2012/12/05/#ubuntu-desktop.txt

[00:39] <robru> cyphermox, ping
[00:57] <RAOF> Have I mentioned before how much I loathe SRU review of vala projects?
[00:58] <RAOF> GAH
[01:00] <TheMuso> RAOF: I can imagine, generally a vala + c diff correct?
[01:00] <RAOF> Indeed.
[01:00] <TheMuso> That puzzles me, the C files really don't need to be shipped.
[01:00] <TheMuso> At least thats what I thought.
[01:01] <RAOF> And the C is entirely impenetrable, but a near-irresistible lure for review eyeball.
[01:02] <RAOF> AIUI the recommended process for shipping vala is to generate the C at dist time, and not use vala at build time.
[01:02] <TheMuso> Ah ok.
[01:02] <TheMuso> Yeah the C is interesting to read.
[01:02] <RAOF> This seems about as sensible to me as shipping the assembler generated by gcc as the source form.
[01:02] <TheMuso> hahaha
[01:03] <TheMuso> I agree.
[01:04] <RAOF> Ok. There's an evening ping for kenvandine pending.
=== rickspencer3_ is now known as rickspencer3
[01:39] <ritz> robert_ancell morning
[01:39] <robert_ancell> ritz, hello
[01:40] <ritz> robert_ancell https://bugs.launchpad.net/lightdm/+bug/1071870 any thoughts ?
[01:40] <ubot2> Launchpad bug 1071870 in Light Display Manager "Dont verify the presence of .face file " [Undecided,New]
[01:41] <ritz> this would be seen when accountsservice is disabled
[01:41] <robert_ancell> ritz, right, why not just disable the user list in this case?
[01:43] <robert_ancell> ritz, the problem of removing this code is then you can't reliably let a greeter know where the user icon is
[01:43] <robert_ancell> which is something that accounts service solved
[01:47] <ritz> robert_ancell hmm, is it really required that the greeter know the user icon reliably ?
[01:47] <veebers> Hi all, after a dist-upgrade when I ssh into my test machine, I don't seem to be able to connect to the SessionBus as expected.
[01:47] <robert_ancell> ritz, yes, otherwise how can it display the icon?
[01:47] <ritz> I am new to this, and not sure.
[01:47] <robert_ancell> ritz, the problem was there wasn't a consistent name for the icon
[01:48] <veebers> a terminal on the machine itself works as expected, but not over ssh. using a python shell to test, when creating a SessionBus object, I seem to be getting something that's not the session bus back (i.e.  no com.canonical.Unity)
[01:48] <veebers> would anyone be able to shed some light on this?
[01:51] <ritz_> robert_ancell sorry, was dc.
[01:52] <ritz_> robert_ancell how about lazy loadin this property ?
[01:52] <ritz_> only when this property is queried for
[01:52] <ritz_> when AccountsService is disabled
[01:53] <robert_ancell> ritz_, yes, that would be fine
[01:54] <ritz_> would this be an acceptable fix ?
[01:54] <thumper> hi folks
[01:54] <ritz_> cool. thanks. will look this up
[01:55] <thumper> does anyone know when glib reads the environment for determining the XDG path for finding desktop files?
[01:55] <thumper> I'm wanting to control the path for some unit tests
[01:55] <thumper> do I need to modify the path before glib is initialized?
[01:55] <thumper> or does it look at the environment each time?
[01:55] <robert_ancell> ritz_, yes, I've updated the bug to say that
[01:56] <ritz_> oh, I had not refreshed ther page
[01:57] <robert_ancell> ritz_, I only did it a few second ago :)
[02:36] <mdeslaur> duflu: so, apparently you're the person I need to bribe to get bug 1037164 fixed...
[02:36] <ubot2> Launchpad bug 1037164 in compiz (Ubuntu Quantal) "Clicking on snapped windows in a different workspace produce unexpected results" [Medium,Triaged] https://launchpad.net/bugs/1037164
[02:38] <mdeslaur> duflu: do you like scotch? :)
[02:39] <duflu> mdeslaur: One the phone
[02:39] <duflu> -e
[02:53] <duflu> mdeslaur: You would have to convince the squad lead and/or JohnLea and Olli to get something like that on the priority lists. Yes, I like scotch but seem to have plenty in reserve right now :)
[02:54] <mdeslaur> d'oh :)
[02:54] <mdeslaur> this stupid bug has been driving me insane since natty
[02:55] <mdeslaur> duflu: who's the squad lead?
[02:55] <duflu> mdeslaur: What would be better would be to encourage more community contribution of fixes. We love them
[02:55] <duflu> mdeslaur: bregma
[02:55] <duflu> mdeslaur: In the mean time, you can probably just turn off the Snapping Windows plugin in CCSM. Does that help?
[02:56] <mdeslaur> uh, good question, let me check
[02:57] <mdeslaur> duflu: unfortunately not, even with that plugin disabled, the window moves half-way into the next workspace
[02:58] <duflu> mdeslaur: OK, thanks. And sorry. Using quantal?
[02:58] <mdeslaur> yeah
[02:59] <duflu> mdeslaur: I'll keep in mind for when I don't feel like paying attention to the priority lists :)
[03:00] <mdeslaur> duflu: hehe, thanks
[03:00] <cyphermox> robru: pong?
[03:00] <cyphermox> still around?
[03:00] <robru> cyphermox, oh hey, I am just heading out the door actually
[03:00] <robru> cyphermox, I sent you an email
[03:00] <cyphermox> ok!
[03:03] <robru> g'night!
=== m_conley is now known as m_conley_away
=== Ursinha is now known as Ursinha-afk
=== Ursinha-afk is now known as Ursinha
[06:17] <didrocks> good morning!
[06:26] <RAOF> didrocks: Yo!
[06:28] <didrocks> hey RAOF! double bath time soon? :)
[06:28] <RAOF> Yeah, I guess so.
[06:28] <RAOF> Zoë's asleep at the moment, though.
[06:28] <didrocks> otherwise, how are things going? :)
[06:29] <RAOF> Would you be askable about the Precise Unity 2d SRU?
[06:30] <didrocks> RAOF: I think it's Mirv only, it's quite far from my head ;)
[06:30] <didrocks> mirv should be around btw :)
[06:30] <didrocks> RAOF: I remember having to ask for some reverts, that's it
[06:30] <didrocks> and they are done IIRC before I uploaded that
[06:30] <RAOF> I think you might have missed some :)
[06:31] <didrocks> urgh? what do you see as irrelevant?
[06:31] <didrocks> https://code.launchpad.net/~unity-2d-team/unity-2d/trunk
[06:31] <didrocks> http://bazaar.launchpad.net/~unity-2d-team/unity-2d/trunk/revision/1142
[06:31] <RAOF> Or, more to the point, there seems to be some refactoring in there that may be fixing a bug, but the bug that it might be fixing isn't filed in Launchpad.
[06:32] <didrocks> as you can see, I asked for more than on revert :)
[06:32] <didrocks> but just to check I didn't sponsor the wrong .changes?
[06:32] <didrocks> you don't see those, right?
[06:33] <RAOF> No, I don't see those.
[06:33] <didrocks> ok, at least, we are on a common ground :)
[06:35] <RAOF> Some of the ‘bugs’ that it fixes seem a bit shaky for an SRU - there seemed to be a couple of ‘change this behaviour to be better’ bugs. The change that I'm least sure of is related to https://bugs.launchpad.net/unity-2d/+bug/977262 , though.
