-sgml/libxml2.git/diff/?h=debian/2.9.2%2bzdfsg1-4&id=76c19f4d5b3328c05649314336d27c1f44a49e96&id2=fccd80a7d35b9eefabcc384d5c9fe89d9545dd2215:43
willcooke3. ubiquity slideshow update for Ubuntu Kylin15:43
willcooke4. fcitx upstream bug triaging/analysis:15:43
willcooke    a) misconfiguration cases - a lot for those who turn off the15:43
willcookeinstallation of Recommends packages15:43
willcooke    b) fcitx ^w out-of-process UI has issues when the desktop runs15:43
willcookewith thin client configurations15:43
willcooke5. pinyin search packaging: in progress15:43
willcooke#topic hikiko15:43
=== meetingology changed the topic of #ubuntu-desktop to: Home of the Desktop Team, https://wiki.ubuntu.com/DesktopTeam | For help or questions, try #ubuntu | Breaking your systems: GTK 3.16 - file bugs with tag "gtk316" | Desktop Team Weekly Meeting - 2015-09-22 | Current topic: hikiko
hikikohi, : closed some bugs, fixed the cube and deformed cube caps images that appeared off-center, and I am going to investigate this shadow bug and  a few other issues that were revealed while I was trying to reproduce the problem and fix the cubeaddon eof15:44
willcookethanks hikiko15:44
willcooke#topic Laney15:44
=== meetingology changed the topic of #ubuntu-desktop to: Home of the Desktop Team, https://wiki.ubuntu.com/DesktopTeam | For help or questions, try #ubuntu | Breaking your systems: GTK 3.16 - file bugs with tag "gtk316" | Desktop Team Weekly Meeting - 2015-09-22 | Current topic: Laney
Laney• Normal release team stuff (FFes and so on)15:45
Laney• Test/sync glib15:45
Laney• Work on some s-l-o-w machines to address a gdk-pixbuf build failure - Company committed it upstream & I just uploaded to Debian, will sync if it works.15:45
Laney• gtk 3.16.715:45
Laney• glib 2.45.815:45
Laney∘ smooth some test failures (upload/fwd a fix for tracker, some retries)15:45
Laney• Patch pilot, concentrated on http://reqorts.qa.ubuntu.com/reports/sponsoring/ubuntu-desktop.html - almost empty, help appreciated on the remainders.15:45
Laney• Notice Unity was deleting launcher icons when applications were upgraded, slight behaviour change in glib, fix that.15:45
Laney• Hopefully fix some trigger loops to avoid (dist-) upgrade failures15:45
Laney• Look at new default wallpaper, needs greyscale version15:45
Laney• Update gst-fluendo-mp3 in Debian to new release & drop 0.10 version15:45
Laney• Fix toolbar fullscreen hiding in eog, not uploaded since waiting for other fixes15:45
Laney15:45
larsuLaney made the glib top contributors list this cycle15:46
LaneyHAHA15:46
LaneyI think the barrier was pretty low15:46
willcookethanks Laney15:46
willcooke#topic larsu15:46
=== meetingology changed the topic of #ubuntu-desktop to: Home of the Desktop Team, https://wiki.ubuntu.com/DesktopTeam | For help or questions, try #ubuntu | Breaking your systems: GTK 3.16 - file bugs with tag "gtk316" | Desktop Team Weekly Meeting - 2015-09-22 | Current topic: larsu
Laneythanks larsu ♥15:46
seb128Laney, we should have a look at updating the fluendo plugin in the partner archive15:46
larsu- investigate and fix unity crasher (bug #1495173)15:46
larsu- unsuccessfully hunt down dbus-monitor bug (file it instead)15:46
larsu- [in progress] geonames: package, review, think about how to efficiently incorporate translations15:46
ubot5bug 1495173 in Unity "compiz crashed with SIGSEGV in g_bit_lock()" [High,Fix committed] https://launchpad.net/bugs/149517315:46
larsu- [in progress] unity-control-center: draft branch that works with geonames15:46
larsu- [in progress] eog: fix fullscreen toolbar; add traditional menubar15:46
larsu- [in progress] investigate making all output go to the journal and use gnome-logs15:46
larsu</lars>15:47
larsus/>/u>/15:47
willcookethanks larsu15:47
larsusorry about all the in progress stuff15:47
* larsu should wrap up15:47
willcookeseb128, larsu shall we talk fluendo now or at the end?15:47
seb128yes15:47
