16:31 <balloons> #startmeeting clock app
16:31 <meetingology> Meeting started Thu Aug  7 16:31:55 2014 UTC.  The chair is balloons. Information about MeetBot at http://wiki.ubuntu.com/meetingology.
16:31 <meetingology> 
16:31 <meetingology> Available commands: action commands idea info link nick
16:31 <dholbach> hey hey
16:32 <nik90> hey
16:32 <balloons> howdy
16:32 <nik90> looks like a hectic schedule to get past TRAINCO
16:33 <balloons> So, I'm filling in for popery again. On my agenda is basically just the clock reboot stuff.
16:33 <nik90> ok
16:33 <dholbach> nik90, I just wanted to give you a quick update on the reboot packaging
16:33 <dholbach> nik90, sergiusens said that I should run the tests not during the build but as an autopkgtest, which I'm looking into right now
16:34 <nik90> dholbach: I noticed that sergiusens added some inline comments after I requested him to take a look at it again few hours back
16:34 <nik90> ok
16:34 <dholbach> nik90, shall I file bugs for all the stuff I can still see as "needs doing" for getting the reboot bits properly packaged?
16:34 <dholbach> nik90, the stuff about autopkgtest was the online comment I saw
16:34 <dholbach> and I don't consider it super important
16:34 <nik90> dholbach: yes please do,
16:34 <dholbach> the other stuff is much more important, ie: install files into the right location
16:34 <nik90> ok
16:34 <dholbach> alrightie - shall I use the 'reboot' tag for that, or something?
16:35 <nik90> yeah sure, I will be any way moving them to the "RTM" milestone that I have dedicated for the clock app reboot
16:35 <dholbach> ok
16:35 <dholbach> that's all from my end
16:36 <balloons> dholbach, having a quick look at https://code.launchpad.net/~dholbach/ubuntu-clock-app/reboot-packaging/+merge/229173 my guess is you could leave autopilot out of it.. I would kill the clock-app-autopilot package as well
16:36 <nik90> I will keep pinging sergiusens to fix the cmake stuff since I would ideally like to replace the old clock app by next week
16:36 <balloons> although, jenkins core apps mp still does use the -autopilot packages... mrhmmm
16:36 <dholbach> yeah, that's something which needs fixing as well
16:36 <dholbach> I'll get right on to filing those bugs now
16:37 <nik90> balloons: btw autopilot3 vis is freaking awesome. It actually has a highlight feature which show which UI is being selected in the vis tool
16:38 <dholbach> https://bugs.launchpad.net/ubuntu-clock-app/+bugs?field.tag=reboot
16:38 <dholbach> (more to come :))
16:39 <nik90> dholbach: ack :)
16:39 <nik90> balloons: I am slowly diverting my attention to the AP tests as well to help out carla
16:39 <nik90> I just have about 5-6 branches currently waiting in queue. It requires some one to do the review process
16:39 <balloons> nik90, yea glad you like it.. when the py2 to py3 split happened, I lost that version of vis and never realized it has simply been packaged as python3-vis
16:40 <nik90> Usually I ask victor thompson to do the review, but I can see that he is swamped with the music app MPs so didnt want to bother him
16:41 <balloons> nik90, mm.. true, you need a good qml dev to review don't you
16:41 <nik90> balloons: actually for these latest MPs not really.
16:41 * balloons is looking through them
16:41 <nik90> the latest MPs are basically borrowing components from the address book into the clock app
16:42 <balloons> I see you have renato and charles looking at them
16:42 <nik90> balloons: so the borrowed components do not need to be reviewed since they were already reviewed by addressbook devs
16:42 <nik90> balloons: I just need someone to review the small integration code in the clock app
16:42 <nik90> balloons: charles reviewed the c++ code, and not the qml part
16:43 <nik90> hmm I could ask renato if he is free to review those MPs actually
16:44 <balloons> the qml is a small piece. I could review if you wish.. insomuch as the code is reasonable and it works for m
16:44 <nik90> balloons: yeah go for it
16:44 <balloons> this looks doable for instance :-) https://code.launchpad.net/~nik90/ubuntu-clock-app/add-vibration-support/+merge/229327
16:45 <balloons> k, so once these mp's are landed, there is the trouble of traincon 0
16:45 <nik90> exactly, only small bits
16:45 <dholbach> nik90, https://bugs.launchpad.net/ubuntu-clock-app/+bugs?field.tag=reboot is what I could think of
16:45 <nik90> balloons: Only https://code.launchpad.net/~nik90/ubuntu-clock-app/add-vibration-support/+merge/229327 and https://code.launchpad.net/~nik90/ubuntu-clock-app/add-custom-sound-backend/+merge/225812 are affect by traincon 0
16:46 <nik90> balloons: others are localised to clock app
16:46 <nik90> dholbach: thnx. will triage after this meeting
16:46 <dholbach> great
16:47 <balloons> so nik90 I know friday was originally a goal for delivering a click correct?