[06:35] <ubot2> Launchpad bug 977262 in unity-2d (Ubuntu Precise) "Unity 2D doesn't load 24 bit icons" [Undecided,Triaged]
[06:36] <Mirv> RAOF: hello
[06:37] <Mirv> RAOF: although sil2100 was the one who worked on that, I just pinged about it :)
[06:37] <didrocks> Mirv: can you see that with RAOF? I'll read the scrollback, but I'm on the unity release :)
[06:37] <didrocks> RAOF: ah this one, we had a big debate about it, but design convinced me in the end
[06:37] <didrocks> RAOF: it's more "loading the good quality", so enhancement, which is seen as a "bug" for design
[06:38] <didrocks> (to load crappy icons)
[06:38] <didrocks> RAOF: maybe it's part of those where we decided to ask charline about those?
[06:38] <didrocks> (also, we only have ? for some of them when they only provide a 24 bits icon)
[06:38] <didrocks> and this can be seen as a bug to be fixed
[06:39] <RAOF> No, I'm not concerned with that bug being SRUable; it seems reasaonable to fix.
[06:39] <didrocks> ah, it's about the code itself?
[06:40] <RAOF> It's just not clear to me what part of that bug is actually being fixed in 5.14.0-0ubuntu1, and the code doesn't help much.
[06:40] <Mirv> the commit message is [lib] Support 24 bit pixbuf to QImage conversion (e.g. for kmag icon in Launcher)
[06:41] <Mirv> and additionally another commmit that fixes the colormapping
[06:41] <Mirv> (from the merges linked in the bug report)
[06:43] <RAOF> The diff between what's currently in precise-proposed and what's in 5.14.0 - http://launchpadlibrarian.net/121657028/unity-2d_5.12.0-0ubuntu1.2_5.14.0-0ubuntu1.diff.gz - only seems to be changing it so that 3-channel pixmaps don't get RGB-byteswapped.
[06:44] <RAOF> Although it does it in an SRU-unfriendly way by renaming variables and changing indentation :)
[06:46] <RAOF> Is that the change that's expected? It does not appear to match the bug description - AFAICT it's *not* adding support for 24bit pixmaps, just changing the handling of the colour channels.
[06:48] <Mirv> RAOF: good point, it was already fixed in 5.12 and only the last commit (https://code.launchpad.net/~aacid/unity-2d/more_24_bit_stuff/+merge/105797) is now in 5.14
[06:48] <Mirv> ie. loading more types of 24bit pixmaps correctly
[06:49] <Mirv> so indeed there is no clear bug decribing which icons do not load correctly, aside from the automated test
[06:51] <Mirv> terrible diff
[06:51] <RAOF> *Yes*
[06:52] <RAOF> Could I therefore please have a bug for the actual problem being fixed? ☺
[06:53] <RAOF> Because as it stands it's highly likely that if I accept the SRU no-one is going to know how to validate that fix.
[06:54] <RAOF> I'll reject the existing upload. If you upload with the changelog pointing to the new, accurate, bug I'll accept it immediately.
[06:54]  * RAOF → Zoë bath
[06:55] <Mirv> I'll ask greyback when he wakes up
[07:01] <didrocks> pitti: when you have some time, can please have a look at bumping https://launchpad.net/~ubuntu-unity/+archive/daily-build/+build/4037566? It seems we will miss daily builds again otherwise
[07:01]  * didrocks now wonders what to do about the unity stack, we wanted to release it today
[07:01] <didrocks> but if I launch a build of it, powerpc won't ever be built in time
[07:26] <didrocks> (thanks to whoever bumped it)
[08:42] <pitti> Bonjour mes amis
[08:42] <didrocks> hey pitti!
[08:42] <didrocks> how is the weather in London?
[08:42]  * pitti gets the score bumping boot
[08:42] <pitti> didrocks: rainy :)
[08:43] <didrocks> ah, real UK then!
[08:43] <pitti> didrocks: well, that thing already built
[08:43] <didrocks> pitti: yeah, see my 08:26:04 didrocks | (thanks to whoever bumped it)
[09:27] <duflu> smspillaz: ping
[09:37] <ogra_> didrocks, hmm, so trying to install from the PPA removes the unity package on my nexus 7 ... and i end up with just nautilus on my desktop
[09:37] <didrocks> ogra_: yeah, previous build because of a build-dep failed if you look at the ppa :)
[09:37] <didrocks> ogra_: there is just a rebuild under progress, so armhf should be fine
[09:38] <didrocks> (it was an arch async)
[09:38] <ogra_> ok, i'll try again once thats done
[09:39] <didrocks> ogra_: yeah, builders seems happy and at most will start in an hour on powerpc
[09:43] <RAOF> Hm. When's Ken going to be on?
[09:44] <didrocks> probably still in 2-4 hours, you would be in bed I guess ;)
=== vrruiz_ is now known as rvr_
[09:58] <seb128> didrocks, RAOF: rather 4 hours than 2 ... what's the question/issue?
[09:58] <didrocks> I guess it's in one of the webapps SRU
[09:59] <RAOF> seb128: He wanted to prod me about an SRU, I wanted to get some clarification about it.
[09:59] <seb128> RAOF, which one? I can maybe help providing the infos...
[10:00] <RAOF> seb128: gnome-control-center-signon - there seems to be some unrelated changes in cc-credentials-accounts-model.vala
[10:08] <RAOF> seb128: Know about that one?
[10:08] <duflu> RAOF: ping
[10:08] <duflu> Ah, there you are
[10:09] <RAOF> duflu: What's cookin'?
[10:10] <duflu> RAOF, tvoss: It's been a few months since I did any serious EGL/GLES work, but I'm left scratching my head in disbelief there is no explicit extension for vblank syncing.
[10:10] <duflu> Is it assumed to be implied by the driver, which hopefully does page flipping?
[10:10] <RAOF> duflu: SwapBuffers + OML_sync?
[10:10]  * duflu looks
[10:11] <RAOF> duflu: http://www.opengl.org/registry/specs/OML/glx_sync_control.txt
[10:11] <seb128> RAOF, looking, but no, seems an undocumented change ... I guess Ken should do another upload with it documented or rolled out
[10:11] <duflu> RAOF: EGL, not GLX :)
[10:12] <duflu> How the hell did I miss that in the GLES work? Mebbe it was always implied as part of swapbuffers if implemented as flipping
[10:13] <duflu> I ask because the Nexus 7 is tearing, and I realized Compiz has no explicit extension support for avoiding that under EGL
[10:15] <RAOF> duflu: set eglSwapInterval to 1 (or more) + eglSwapBuffers?
[10:15] <RAOF> Although the default eglSwapInterval should be 1
[10:16] <duflu> RAOF: OK, I'm clearly blind. Feel free to ignore my stupidity. Still worried about the Nexus though
[10:16] <RAOF> ( http://www.khronos.org/registry/egl/specs/eglspec.1.4.20110406.pdf )
[10:18]  * duflu needs more sleep and/or coffee. Maybe not together
[10:28] <xnox> So we currently have bug 1060327
[10:28] <ubot2> Launchpad bug 1060327 in compiz (Ubuntu) "compiz crashed with SIGSEGV in compiz::opengl::bindTexImageGLX() from TfpTexture::bindTexImage() from ... from GLTexture::bindPixmapToTexture() from DecorTexture::DecorTexture()" [High,Confirmed] https://launchpad.net/bugs/1060327
[10:28] <xnox> is this being work on?