willcookes/larsu/laney15:47
seb128larsu, good list of in progress ;-)15:47
larsuwillcooke: you and didrocks....15:48
seb128larsu, btw I tried gnome-logs, it's nice but we loose e.g access to Xorg.0.log15:48
seb128so unsure if we should switch this cycle15:48
seb128or plan better for next cycle15:48
larsuyes, wait for next15:48
larsuit's too tight15:48
seb128k15:48
larsualso missing traditional menu and title bar15:48
seb128willcooke, sorry for interrupting ;-)15:48
willcookenw15:48
seb128larsu, well, that's ok15:48
seb128more annoying at the windows corners15:49
seb128but that's not specific to gnome-logs15:49
larsuseb128: blame Trevinho...15:49
seb128yeah, we need to put that bug on the lts list15:49
didrocks+115:49
larsuseb128: I'm all for doing the switch in gtk, but then we lose shadows15:49
seb128Trevinho, do you have a unity/compiz bug for that?15:49
Trevinhoseb128: mh, I think i read something, let me check15:50
seb128Trevinho, thanks15:50
seb128willcooke, I think we can continue with the meeting ;-)15:50
willcooke:)15:50
willcooke#topic qengho15:50
=== meetingology changed the topic of #ubuntu-desktop to: Home of the Desktop Team, https://wiki.ubuntu.com/DesktopTeam | For help or questions, try #ubuntu | Breaking your systems: GTK 3.16 - file bugs with tag "gtk316" | Desktop Team Weekly Meeting - 2015-09-22 | Current topic: qengho
qengho* fixed new armhf build problem with chromium-browser.15:50
qengho* l10n and Athlon SSE worries worked out, so giving to #security for release.15:50
qengho* learning how to work on firefox. Hopefully will prepare next release.15:50
qengho* looked for flights, but haven't picked one yet.15:50
qenghoEOF15:50
willcookethanks qengho15:51
willcooke#topic seb12815:51
=== meetingology changed the topic of #ubuntu-desktop to: Home of the Desktop Team, https://wiki.ubuntu.com/DesktopTeam | For help or questions, try #ubuntu | Breaking your systems: GTK 3.16 - file bugs with tag "gtk316" | Desktop Team Weekly Meeting - 2015-09-22 | Current topic: seb128
Laneyqengho: will it fix the autopkgtests?15:51
seb128• synced some bugfix updates from debian15:51
seb128• desktop updates (totem, webkitgtk, bluez)15:51
seb128• NEW reviews15:51
seb128• fixed u-c-c/bluez5 regression with ssp devices15:51
seb128• reviewed e.u.c top errors, reported some bugs on launchpad and upstream, pinged some people about some of the issues15:51
seb128• helped with the unity bugs triaging15:51
seb128• investigated software-center/unity integration issue15:51
seb128• looked at syslog warnings and reported some bugs with details about some of those15:51
seb128• reviewed packaging & tried new geonames library for larsu15:51
seb128• contributed to https://bugs.launchpad.net/ubuntu/+bugs?field.tag=rls-w-incoming ... if you have some free cycles please try to have a look and pick a few bugs15:51
seb128we have quite some annoying bugs on that list ^, would be nice to tackle some before wily if we can15:52
qenghoLaney: No, it won't. That's some dependency of a dependency of a dependency problem that I haven't tracked down.15:52
seb128willcooke, I wonder if we need another list for wily, or maybe milestone as well15:52
Laneyyou're meant to triage from the incoming list15:52
larsuseb128: thanks for looking at those warnings!15:52
seb128we dumped "would be nice to fix for the lts" on rls-w-incoming15:52
willcookeseb128, well I was adding to that list because X-incoming doesn't (didnt) exist15:52
seb128Laney, yeah, but we didn't use it this way, we used it as a lts todolist...15:53
didrocksor use priority to order the list?15:53
seb128Laney, see ^15:53
willcookeso I think if we can do anything on that list now, great - but then we can do a proper review at the sprint at take across the ones we want15:53
Laneyit can be both15:53
seb128didrocks, launchpad doesn't have priority, importance you mean?  but some "low" importance are easy to fix and should be done for wily imho15:53