16:47 <nik90> balloons: yeah I generally try to update the app in the store on fridays
16:48 <nik90> balloons: by the discussion in ci-eng, they really want trainco 0 to end by tomorrow
16:48 <balloons> but if traincon0 is still on, what's the plan? Will you simply delay until it's lifted, or try and release something?
16:49 <nik90> if traincon0 is still on, then I will have to postpone to Monday since I don't really have much to release
16:49 <nik90> everything new since the last update is in the MP's you saw
16:50 <balloons> nik90, ok sounds fine. So apart from dholbach's mp, everything should be ready to land. I'll try and review each mp
16:50 <dholbach> well, I'll need some help
16:51 <balloons> dholbach, yes, I'm wondering what is needed to land yours
16:51 <nik90> balloons: thnx
16:51 <dholbach> I merely came in to help a bit and had a look at some of the packaging
16:51 <balloons> nik90, don't let me forget about talking tests before we end
16:51 <balloons> dholbach, and now you are sucked in :-)
16:51 <nik90> dholbach: yup no exits() I am afraid :D
16:51 <dholbach> but it looks like somebody needs to put in some work into the build system to fix all the paths, etc
16:52 <dholbach> hah
16:52 <dholbach> then you're stuck with me
16:52 <dholbach> that doesn't help you ;-)
16:52 <nik90> u make it sound bad :P
16:52 <balloons> so you want make autopilot to work and be a thing
16:53 <nik90> balloons: is that directed at me
16:53 <dholbach> I don't think that running the tests is the biggest problem
16:53 <nik90> ?
16:53 <balloons> I'm just confused what else is intended in the mp..
16:53 <dholbach> if we want this packaged, we need to make sure that all the files and directories go into the right places, if we install into system locations
16:53 <dholbach> balloons, packaging
16:53 <balloons> I'm asking because I'm curious about simplifying what you are trying to do
16:54 <dholbach> just taking ubuntu-clock-app as an upstream packaging and make it work as a .deb
16:54 <dholbach> just taking ubuntu-clock-app as an upstream package and make it work as a .deb
16:54 <nik90> balloons: the cmake instructions required to make it install at the correct place is horribly broken
16:54 <dholbach> but right now it does not install its bits into the right places
16:54 <nik90> once that is fixed, the debian packaging should fall in place pretty quick
16:54 <dholbach> yep
16:54 <balloons> ok. So, I as I said, I don't think the tests need to be packaged.. The trouble is I remembered jenkins uses them, so I suppose that's ok to package them for that reason.
16:55 <balloons> shouldn't you be able to get away with not having a make_autopilot though? I don't know of another project that does it that way
16:55 <dholbach> sure, we can disable all the tests
16:55 <nik90> balloons: this includes the .desktop files, c++ backends as well, not just the autopilot package
16:55 <balloons> that would more or less let this land then right?
16:55 <dholbach> hat won't fix 1354079, 1354081, 1354082, 1354083 though
16:55 <nik90> dholbach: well we did get "make autopilot" working at the end
16:56 <balloons> if we could split the mp and file a bug, I'm open to such ideas
16:56 <balloons> bug 1354079
16:56 <balloons> no bugbot :-(
16:56 <dholbach> nik90, it didn't work for me in a chroot, it passed though, because 'make autopilot' did not return its return code
16:56 <nik90> dholbach: oh
16:57 <nik90> balloons: https://bugs.launchpad.net/ubuntu-clock-app/+bugs?field.tag=reboot summarises all the bugs with the MP
16:57 <dholbach> I wonder if add_custom_target() in CMakeLists.txt ignores the retcode
16:57 <dholbach> it exploded and the retcode was 0 all the time
16:57 <balloons> well I would push you to adopt autopkgtest as sergio recommends, but my brain is churning on how that works with debian packaging and cmake..