[10:29] <didrocks> xnox: can we avoid having 2 different people asking the same question on 2 different channels?
[10:29] <didrocks> xnox: so let me copy the question on #ubuntu-release
[10:29] <didrocks> 11:28:24 didrocks | gema_: is this crash on top of errors.ubuntu.com
[10:29] <xnox> didrocks: sure which channel should we talk on?
[10:29] <didrocks> here is fine, but just pick one :)
[10:50] <RoyK> Why is the 32bit version still flagged as 'preferred'? I mean - there's hardly any 32bit hardware out there, and 32bit isn't very efficient on 4GB RAM or more, and most machines these days have that
[10:50] <seb128> RoyK, why is it not efficient if you have > 4GB ram?
[10:51] <mlankhorst> it really isn't..
[10:51] <mlankhorst> forces kernel to use indirect addressing for high memory
[10:51] <mlankhorst> although 64-bit kernel with 32-bit userspace is fine most of  the time..
[10:52] <RoyK> seb128: it uses PAE for anything >2GB, and PAE isn't very efficient comparing to native 64bit addressing
[10:53] <mlankhorst> RoyK: >2gb? I thought anything above 1 gb would be highmem already with the 3/1g kernel split..
[10:54] <seb128> RoyK, in any case read https://lists.ubuntu.com/archives/ubuntu-devel/2012-April/035054.html on the topic
[10:54] <RoyK> mlankhorst: yeah
[10:55] <seb128> RoyK, https://lists.ubuntu.com/archives/ubuntu-devel/2012-April/035088.html was the conclusion
[10:57] <mlankhorst> I thought it was impossible to boot the 32-bit image on true 32-bit hardware because pae is enabled in the default image? then again I no longer have true 32-bit hardware to test..
[10:58] <ogra_> mlankhorst, thats only true for non-PAE capable 32bit hardware (i.e. rather old CPUs like P II, some pentium M's etc)
[10:59] <ogra_> recent 32bit systems should still work fine
[11:05] <didrocks> ogra_: compiz and unity are still building on armhf, will tell you once done
[11:06] <ogra_> didrocks, i have the tab open in front of me :)
[11:06] <ogra_> well, on my most left screen ... not actually in front :)
[11:06] <didrocks> ogra_: staring! :)
[11:07] <didrocks> ogra_: but feel free to install them in the desktop
[11:07] <didrocks> ogra_: and shout if anything bad happens :)
[11:08] <ogra_> nah, i wont taint my precise desktop install :P
[11:09] <didrocks> ogra_: hum, I won't comment :p
[11:09] <ogra_> took me long enough to get a 3 monitor setup working with full 5760x1920 in a usable way
[11:10] <ogra_> i wont fiddle with that machine :)
[11:10] <ogra_> i might test on my chromebook raring install though
[11:10] <ogra_> but thats only arm as well :P
[11:10] <didrocks> héhé
[11:40] <codfather>  Good morning from the UK. I would like some help with a question on 12.10 and dnsmasq. Does anyone know where the upstream DNS servers are now stored? They are not in /etc/resolv.conf or /var/run/nm-dns-dnsmasq.conf.
[11:41] <codfather> any information or pointers to documentation gratefully received - thanks in advance
[12:07]  * ogra-cb hugs didrocks 
[12:08] <ogra-cb> didrocks, all PPA stuff seems to work fine on the nexus
[12:08] <ogra-cb> didrocks, the workspace switcher icon is missing for me though ...
[12:08] <ogra-cb> shows a "broken icon" image
[12:22] <seb128> GunnarHj, hey, I'm reassigning those merge requests, I'm not working on xubuntu or ubuntustudio and I prefer the people from those flavors to decide when is the right time to change their seed for their images
[12:25] <GunnarHj> seb128: Ok. Please note, though, that they include language-selector-gnome in their seeds, and im-config is a dependency of language-selector-gnome.
[12:26] <seb128> GunnarHj, well, still the people in charge of those images should be the ones doing the merges
[12:27] <GunnarHj> seb128: I see.
[12:27] <seb128> GunnarHj, I updated the reviewer for that, but I will make sure things get coordinated so no worry
[12:27] <GunnarHj> seb128: Great.
[12:28] <GunnarHj> seb128: Sent a reminder to Aron, btw, but haven't heard from him yet.
[12:28] <seb128> GunnarHj, ok, I think I will just merge your changes and we can do a follow up update if he has comments or issues
[12:29] <GunnarHj> seb128: Sounds good. The function we discussed will still be changed sooner or later. There is a FIXME remark in the file.
[12:46] <didrocks> 13:44:20   didrocks | ogra-cb: \o/
[12:46] <didrocks> 13:44:30   didrocks | ogra-cb: broken icon?
[12:46] <didrocks> 13:44:45   didrocks | ogra-cb: the icon should shows on which ws you are on
[12:46]  * didrocks hugs ogra-cb back
[12:46] <ogra-cb> didrocks, well, seems it doesnt find any icon at all ... WS switcher works though
[12:47] <ogra-cb> i just get the question mark icon for it
[12:48] <didrocks> interesting
[12:49] <didrocks> ogra-cb: you do have latest unity-asset-pool isn't it?
[12:50] <ogra-cb> it has "daily" in the version
[12:50] <didrocks> so, it should be fine, weird
[12:50] <didrocks> well, let's not block on that for now
[12:50] <ogra-cb> 0.8.24daily12.12.03-0ubuntu1
[12:50] <didrocks> I think you will agree :)
[12:50] <didrocks> oh
[12:50] <didrocks> 12.12.03?
[12:50] <ogra-cb> nah, definitely not a blocker
[12:50] <didrocks> you should have 12.12.05
[12:50] <didrocks> you did dist-upgrade?
[12:50] <ogra-cb> ah, intresting
[12:51] <ogra-cb> yes
[12:51] <didrocks> apt-cache policy is still telling there is only .03 for you?
[12:51] <didrocks> (2 days! you are so old!!! ;))
[12:52] <ogra-cb> oh, a new dist-upgrade call wants to upgrade 31 packages (many unity ones among them)
[12:52] <ogra-cb> let me run that :)
[12:52] <ogra-cb> yeah, seems that was caused by my first try where not everything was ready
[12:53]  * ogra-cb just seems to get 12.12.05 for everything now
[12:53]  * ogra-cb prays that doesnt regress
[12:53] <didrocks> interesting :)
[12:53] <didrocks> ogra-cb: keep me posted!
[12:53] <ogra-cb> will do
[12:58] <ogra-cb> didrocks, 12.12.05 works fine, workspace switcher has an icon too now
[12:59] <ogra-cb> ready to upload i'd say
[12:59] <ogra-cb> at least from an arm perspective
[12:59] <didrocks> ogra-cb: waiting on powerpc to finish and will do :)
[12:59] <didrocks> good good news \o/
[12:59] <ogra-cb> yeah, thanks so much for the hard work
[13:01] <didrocks> no worry :)
=== tkamppeter_ is now known as tkamppeter
[13:19] <psivaa> didrocks: just in case you did not notice in release channel, i could not verify the ppa, in the live session, persistence mode in vms with kvm is not working. but pkill -9 X after the black blank screen brings up the live session properly even without installing the ppa
[13:20] <didrocks> psivaa: yeah, I did notice, I think, let's try with tommorrow's daily
[13:20] <didrocks> psivaa: it will get the new stack
[13:20] <psivaa> didrocks: ack thanks
[13:20] <didrocks> thanks to you :)
[13:32] <didrocks> only the lens shopping is holding back now! waiting for powerpc to publish!