Laneyto target bugs, nominate and remove the tag15:53
Laneyreview the rest of them to move to rls-x-incoming15:53
willcookeseb128, oh - right, what you said.  Misread it.15:54
didrocksseb128: importance, yeah, sorry, I think we could focus on the critical/high one. Doesn't prevent anyone having some free time to pick an easy "low" one15:54
seb128right15:54
seb128we should probably make a rls-x-incoming and move things there15:54
willcooke+115:54
seb128good15:54
seb128thanks15:54
Laneywhy?15:54
Laneyor, why before this release?15:55
seb128Laney, to have a better view of what is for wily15:55
seb128and what is not15:55
Laneythat view is supposed to be http://reqorts.qa.ubuntu.com/reports/rls-mgr/rls-w-tracking-bug-tasks.html15:55
seb128so we just need to accept nominations?15:55
Laneynominate and remove the tag15:56
* Laney just checked that old email15:56
seb128k15:56
seb128wfm15:56
seb128going to do that15:56
seb128Laney, thanks ;-)15:56
Laneythen we can see what's left after release to move over15:56
Laneyty15:56
seb128I didn't use that page for a while, I forgot how it was working15:56
Laneytry with some test one at first maybe15:56
Laneyto make sure it works15:56
seb128right15:56
willcookeok, that sounds good15:57
willcookeso keep adding to rls-w-incoming for now then15:57
seb128:-)15:58
* willcooke checks the sprint agenda15:58
willcookeadded15:59
willcooketo the top15:59
willcooke#topic sweet5hark15:59
=== meetingology changed the topic of #ubuntu-desktop to: Home of the Desktop Team, https://wiki.ubuntu.com/DesktopTeam | For help or questions, try #ubuntu | Breaking your systems: GTK 3.16 - file bugs with tag "gtk316" | Desktop Team Weekly Meeting - 2015-09-22 | Current topic: sweet5hark
willcooke- LibreOffice conference15:59
willcooke#topic TheMuso15:59
=== meetingology changed the topic of #ubuntu-desktop to: Home of the Desktop Team, https://wiki.ubuntu.com/DesktopTeam | For help or questions, try #ubuntu | Breaking your systems: GTK 3.16 - file bugs with tag "gtk316" | Desktop Team Weekly Meeting - 2015-09-22 | Current topic: TheMuso
willcooke* Finished  packaging up PulseAudio 6.99.2, which will become 7.0 shortly. It can be found in ppa:ubuntu-audio-dev/pulse-testing, amd64 and i386 only, armhf failed to build due to qemu related segfault.16:00
willcooke* Fleshed out a little more design for the accessibility profile blueprint, and finished drafting a blog post about it.16:00
willcooke* Started preparing at-spi and orca package updates to be put into the Ubuntu Accessibility dev PPA for general use. Updates will be provided from ppa:accessibility-dev/ppa for Orca, and the at-spi stack, for the LTS, vivid, and wily.16:00
willcooke#topic tkamppeter16:00
=== meetingology changed the topic of #ubuntu-desktop to: Home of the Desktop Team, https://wiki.ubuntu.com/DesktopTeam | For help or questions, try #ubuntu | Breaking your systems: GTK 3.16 - file bugs with tag "gtk316" | Desktop Team Weekly Meeting - 2015-09-22 | Current topic: tkamppeter
tkamppeter- hplip: Backported hpcups crash fix from upstream, use pkexec instead of gksu for hp-plugin-ubuntu16:00
tkamppeter- cups: USB-backend: Added new quirk rules for delayed closing of the backend for several older HP LaserJet printers16:00
tkamppeter- ippusbxd: Released 1.23 which logs all communication content as hex dump in debug mode and actually automatically terminates when the printer is disconnected or turned off.16:00
tkamppeter- Bugs.16:00
willcooketjamls tkamppeter16:00
willcookeerm, thanks tkamppeter16:01
willcooketkamppeter, did your parcel arrive yet?16:01
willcooke#topic Trevinho16:02
=== meetingology changed the topic of #ubuntu-desktop to: Home of the Desktop Team, https://wiki.ubuntu.com/DesktopTeam | For help or questions, try #ubuntu | Breaking your systems: GTK 3.16 - file bugs with tag "gtk316" | Desktop Team Weekly Meeting - 2015-09-22 | Current topic: Trevinho