16:57 <nik90> ok everything pretty much rests on sergiusens and his cmake fixes.
16:57 <dholbach> I'm looking into the autopkgtest now
16:57 <dholbach> but that's the least of our worries
16:57 <dholbach> yes
16:58 <dholbach> I'll file a bug about the autopkgtest too
16:58 <dholbach> and update it
16:58 <nik90> ok
16:58 <nik90> balloons, dholbach: I will keep a tab on sergiusens to see if he can provide us a working diff for the cmake mess
16:58 <nik90> by today or tomorrow
16:59 <dholbach> rock and roll!
16:59 <balloons> sounds like you guys are more in tune with what needs done, so :-)
16:59 <dholbach> nik90, http://paste.ubuntu.com/7981269/
17:00 <dholbach> I'll file a separate bug for that as well
17:00 <balloons> let's end by chatting about the AP tests then.. How's things on that front?
17:00 <nik90> dholbach: ok
17:01 <nik90> balloons: we got the alarm tests up and running as I mentioned yesterday. I broke them with my 5 MP's but fixed them immediately as well :D
17:01 <balloons> heh.. any concerns or needs on that front?
17:01 <nik90> I have asked carla to wait for these MP's to land before starting with the world clocks since the UI is changing a bit
17:01 <balloons> one of the things to look out for is perhaps you using new components that don't have, but could use, helpers
17:02 <nik90> what do you mean ^^
17:02 <balloons> so Carla is also waiting then for these mp's to land before further work?
17:02 <nik90> yes
17:02 <nik90> let me explain on that
17:02 <balloons> nik90, I mean, something like your class PageWithBottomEdge
17:03 <balloons> I'm curious if that is or could be a helper.. off the cuff, seems like it
17:03 <dholbach> got another meeting coming up... so I'll chat with you tomorrow again :)
17:03 <nik90> we borrowed the custom swipe delete component from the address book which will make the swipe delete helper from the SDK fail which is also what carla used
17:03 <balloons> cu dholbach
17:03 <nik90> dholbach: ack. see you
17:03 <balloons> mm.. why a custom component?
17:04 <nik90> so I had to fix that in my Mps. Obviously these UI improvements came with my branch, so carla couldn't have seen that
17:04 <balloons> because it's better, heh?
17:04 <nik90> because design required it, yes
17:04 <nik90> it was supposed to come from the SDK, but won't make it in time for RTM
17:04 <balloons> gotcha.. ok.. so then there should indeed be a helper for i
17:04 <balloons> *it.. just in this interim you are providing both if you will
17:05 <nik90> yeah
17:05 <balloons> kk.. so yea, that's the type of stuff I wanted you to be on the lookout for
17:05 <nik90> I borrowed the PageWithBottomEdge AP helper from the address book actually, so it should be stable enough
17:05 <balloons> and I think that's it from me, unless you have other questions
17:05 <nik90> this way if adress book ap fails, it will fail the clock app as well, but at least we will know why
17:05 <balloons> yea, the worry about borrowing helpers is splintering code.. it makes messes and painful when things change
17:06 <balloons> we've been there before
17:06 <nik90> tell me about it :/ ...luckily we have this situation only for another month or so before it lands in the SDK properly
17:06 <nik90> I am keeping an active look out at the address book trunk for new merges
17:07 <nik90> but yeah that's was it from me as well
17:07 <balloons> nik90, it would probably be useful to have a bug tracking this.. so at least it's listed somewhere .. what do you think?
17:07 <nik90> regarding the component landing in SDK?
17:07 <balloons> I'm thinking a bug that says we need a helper for.. well.. maybe it's not so easy
17:08 <balloons> yep, nvm
17:08 <balloons> :-)
17:08 <balloons> ok, so with that, time to close
17:08 <nik90> I have added a README.upstream which keeps track of this for my sanity :D
17:08 <balloons> #endmeeting