[13:36] <ogra-cb> go go go !
[13:37] <didrocks> ogra-cb: my checker is still saying it's not published (checking every 5 times ;))
[13:37] <ogra-cb> :)
[13:37] <didrocks> then, as there are packaging change, I need to ack them ;)
[13:39] <didrocks> publisher phase now \o/
[13:40] <didrocks> btw, you can follow that live on https://jenkins.qa.ubuntu.com/view/cu2d/view/Unity%20Head/job/cu2d-unity-head-3.0publish/lastSuccessfulBuild/ :)
[13:41] <ogra-cb> heh
[13:42] <didrocks> (it's blocked because of the package_changes, which is to expect as we converted them all to dh9 ;))
[13:42]  * didrocks makes a second review
[14:04] <didrocks> ok, all reviewed
[14:04] <didrocks> now, let's "force" publishing
[14:05] <ogra_> you publish into the archive out of the PPA ?
[14:05] <didrocks> yep
[14:05] <ogra_> wow
[14:06] <didrocks> now, it's proposing all the packaging changes back to upstream :)
[14:10] <didrocks> and done \o/ https://jenkins.qa.ubuntu.com/view/cu2d/view/Unity%20Head/job/cu2d-unity-head-3.0publish/17/console
[14:10] <didrocks> (all will be merged automatically by the upstream merger)
[14:10] <ogra_> so next publisher run the raring archive will have it ?
[14:10] <didrocks> just waiting for a distro sync which is… now!
[14:10] <ogra_> yay
[14:11] <didrocks> 2012-12-05 14:11:07,690 INFO Found packagelist_rsync_cu2d-unity-head
[14:11] <didrocks> \o/
[14:11] <dbarth_> desrt: ping?
[14:11] <didrocks> 2012-12-05 14:11:09,594 ERROR The project unity-scope-gdrive is not in the allowed stack to be copied to distro. Rejecting.
[14:12] <didrocks> great, I was trying to not refresh the blacklist to test it :)
[14:12] <didrocks> let's free that one as well
[14:15] <didrocks> and done :)
[14:16] <didrocks> http://cdn.memegenerator.net/instances/400x/31326392.jpg
[14:32] <mspencer> mpt: I have another question regarding Contributor Console.
[14:32] <mpt> mspencer, good
[14:32] <mpt> :-)
[14:35] <mspencer> mpt: I have a couple of ideas for features to add to the Bugs panel. Do you want me to file them as bugs even though Contributor Console hasn't been released yet or do you just want me to talk to you directly about new ideas?
[14:35] <mpt> mspencer, either is fine by me
[14:36] <desrt> dbarth_: hi
[14:37] <mspencer> mpt: Okay. What do you think about adding the ability to see bugs that have been reported and the ability to see bugs that have been saved for reporting later?
[14:37] <mspencer> mpt: Then maybe have a button to get the code for the package a bug is filed for?
[14:38] <cyphermox> good morning
[14:38] <desrt> cyphermox: bonjour
[14:38] <cyphermox> desrt: sup
[14:38] <desrt> not my brain, yet
[14:39] <cyphermox> meh, it's before 11am.
[14:39] <cyphermox> nobody's up that early.
[14:40] <dbarth_> desrt: hi, i have a questoin about gsettings schema introspection
[14:41] <dbarth_> desrt: is there an api to determine the type of a key?
[14:41] <mpt> mspencer, is "the ability to see bugs that have been reported" the same as the "Show Bug Reports" button, or something different?
[14:41] <desrt> dbarth_: yes.  two, in fact.
[14:41] <dbarth_> desrt: or am i supposed to be parsing the xml
[14:41] <dbarth_> ah, nice
[14:42] <desrt> dbarth_: i have to wonder why you would want to do such a thing, though
[14:42] <dbarth_> desrt: i'm trying to enumerate unity properties and sync them with u1
[14:42] <mspencer> mpt: Different. I meant bugs that had been reported by the user.
[14:43] <dbarth_> desrt: so the code has to be able to determine the type of keys at runtime
[14:43] <desrt> doesn't the compizconfig abstraction have this information in it?
[14:43] <mpt> mspencer, oh, I see
[14:43] <desrt> or are you doing this from outside of compiz?
[14:43] <dbarth_> desrt: ie, i tell it: go enumerate com.canonical.Unity.Launcher and turn that into a hash that i can send to the u1db api
[14:43] <desrt> oh.  u*1*.  i saw 'ui' :)
[14:44] <desrt> dbarth_: right.  okay.
[14:44] <dbarth_> right, Ubuntu One
[14:44] <desrt> dbarth_: so first you will want to get a list of all the keys, i guess
[14:44] <dbarth_> i have that yup
[14:44] <desrt> that's g_settings_list_keys()
[14:44] <dbarth_> yup
[14:44] <desrt> then you can just get each value with g_settings_get_value()
[14:44] <desrt> and then you can find out its type in a lot of ways
[14:44] <desrt> depending on how you want to store it....
[14:44] <mpt> mspencer, that seems like a good idea, indeed :-)
[14:44] <dbarth_> hmm, i thought i had tried that
[14:44] <desrt> the easiest thing to do is probably g_variant_print()
[14:44] <dbarth_> it returns a gvariant, right
[14:45] <desrt> then you will get a string that you can parse on the other side
[14:45] <dbarth_> right
[14:45] <dbarth_> hmm, i remember getting an error with variants, so i tried to use native type instead
[14:45] <desrt> but you can also query the type if you feel like handling it manually
[14:45] <desrt> g_variant_get_type()
[14:45] <dbarth_> maybe that's the python bindings
[14:45] <dbarth_> ok, i will try that again
[14:45] <dbarth_> desrt: thanks
[14:45] <desrt> the python bindings should have OK support for variants
[14:46] <desrt> since about two years ago
[14:48] <mpt> mspencer, and seeing bugs that were saved earlier is in that bug report that you originally saw, right?
[14:48]  * mpt tries to find it
[14:49] <mpt> mspencer, bug 657275
[14:49] <ubot2> Launchpad bug 657275 in apport (Ubuntu) "ubuntu-bug should save reports offline automatically rather than giving a cryptic error message" [Wishlist,In progress] https://launchpad.net/bugs/657275
[14:49] <mspencer> mpt: Yes.
[14:49] <mpt> mspencer, so basically you'd like a design for those two?
[14:49] <mspencer> mpt: Yes, that would be great.
[14:50] <dbarth_> desrt: g_variant_get_type_string does the trink
[14:50] <desrt> dbarth_: honestly, i think you should just go straight for g_variant_print(v, TRUE)
[14:51] <desrt> if you want to step up your game a little, you could say FALSE and re-provide the type information on the other side
[14:51] <mspencer> mpt: About filing bugs that are feature requests, if I or someone else files one, I'm going to assign it to you, then you design it, then you unassign yourself, right?