Trevinho· Debugging webkit-gtk to get proper OSB theming. Speaking with upstream.16:02
Trevinho· Fixed unit tests on nux, now they run at every build.16:02
Trevinho· Fixed nux events handling on mouse-release when a new view was getting the mouse input16:02
Trevinho· Some tweaking on bugs management scripts16:02
Trevinho· Show-menus-tuning unity branch16:02
Trevinho· New unity/nux landing16:02
Trevinho· Add selective grab to compiz and use it in move and resize plugins16:02
Trevinho· New fixes for gsettings in libunity16:02
Trevinho· Prepared no-change SRU for compiz in vivid to get dbgsym's16:02
Trevinho· Finished work on new dash OSB (after fighting with nux bugs and oddities), needs cleanup16:02
Trevinho· Got sprint flights16:02
Trevinho· Reviews16:02
Trevinho16:02
willcooke:) thanks Trevinho16:03
willcooke#topic robert_ancell16:03
=== meetingology changed the topic of #ubuntu-desktop to: Home of the Desktop Team, https://wiki.ubuntu.com/DesktopTeam | For help or questions, try #ubuntu | Breaking your systems: GTK 3.16 - file bugs with tag "gtk316" | Desktop Team Weekly Meeting - 2015-09-22 | Current topic: robert_ancell
willcooke- Attended XDC 201516:03
willcooke- Released simple-scan 3.18.016:03
willcooke#topic any other business16:03
=== meetingology changed the topic of #ubuntu-desktop to: Home of the Desktop Team, https://wiki.ubuntu.com/DesktopTeam | For help or questions, try #ubuntu | Breaking your systems: GTK 3.16 - file bugs with tag "gtk316" | Desktop Team Weekly Meeting - 2015-09-22 | Current topic: any other business
willcookePlease get your travel auth in if you haven't already16:04
willcookeAnyone else?16:04
desrti'd like to share this:16:04
desrthttp://www.animalstown.com/animals/animals-x.php16:04
willcooke:D16:04
Laneydo I need to do auth to buy the train ticket myself?16:05
=== tkamppeter_ is now known as tkamppeter
willcookeLaney, nah, just expense it16:05
Laneyyessssssssssss16:05
willcookewell, that's what I'm doing16:05
andyrockdo we need to ask for a pass card to enter in the office?16:06
Laneynah16:06
andyrock\o/16:06
willcookeandyrock, you'll get a visitors badge16:06
* willcooke will make sure reception are expecting you16:06
desrtare we staying at citizenM?16:06
willcookedesrt, we are16:07
* Trevinho has an office pass :P16:07
desrtawesome16:07
willcookeI find the place a bit oppressive tbh16:07
desrtit's modern and not too fancy16:08
desrtand the plumbing works properly16:08
willcookehahaha16:08
willcookeand the windows don't rattle16:08
willcookeand there are less ghosts16:08
desrtthey also have international power plugs in the room16:08
desrtbecause, ya, sometimes hotel guests come from other countries16:09
Trevinhofor some reason nobody consider the Italian plug anyway (with ground in the middle)16:09
willcooke???16:09
Laneythey have a bed which you can sleep in any way round16:09
Laneyand colouful showers16:09
Laneyso it wins for me16:09
* willcooke likes the breakfast 16:09
desrtthey also have a nice downstairs area16:09
desrtyes.  breakfast is very very good there.16:09
desrtthe staff is also friendly and helpful16:10
willcookeI'm probably only going to stay over weds & thurs night and commute the rest of the time.  Because babies16:10
willcookeanyway - we can continue this after the meeting16:10
willcooke#endmeeting16:10
=== meetingology changed the topic of #ubuntu-desktop to: Home of the Desktop Team, https://wiki.ubuntu.com/DesktopTeam | For help or questions, try #ubuntu | Breaking your systems: GTK 3.16 - file bugs with tag "gtk316"
meetingologyMeeting ended Tue Sep 22 16:10:42 2015 UTC.16:10
meetingologyMinutes:        http://ubottu.com/meetingology/logs/ubuntu-desktop/2015/ubuntu-desktop.2015-09-22-15.30.moin.txt16:10
willcookeThanks a lot everyone16:10
desrttheir checkin system is awful, though16:10
andyrocksorry for being late16:10
desrtditto16:10
* desrt goes back into airplane mode again16:11