[14:52] <mpt> mspencer, right, I've just done that with bug 657275 and bug 1086825
[14:52] <ubot2> Launchpad bug 657275 in apport (Ubuntu) "ubuntu-bug should save reports offline automatically rather than giving a cryptic error message" [Wishlist,In progress] https://launchpad.net/bugs/657275
[14:52] <ubot2> Launchpad bug 1086825 in Contributor Console "Can't easily access bug reports I'm involved in" [Undecided,New] https://launchpad.net/bugs/1086825
[14:52] <mspencer> mpt: Okay, thanks.
[14:53] <mpt> cool
[14:54] <dbarth_> desrt: ok; getting there
[15:06] <codfather> I would like some help with a question on 12.10 and dnsmasq. Does anyone know where the upstream DNS servers are now stored? They are not in /etc/resolv.conf or /var/run/nm-dns-dnsmasq.conf.any information or pointers to documentation gratefully received - thanks in advance
[15:26] <seb128> cyphermox, hey, I'm rejecting evolution-indicator from NEW, src/xutils.c,h are LGPL but debian/copyright only list GPL, we also need a COPYING.LGPL shipped in the source ... can you get those fixed and reupload?
[15:26] <cyphermox> yup
[15:26] <seb128> cyphermox, thanks
[15:27] <xclaesse> how do I take screenshot of a single window with unity? alt-printscreen takes a shot of the dash because alt open it
[15:28] <seb128> xclaesse, hum, that's not supposed to happen... only a tap on alt is supposed to open the hud (the dash interface to browse menu)
[15:29] <seb128> xclaesse, in practice most people re-map the hud away from alt in system settings
[15:30] <xclaesse> seb128, what actually happens: 1) press alt 2) press printscreen 3) release printscreen (shot is not yet take) 4) release alt and dash opens then shot is taken
[15:30] <seb128> xclaesse, for some reason the take screenshot is ctrl-printscreen here
[15:31] <seb128> I wonder if I remapped it :p
[15:37] <chrisccoulson> hmmm, why didn't i discover JS proxies before!
[15:39] <mvo> mterry: is it intentional that update-manager -d only shows information about the new release after all updates are installed? looking at the code it is, but it will be confusing for some users (well, me at least :)
[15:39] <mterry> mvo, it was intentional since I thought that's the recommended way to upgrade (i.e. to make sure that users have latest update manager and all that
[15:41] <mvo> mterry: right, its generally a good idea I was just confused that there is no indication that the new release is available in the UI and only after applying the missing updates it was there. not a big deal, just wanted to mention it
[15:41] <mterry> mvo, yeah I'm fine with that.  I think that's how the spec is written too.
[15:43] <mvo> ok
[15:54] <didrocks> mterry: hey, how are you?
[15:55] <mterry> didrocks, good.  I've been running the PPA.  No obvious problems
[15:55] <didrocks> mterry: it's in the distro right now \o/
[15:55] <didrocks> so we got our first daily release (but without autopilot, just dogfooding)
[15:55] <kenvandine> yay!
[15:55] <didrocks> :)
[15:56] <didrocks> mterry: so, now that we are back to a normal life, I let the merge reviews under your vigilance :)
[15:56] <mterry> didrocks, hah, OK
[15:56] <didrocks> mterry: FYI, I fixes some mismatch in the build-deps
[15:56] <didrocks> mterry: however, nux has a lot of rdepends in the .pc file not listed in its -dev
[15:56] <didrocks> (I blame the merging of multiple .so)
[15:56] <mterry> didrocks, I noticed you also did a set of changelog updates a while back.  Did I miss some things or were there just parallel uploads to raring?
[15:57] <didrocks> mterry: some were parallel uploads, but some were missing :)
[15:57] <mterry> didrocks, ah well.  Thanks!
[15:57] <didrocks> mterry: not a biggie, now that's done, will be easier to catch I guess!
[15:57] <didrocks> mterry: I plan to discuss later this week about how the daily process work in depth
[15:57] <didrocks> mterry: for the nux thingy, just ping me (or I'll watch)
[15:58] <didrocks> mterry: oh, and btw http://pad.ubuntu.com/ubuntu-unity is resolved (well almost, some are still not merged)
[15:58] <didrocks> can you track them?
[15:58] <didrocks> and note everything not compliant :)
[16:00] <mterry> didrocks, what's the nux thingy again?
[16:00] <didrocks> mterry: oh, if you look at the -dev
[16:01] <mterry> didrocks, OK
[16:01] <didrocks> it doesnt dep on every requirement of the .pc
[16:01] <didrocks> doesn't*
=== m_conley_away is now known as m_conley
[16:01] <didrocks> so I guess a sanity check (even if we are the only nux project :p)
[16:01] <mterry> didrocks, OK, I'll try to keep on top of branches without tests for ya
[16:01] <didrocks> thank you!
[16:01] <mterry> didrocks, and I'll do a pass on nux and try to propose a review
[16:02] <didrocks> excellent, thanks a lot :)
[16:02] <mterry> kenvandine, btw, you should be able to run deja-dup's trunk tests again
[16:10] <kenvandine> mterry, cool, i'll try
[16:11] <didrocks> mterry: kenvandine: cyphermox: I think I'll give you a short introduction to this daily-build thingy, do you prefer before those holidays or after?
[16:11] <kenvandine> after please... i'll be out for the rest of the year and will surely forget everything :)
[16:12] <didrocks> ok ;) making sense, I'll maybe just give mterry some introduction first as he will be the only desktop soldier starting on 14th
=== gatox is now known as gatox_lunch
[16:13] <didrocks> kenvandine: seb128: do you know why we have debian/places in ubuntu-mono btw?
[16:13] <mterry> :)
[16:13] <didrocks> mterry: we'll all think about you! (while drinking and celebrating) :-)
[16:13] <kenvandine> didrocks, i don't
[16:13] <mterry> hah
[16:14] <didrocks> kenvandine: I'm about to kill that if seb128 doesn't answer in 3
[16:14] <didrocks> 2
[16:14] <didrocks> 1
[16:14] <seb128> didrocks, no
[16:14] <didrocks> :)
[16:14] <didrocks> ok, let's move it!
[16:14] <seb128>   [Paul Sladen]
[16:14] <seb128>   Pre-caching: buildd build fails to pick up underlying raster wallpaper
[16:14] <seb128>   from 'ubuntu-wallpapers', leaving the icon empty unless highlighted.
[16:15] <didrocks> I saw that, its not enlightening to me :)
[16:15] <seb128> didrocks, the bzr log says ... which was when they tried to include the wallpaper image in the svg ... which was great because it made stuff like nautilus take one extra second to start because it was resizing wallpapers to fit in each icon size variant ...
[16:15] <didrocks> ah that
[16:15] <didrocks> you enjoyed it!
[16:15] <seb128> yeah, so much that I nuked it :p
[16:15] <didrocks> heh
[16:15] <didrocks> so moving the .png
[16:15] <didrocks> and installing the new ones
[16:16]  * didrocks is a bzr mv away ;)
[16:17] <mspencer> mpt: I've got another idea - what about adding an option to open Bazaar Explorer (shown only if it is installed) when getting the code for a package/project?
[16:18] <seb128> didrocks, drop the debian/rules "# Go on! Give up; admit defeat by the buildd. " hack as well I think
[16:18] <mpt> mspencer, sure, good idea
[16:18] <didrocks> seb128: yeah, I saw it
[16:18] <didrocks> it's horrible
[16:18] <didrocks> I'm waiting for robru who started on it
[16:18] <didrocks> and will probably finish and do the cping
[16:19] <mspencer> mpt: Should I report that as a bug and assign you to it?
[16:19] <mpt> mspencer, ok :-)
[16:19] <mspencer> mpt: Thanks!
[16:20] <seb128> didrocks, hum, trying a local build, dropping that dir results in missing icons...
[16:21] <didrocks> seb128: yeah, because of the cp in debian/rules
[16:21] <didrocks> seb128: we just need to relocate in ./common/places
[16:21] <seb128> didrocks, right
[16:21] <didrocks> (. being the root of the package)
[16:21] <didrocks> and do the cp in the packages from it
[16:21] <didrocks> will be easy, no worry :)
[16:21] <seb128> ok ;-)
[16:22] <didrocks> and removing tonnnnnnnnnnnnsss of comments :)
[16:22] <desrt> seb128: so i think we should take 3.8 this cycle :)
[16:22] <seb128> desrt, suuuure
[16:22] <desrt> seb128: i'm fixing all the issues for you ahead of time :p
[16:22] <seb128> desrt, I see and I appreciate it, thanks ;-)
[16:22] <desrt> but what's the point if you don't take it? :p
[16:22] <didrocks> desrt: seb128 already started to push it :p
=== ejat- is now known as ejat
[16:27] <mspencer> mpt: I've filed this as bug 1086870
[16:27] <ubot2> Launchpad bug 1086870 in Contributor Console "Should be able to open Bazaar Explorer when getting code" [Undecided,New] https://launchpad.net/bugs/1086870
[16:27] <mpt> ok
=== VD is now known as Guest23923
=== gatox_lunch is now known as gatox
[17:47] <didrocks> ogra_: unity, compiz and nux seems to be blocked because of edubuntu alpha1 freeze
[17:48] <didrocks> ogra_: so you maybe will have to wait after this to rebuild your image
[17:53]  * didrocks waves good evening
=== m_conley is now known as m_conley_away
=== m_conley_away is now known as m_conley
=== VD is now known as Guest64385
=== m_conley is now known as m_conley_away
=== m_conley_away is now known as m_conley
[18:55] <dobey> pitti: is there any way to programmatically get the pygobject version number in python, when using gir bindings?
[18:55] <pitti> $ python3 -c 'import gi; print(gi.version_info)'
[18:55] <pitti> (3, 7, 2)
[18:55] <pitti> dobey: ^ that's the official API
[18:56] <dobey> ah ok
[18:56] <dobey> thanks
[18:56] <dobey> now i'll have to patch twisted
[18:58] <mterry> Sweetshark, so should I look at bug 1034560 now?  We want libreoffice-report-builder in raring?
[18:58] <ubot2> Launchpad bug 1034560 in libserializer (Ubuntu) "[MIR] libloader-java, libformula-java, librepository-java, libfonts-java, libserializer-java " [Undecided,New] https://launchpad.net/bugs/1034560
[19:11] <Sweetshark> mterry: yep.
[19:16] <plars> seb128: ping?
[19:16] <Sweetshark> mterry: http://anonscm.debian.org/gitweb/?p=pkg-openoffice/libreoffice.git;a=blob;f=rules;h=2ed34ada677a8a688eda6511093683a0699361aa;hb=651089892251843ef3ae93d1499121957e25ac11#l848 shows the additional deps that would be ensued by enabling report-builder ..
[19:17] <plars> seb128: could I push https://bugs.launchpad.net/ubuntu/+source/compiz/+bug/1086036 over to you or didrocks (or anyone else that might want to take a look)?
[19:17] <ubot2> Launchpad bug 1086036 in compiz (Ubuntu) "Compiz driven ubiquity-dm crashes when booting raring" [High,Confirmed]
[19:37] <seb128> plars, hi
[19:37] <plars> hi seb128
[19:37] <seb128> plars, seems like duflu looked at it, he's the best placed to debug the issue
[19:37] <seb128> plars, but try to ping bregma about it, he can probably help you
[19:38] <plars> bregma: ping?
[19:39] <seb128> plars, btw nothing changing in the compiz stack "between the 20121129 and 20121202"
[19:39] <plars> seb128: but it was switched out for metacity apparently
[19:40] <seb128> plars, but it seems like around the time ubiquity started using compiz
[19:40] <plars> right
[19:40] <seb128> plars, right, so that's not very helpful info
[19:40] <seb128> "compiz started to have an issue when we started to use it"
[19:40] <seb128> sure you didn't have issue when you were not using it...
[19:40] <seb128> it doesn't mean the issue was not there
[19:41] <plars> seb128: compiz was merely our best guess based on what we are seeing, but we need help debugging where the problem actually lies.  There was a corresponding crash with compiz that was suspicious, but may or may not be the underlying cause
[19:42] <plars> seb128: it could, in fact, be what duflu suspected, and a dup of this 2 month old compiz bug.  But that bug hasn't seem to have made any progress and this is impacting daily raring images now, so it needs some attention
[19:42] <seb128> plars, right, well the #ps team (e.g bregma and duflu) are the best placed to help you
[19:42] <plars> seb128: thanks I'll talk to them
[19:42] <seb128> let us know how it goes
[19:43] <bregma> plars, our best bet is to wait until duflu is on and get his opinion on possible causes
[19:44] <seb128> bdrung, thanks for the file-roller work!
[19:44] <bdrung> seb128: you're welcome
[19:45] <bdrung> seb128: i recommend to add bugs to attract new developers ;)
[19:45] <seb128> hehe ;-)
[19:48] <bdrung> that's the normal way for me to become a maintainer of a package
[19:56] <robert_ancell> tedg, is lp:indicator-network obsolete? I'm working on a Network Manager indicator and wondering if we should take over that branch. Also do you know who maintains ~indicator-developers?
[20:00] <robert_ancell> seb128, fricking copyrights :)
[20:04] <seb128> robert_ancell, hey
[20:04] <seb128> robert_ancell, don't tell me, I hate them!
[20:04] <robert_ancell> seb128, I hoped you'd just patch it for me :)
[20:05] <seb128> robert_ancell, I did!
[20:05] <seb128> robert_ancell, not nice from you though :p
[20:06] <robert_ancell> seb128, your emails weren't in unified diff format ;)
[20:07] <tedg> robert_ancell, It's basically connman based, we've intended to port it to network manager eventually, but I don't think that's staffed currently.
[20:07] <seb128> robert_ancell, heh, I did upload to the queue for you!
[20:07] <robert_ancell> seb128, oh cheers, I didn't see that
[20:09] <robert_ancell> tedg, right, we have a working nm indicator but it's not based on the old indicator-network codebase - shall push the old lp:indicator-network to another branch and link lp:indicator-network to the new one?
[20:09] <seb128> robert_ancell, I posted "I rejected but I uploaded a fixed version" (or something around those lines)
[20:09] <seb128> robert_ancell, for both seahorse-share and realmd
[20:09] <robert_ancell> seb128, oh too early for me!
[20:10] <tedg> robert_ancell, ?  I'm confused.