* Laney likes being called LANE IAIN/MR16:11
* desrt does not care for this16:11
Trevinhoslash mister?16:12
Laneysometimes IAINMR indeed16:12
didrocksthanks16:12
Laneythat form is usually on travel tickets16:12
Laney(amity travel are weird)16:12
Trevinhoah, sure... I'm MARCOMR...16:13
Trevinhoor, according to my fiber provider Marco Partita Iva (= Marco VAT)16:13
sethjCan anyone else reproduce this? https://bugs.launchpad.net/ubuntu/+source/unity/+bug/1286910 I see andyrock marked it as confirmed, so I'm not sure if that means he reproduced it or if he was just fixing the status.16:13
ubot5Ubuntu bug 1286910 in unity (Ubuntu) "Lock screen uses last active window's language instead of default keyboard language" [Medium,Confirmed]16:13
sethj(I cannot FWIW)16:14
andyrocksethj: well i can16:14
sethjoh hi anyrock, I thought you were out for the day. What steps did you take to reproduce it?16:15
seb128unsure of the specifics16:15
seb128but changing the lock screen layout doesn't change the session one16:15
seb128so you change, unlock and need to change again16:15
seb128which is annoying16:15
andyrocknope i was just late for the meeting :D16:15
andyrockwell i switch keyboard layout before locking the screen16:16
andyrocklet's say from italian to english16:17
andyrockand in the lockscreen the language is set to english16:17
andyrockit should be italian16:17
andyrockmaybe it is a slightly different bug16:18
sethjstrange. For me, if I changed the layout to Spanish (LA) and then lock the screen the greeter changes to the default language (en-us)16:18
andyrockI'll take a look16:19
sethjI'll give it a shot again a little later.16:19
andyrockcan be annoying16:19
andyrockif you don't notice that16:19
andyrockyou are going to blame the lockscreen becuase it does not accept your password16:19
andyrock:D16:19
andyrocksethj: btw I can't reproduce it 100%16:23
andyrockbut it happens16:23
Trevinhoandyrock: but isn't that more related to the indicator to unity itself?16:25
andyrockTrevinho: yep16:25
attenteseb128: there's this comment by ChrisTownsend: https://bugs.launchpad.net/unity/+bug/1324489/comments/116:28
ubot5Ubuntu bug 1324489 in Unity "Keyboard layout from lockscreen doesn't carry over to session" [Undecided,Opinion]16:28
attentewe did that by design and it would've made our lives easier to not implement the current behaviour at all16:29
seb128attente, ok, I can see why, that makes sense16:30
seb128I guess it's not really possible to carry for all usecases16:30
seb128though I would argue that people who want to have their word editor in english have the "by window layout tracking"16:30
seb128and don't change it for the session16:30
seb128hum17:00
seb128Trevinho, andyrock, hey, I just had issues today on my laptop and my test machine, going back from the greeter to my session made me enter my password twice (e.g the screenlock in the session didn't dismiss after entering the password on the unity-greeter), is that a known issue?17:01
seb128that's on wily17:01
andyrockseb128: nope first time I don't see this bug17:01
andyrockdoes it happen all the time?17:02
=== alan_g is now known as alan_g|EOD
seb128yes17:02
seb128just tried17:02
Laneyttyl!17:02
seb128Laney, have a good evening!17:03
Trevinhoseb128: it might happen in some cases, but it generally shouldnt... as we listen to a logind signal to unlock17:05
seb128Trevinho, in never happened before yesterday or today, I wonder if that's a regression17:05
seb128different topics but I've those warnings in my log17:08
seb128[system] Rejected send message, 9 matched rules; type="method_return", sender=":1.6" (uid=0 pid=874 comm="/usr/lib/accountsservice/accounts-daemon ") interface="(unset)" member="(unset)" error name="(unset)" requested_reply="0" destination=":1.66" (uid=1000 pid=2111 comm="/usr/lib/i386-linux-gnu/indicator-messages/indicat")17:09
seb128jdstrand, tedg, ^ do you know if that's known (wily desktop)17:09
jdstrandI've not seen it. mdeslaur, have you? ^17:11
seb128jdstrand, you don't have any of those in your auth.log?17:11
seb128weird, I just booted my test laptop and it has the same17:11
mdeslaurI don't run wily, so hven't seen it yet17:11
seb128(well after using xchat-gnome and receiving a message)17:11
jdstrandseb128: ah, I am still on vivid. I thought you were wondering if I've seen bugs, etc17:12
seb128jdstrand, any info that would be useful in a bug report? against what should that be reported?17:13
seb128reported bug #149859117:15
ubot5bug 1498591 in indicator-messages (Ubuntu) "Rejected send message auth.log errors" [Undecided,New] https://launchpad.net/bugs/149859117:15
jdstrandseb128: fyi, this seems to be a policy kit issue/bus policy issue (ie, it isn't an apparmor issue-- the message would look different)17:16
jdstrandthat is a wild guess17:16
seb128jdstrand, ok, I though it would be, but I was unsure if that was due to the dbus mediation17:16
seb128desrt, larsu, ^ do you know if those errors are more likely to be buggy client code? or a dbus issue?17:17
seb128jdstrand, thanks17:17
jdstrandyeah, none of those processes run under vonfinement17:17
jdstrandconfinement17:17
desrtseb128: that looks really weird17:26
desrti can't imagine this is buggy client code... looks rather like a broken security policy17:26
seb128https://bugs.launchpad.net/ubuntu/+source/indicator-bluetooth/+bug/1498594 seems a bit similar17:26
ubot5Ubuntu bug 1498594 in indicator-bluetooth (Ubuntu) "Rejected send message auth.log warnings" [High,New]17:27
desrtseems very similar17:27
desrtboth bluetooth and accountsservice are root17:27
desrtso they should be allowed to do anything as per dbus policy17:27
desrtread: likely apparmor fail17:28
* jdstrand notes that there isn't apparmor policy for those processes17:28
jdstrandnot to mention, the log message would look different17:29
desrtcould it rather be something about the processes on the other end to whom accountsservice is trying to send the message?17:29
desrtthe bluetooth one is different.  i make no statement about that.17:29
jdstrandapparmor would mediate that, but neither process should have a profile17:29
desrtbut rejecting a method_return from a root-owned process?  how could that possibly be happening?17:29
jdstrand(and the message would look different)17:30
tyhickspossibly due to the bus policy17:30
tyhicksthat is in /etc/dbus-1/system.conf17:30
desrt    <allow send_requested_reply="true" send_type="method_return"/>17:31
desrtfor user *17:31
tyhicksbut the logged message is not a requested_reply: requested_reply="0"17:31
* jdstrand wonders if that is overriden in /etc/dbus-1/system.d/* somewhere17:32
desrtlooking at this one: https://bugs.launchpad.net/ubuntu/+source/indicator-messages/+bug/149859117:32
ubot5Ubuntu bug 1498591 in indicator-messages (Ubuntu) "Rejected send message auth.log errors" [High,New]17:32
desrtthis is the really interesting one17:32
desrtRejected send message, 9 matched rules; type="method_return", sender=":1.6" (uid=0 pid=874 comm="/usr/lib/accountsservice/accounts-daemon ")17:32
desrt  destination=":1.66" (uid=1000 pid=2311 comm="/usr/lib/i386-linux-gnu/indicator-messages/indicat")"17:32
desrtone way or another, indicator-messages likely has a whacked apparmor or dbus policy17:34
desrtanother one here: https://bugs.launchpad.net/ubuntu/+source/indicator-applet/+bug/34651317:34
ubot5Ubuntu bug 346513 in indicator-applet (Ubuntu) ""dbus-daemon: Rejected send message" for indicator-applet spams /var/log/auth.log" [Medium,Fix released]17:34
jdstrand$ sudo ps -axZ|grep indicator-m17:39
jdstrandunconfined                       1677 ?        Ssl    0:00 /usr/lib/x86_64-linux-gnu/indicator-messages/indicator-messages-service17:39
jdstrandthat VM is slightly out of date, but it is unconfined17:39
jdstrandI am updating the VM, but I can say for sure we didn't not confine it17:39
jdstrand(as in the security team)17:40