[20:10] <tedg> robert_ancell, You wrote a new indicator?
[20:11] <robert_ancell> tedg, yes
[20:12] <tedg> Uhm, why?
[20:13] <robert_ancell> tedg, based it off the indicator-bluetooth code and both I and cyphermox considered the old i-n overly complicated
[20:13] <seb128> there is a chance the new indicator is faster to write than trying to adapt the 3 years old conman based one
=== Quintasan_ is now known as Quintasan
=== lifeless_ is now known as lifeless
[20:58] <robert_ancell> desrt, hey, if you register an object in GDBus with a property and you notify that property doesn't it generate a D-Bus notify?
[20:59] <robert_ancell> i.e. a PropertiesChanged D-Bus signal
[21:29] <desrt> robert_ancell: which kind of object and how?
[21:29] <micahg> kenvandine: did you discuss the ctypes thing with chrisccoulson, I would think not being able to use a multiarch dir for a library is a bug
[21:29] <desrt> robert_ancell: if you mean register_object() then no
[21:29] <robert_ancell> desrt, yes
[21:30] <kenvandine> micahg, not yet
[21:30] <desrt> robert_ancell: if you mean gdbus-codegen then "maybe"
[21:30] <robert_ancell> desrt, no, from vala
[21:30] <desrt> that's more interesting.  also a maybe.
[21:30] <desrt> i guess it depends on if the signal is part of the tagged-public dbus interface
[21:30] <kenvandine> micahg, i think it should find it without the path
[21:30] <desrt> in that case i think it is emitted
[21:31] <robert_ancell> desrt, it is, but it doesn't seem to generate the signal. I think it's because you potentially want to send only one d-bus signal and GLib generates a notify for each property
[21:31] <robert_ancell> desrt, but in that case it really should be more clearly documented and there should be a bus.notify_properties (object, properties)
=== attente is now known as attente_zzz
[21:33] <robert_ancell> desrt, it's probably more of a vala issue because in C you'd have to handle all the properties anyway
[21:33] <desrt> robert_ancell: the property change signal thing is... controversial
[21:33] <desrt> some people say it's part of the spec, others disagree
[21:33] <desrt> in particular, i think david is a strong advocate and thiago hates it
[21:33] <robert_ancell> desrt, it's practically required and used everywhere
[21:34] <micahg> kenvandine: well, I would think the multiarch path would be accessible without being explicitly defined if that's what you mean
[21:35] <kenvandine> micahg, right
[21:35] <desrt> robert_ancell: i think QtDBus is quite intentional about not supporting it
[21:35] <robert_ancell> groan
[21:36] <robert_ancell> and this is why everyone reimplements essentially a properties interface for their objects :)
[21:36] <desrt> robert_ancell: depending on what you're doing, objectmanager is a good fit
[21:36] <robert_ancell> desrt, I'm trying to make an indicator
[21:36]  * robert_ancell cries
[21:36] <desrt> erm
[21:37] <desrt> isn't there a library for that?
[21:37] <robert_ancell> there's at least 5!
[21:37] <desrt> heh
[21:37] <desrt> so why are you talking to dbus directly?
[21:37] <desrt> pick one :)
[21:37] <robert_ancell> I've solved my issue now, but it felt clunky
[21:37] <desrt> btw...
[21:37] <desrt> any idea if we will be on logind soon?
[21:37] <robert_ancell> you have to make a service to send the icon name from the indicator service to the indicator library
[21:38] <robert_ancell> desrt, seb128 said slangasek said soon
[21:38]  * desrt is getting sick of dealing with consolekit issues
[21:38] <chrisccoulson> micahg, kenvandine, it just uses dlopen(), and it works fine here btw for my addons
[21:38] <robert_ancell> me too
[21:38] <desrt> BSD is gonna be pissed :(
[21:38] <chrisccoulson> eg, https://addons.mozilla.org/en-US/firefox/addon/unityfox/ happily loads libunity from a multiarch location just fine
[21:38] <kenvandine> chrisccoulson, i am pretty sure this was working before in unity-firefox-extension
[21:38] <desrt> robert_ancell: this cycle soon, you think?
[21:39] <robert_ancell> desrt, that's what I've heard
[21:39] <kenvandine> we tested it when we converted it to multiarch
[21:39]  * desrt grins widely
[21:39] <desrt> as soon as there is a stable release of ubuntu sans-CK, support is just going to utterly utterly vanish from GNOME
[21:40] <desrt> which will be awesome for absolutely everyone except the BSDs :/
[21:41] <robert_ancell> desrt, they should be able to do the same thing we're doing and switch to logind
[21:42] <desrt> i thought i understood that logind was wired pretty tightly into some linux-only kernel features...
[21:42] <robert_ancell> desrt, I think that's an exaggeration. We/they might have to drop a few features, but the core part of tracking sessions and seats only really requires PAM afaict
[21:43] <desrt> hmmm
[21:43] <micahg> kenvandine: FWIW, that package was lacking a multiarch label in the control file even though it was converted
[21:43]  * desrt was pretty sure cgroups was a core feature there
[21:43] <robert_ancell> the big new thing is tracking session processes with cgroups, but if you disable that you aren't any worse off that CK
[21:43] <desrt> true enough
[21:43] <robert_ancell> and ideally we/they can enable all that in the future
[21:44] <robert_ancell> even if a different technology is required
[21:44] <desrt> i should email antoine
[21:48] <seb128> desrt, robert_ancell: some systemd trolling?
[21:49] <robert_ancell> seb128, no trolling
[21:49] <robert_ancell> seb128, desrt was asking when we get logind
[21:49] <seb128> this cycle hopefully, I ping slangasek about that last week
[21:50] <desrt> seb128: i can't believe i'm wasting my time fixing all these consolekit issues :p
[21:50] <seb128> he didn't manage to finish looking at that last cycle because secure boot was higher priority on the list
[21:50] <kenvandine> micahg, what was it missing?
[21:50] <seb128> hum, ping->pinged
[21:50] <micahg> kenvandine: Multi-Arch: (same|foreign)
[21:50] <kenvandine> oh... i've never seen that :)
[21:51] <kenvandine> none of my checkouts have that... weird
[21:51] <micahg> kenvandine: it lets dpkg/apt know which arch of a library can suffice for another arch's dependency
[21:51] <seb128> desrt, yeah, maybe for nothing if we get logind next cycle by the time we get GNOME 3.8 ;-)
[21:51] <desrt> seb128: it still suits me
[21:51] <desrt> it's a personal priority for me to keep jhbuild working against the latest stable ubuntu
[21:51] <seb128> desrt, you are happy then, see you didn't waste time ;-)
[21:51] <desrt> and since we don't have a stable release (or any realease) with logind yet...
[21:52] <seb128> desrt, joke aside Steve said it should be doable
[21:52] <micahg> kenvandine: apt-cache show libpoppler28 | grep Multi-Arch
[21:52] <seb128> desrt, he just needs to sort out the weird cgroups usage/requirement
[21:53] <desrt> the cgroups thing is semi-awesome
[21:53] <desrt> but also weird at the same time...
[21:53] <desrt> some things in the session (gvfsd, dconf-service, etc.) are services to the user rather than services to the session
[21:53] <desrt> and they should not have their lifecycle tied to the session cgroup
[21:53] <desrt> i should talk to lennart about that...
[21:55] <achiang> random question because i'm awake... do we have plans to migrate gtk2 apps => gtk3 (similar to python2 => python3) ?
[21:56] <desrt> achiang: been going on for several cycles now...
[21:56] <desrt> mostly consists of pressuring upstreams to get their act together :)
[21:56] <achiang> heh. what are the worst offenders currently?
[21:56] <desrt> ubuntu1 has a python2 problem
[21:56] <desrt> that's the biggest case, i think
[21:57] <seb128> achiang, libreoffice, firefox
[21:57] <desrt> software centre too, but the problem is not too serious (and may be solved)
[21:57] <desrt> ya... and libreoffice/firefox on the gtk side
[21:57] <seb128> desrt, he asked about gtk2 to gtk3
[21:57] <seb128> s-c is gtk3
[21:57] <achiang> damn, those are pretty large programs
[21:57] <desrt> ah.  i thought he asked about both :p
[21:57]  * achiang is happy for lots of extra info :)
[21:57] <seb128> achiang, yeah, as always easy stuff are done for a longtime
[21:57] <desrt> achiang: s-c and u1 are python2 problems... not gtk2 :)
[21:58] <achiang> it's frustrating that we ship gtk2, gtk3, *and* qt4 on the cd
[21:58] <seb128> achiang, usually when something remains to after a few cycle it's because they are hard ones...
[21:58] <achiang> i wonder if we'll have this same problem with qt4 and qt5
[21:58] <seb128> achiang, tell us...
[21:58] <desrt> firefox has gtk3 patches....
[21:58] <seb128> achiang, it's easy enough to kick out qt4
[21:58] <seb128> desrt, do we have adobe plugins for gtk3?
[21:59] <achiang> seb128: ah, good. re: qt4
[21:59] <desrt> seb128: does the flash plugin link against gtk2?
[21:59] <seb128> achiang, qt is only for u1
[21:59] <seb128> desrt, yes
[21:59] <desrt> seb128: that's approximately the stupidest thing i've ever heard
[21:59] <desrt> why!?
[21:59] <seb128> desrt, why do you think it doesn't work in epiphany...
[21:59] <desrt> GtkPlug?
[21:59] <desrt> seb128: i don't really use flash...
[21:59] <seb128> desrt, not sure, the source is not available
[22:00] <desrt> seb128: easy enough to check for symbols...
[22:00] <seb128> desrt, they have an ui in any case to choose quality and other things
[22:00] <desrt> seb128: seems to be using file selector, menus, message dialogs
[22:01] <desrt> input methods...
[22:01] <desrt> clipboard
[22:01] <desrt> lots of stuff, actually
[22:01]  * achiang finds http://worldofgnome.org/libre-office-in-gtk3-and-wayland/
[22:03] <achiang> there's gotta be some equivalent of jwz's law of envelopment... any sufficiently large program must write its own toolkit
[22:04] <desrt> freaking gimp
[22:04] <achiang> xul, nux... ;)
[22:04] <desrt> iirc vlc has its own toolkit as well
[22:04] <TheMuso> VLC uses QT afaik.
[22:05] <desrt> maybe i get it confused with libreoffice's toolkit being called vcl
[22:06] <seb128> don't forget gnome-shell not using gtk but their own stuff as well ;-)
[22:07] <desrt> also meego using its own toolkit :)
[22:07] <seb128> desrt, tizen uses efl ;-)
[22:08] <desrt> finally!  someone who uses an existing product!
[22:08] <desrt> wait... efl?
[22:08]  * desrt falls over
[22:08] <seb128> the world got tired to wait for e17
[22:08] <seb128> somebody had to release a product using efl
[22:08] <seb128> "release"
[22:09] <desrt> hah
[22:09] <seb128> or maybe tizen will be doomed the same way e17 was
[22:09] <seb128> iz efl's fault
[22:09] <desrt> more similar to the way that meego is, i think :)
[22:09] <desrt> linux is cursed
[22:09] <desrt> unless you're google.....
[22:10] <desrt> maybe that's the key... you can only be successful using linux if you carry a significant delta vs. the upstream kernel :)
[22:12] <seb128> I don't like that rule :p
[22:12] <desrt> here's another one, then:
[22:12] <desrt> android doesn't have glib
[22:12] <desrt> you can only be successul if you don't have glib :p
[22:12] <seb128> some people are working on solving that it seems
[22:12] <desrt> (also: dbus, glibc, etc... GNU in general, actually)
[22:13] <seb128> wasn't slomo sending patches for having glib to work on android?
[22:13] <desrt> yes
[22:13] <desrt> they have gstreamer working with it
[22:16] <robert_ancell> tedg, do you know about IdoSwitchMenuItem?
[22:17] <tedg> robert_ancell, A bit, but charles or larsu would probably know more.
[22:17] <robert_ancell> tedg, what is the correct way to set a label on it?
[22:17] <seb128> robert_ancell, indicator-sync is an example of indicator using it
[22:17] <seb128> robert_ancell, in case that helps you
[22:18] <robert_ancell> seb128, yeah, looking at that, but it does it a bit differently
[22:18] <robert_ancell> tedg, is that what the "content area" is for?
[22:18] <tedg> ?
[22:18] <seb128> charles, ^
[22:18] <tedg> I think that you probably don't want to do switches with libappindicator :-)
[22:19] <tedg> There's now way to put the IDO widget in and have it introspected back out.
[22:19] <tedg> That feature was dropped half way when I learned about GMenuModel.  Didn't make sense to complete it.
[22:20] <robert_ancell> tedg, I'm copying how indicator-sync does it and it uses DbusmenuGtk.Client.add_type_handler
[22:20] <tedg> robert_ancell, On the client side, yes.
[22:20] <tedg> robert_ancell, You can't do that on the server side.
[22:20] <robert_ancell> tedg, that's fine
[22:21] <robert_ancell> tedg, ok, it seems to have worked
[22:24] <desrt> achiang: hey... what's the status of power-and-usb-otg-at-the-same-time?
[22:25] <achiang> desrt: hm... that's a better question for ogra or janimo. i don't think it's working though
[22:25] <desrt> achiang: know anything about these 4 contacts on the side of the device?
[22:25] <desrt> rumour has it i can apply +5 and ground to two of them....
[22:25] <achiang> desrt: they're "pogo plugs"
[22:25] <achiang> desrt: supposedly a dock is coming
[22:25] <achiang> but i don't know if it's shipped yet
[22:26] <desrt> ya... i've seen the dock
[22:26] <desrt> just wondering if we support that
[22:27] <desrt> i was reading that some older versions of the android firmware refused to charge from them
[22:27] <achiang> interesting... i don't know anyone who's tried testing
[22:27] <desrt> k.  thanks
[22:28] <robert_ancell> tedg, hey, so I think I have my head around doing indicators right - can I get you to review http://bazaar.launchpad.net/~robert-ancell/+junk/indicator-vala/files/head:/src/ some time?
[22:30] <tedg> robert_ancell, I can, I have a rather long review queue sadly though.  You might ping charles or larsu to see if their's is shorter.  But I can do it if they're busy as well.
[22:30] <robert_ancell> tedg, I know larsu is busy :) I'll send you all an email and I'll just press on for now assuming it's correct
[22:31] <robert_ancell> bbl
=== VD is now known as Guest49227
=== m_conley is now known as m_conley_away