tyhicksjdstrand: and the accounts-daemon is also unconfined?17:40
jdstrand$ sudo ps -axZ|grep accounts-d17:41
jdstrandunconfined                        508 ?        Ssl    0:00 /usr/lib/accountsservice/accounts-daemon17:41
desrtwhere are the current set of profiles maintained?17:41
desrtjust out of curiosity...17:41
jdstrandin the packages themselves17:41
desrtah17:41
jdstrandfor anything shipped by default that is17:41
jdstrandthere are the apparmor-profiles* packages, but they are not installed by default and complain mode only17:42
jdstrandand are*17:42
larsudesrt: where are you seeing this line? The only policy file I ave for accountsservice is in /etc/dbus-1/system.d/17:47
desrtlarsu: in the base policy in system.conf17:47
larsudesrt: isn't that overridden by a file in etc?17:48
tyhicks"Rejected send message" comes from a failed bus_client_policy_check_can_send() call, which is a access control check for the bus config and not the apparmor policy17:48
desrtlarsu: ya.. /etc/dbus-1/system.conf...17:48
larsudesrt: that file is empty for me...17:48
larsuwhat system are you on?17:48
desrtdebian.  assumed it was the same :)17:48
larsuhahahahaha17:48
larsuyou're funny17:48
desrtthanks for your understanding :)17:49
jdstrandnothing changed after upgrading17:50
larsudesrt: https://paste.debian.net/?show=31291217:50
larsudesrt: and I don't have that line in /usr/share/dbus-1/system.conf either17:55
larsuI wonder when this switch happened17:55
larsuwhich package installs that file for you?17:57
desrtdbus18:05
larsudbus ships a rule for accountsservice?18:06
desrtmy accountsservice file is same as yours18:06
larsuor is that a general rule?18:06
desrtdbus ships system.conf18:07
desrtthe accountsservice policy comes with accountsservice, of course18:07
larsusure sure18:07
larsuI have this as well, but in /usr/share18:07
desrtwhere?18:07
desrtyou're living in fancy future-world :)18:07
larsuya18:07
larsuthe one in /etc is still there, but it simply points to this one18:08
larsunow the question remains why this isn't applied :/18:08
larsuseb128: can you reprodcue this error message?18:25
seb128larsu, yes, you should also I guess, if you are on wily18:25
seb128it's in auth.log18:25
larsuI am, but I can't seem to18:25
seb128I've it on my test machine as well18:25
larsuhm ok I'll keep trying18:26
seb128I think it's "use xchat-gnome (or another app integrated to indicator-message) and receive a message"18:26
larsujust making sure :)18:26
seb128larsu, ping?18:26
larsuoh hehe18:27
larsuusing dbus calls directly :)18:27
seb128lol18:27
* larsu is in irssi but thanks18:27
seb128well, I wanted a pong with my name :p18:27
larsuwow indicator-messages is calling this method on the wrong interface18:27
larsuugh nevermind18:28
larsuit creates a proxy for that interface but then calls g_dbus_connection_call() directly (with all of the proxy's information except for the inteerface)18:29
larsu*sigh(18:29
seb128larsu, confirmed, the warnings are displayed when I receive an irc msg18:29
tyhickslarsu: has indicator-messages been doing that for a couple releases now?18:34
tyhicks(ubuntu releases, that is)18:34
larsutyhicks: it appears so - this code was first added in Sep 1418:35
tyhickslarsu: of last year?18:35
larsutyhicks: sorry, yes, September 201418:36
* tyhicks nods18:36
tyhicksnow that I think about it, I'm pretty sure those unrequested reply messages caused me a lot of headache while developing the apparmor mediation in dbus-daemon18:36
larsucalling the method on accountsservice manually doesn't trigger the warning for me18:36
jjohansenyeah those unrequested replies are evil18:37
jdstrandtyhicks: I was going to mention that, but didn't want to confuse the issue18:44
* qengho afk.18:57
chefpvanyone home?20:30
=== Guest61243 is now known as mfisch
=== mfisch is now known as Guest50715
=== Guest50715 is now known as mfisch

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!