15:59:32 <zack> #startmeeting
15:59:32 <MeetBot> Meeting started Fri Jul 24 15:59:32 2015 UTC.  The chair is zack. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:59:32 <MeetBot> Useful Commands: #action #agreed #help #info #idea #link #topic.
15:59:43 <zack> looks like the first topic will be easy :)
15:59:46 <zack> #topic roll call
15:59:51 <zack> presents: zack, clemux, orestis
15:59:58 <zack> (matthieucan and jpleau are on VAC)
16:00:03 <zack> #topic next meeting
16:00:18 <zack> as a reminder, *I* will be on VAC starting next week, for ~2 weeks
16:00:27 <zack> so I will not be around for 2 meetings
16:00:40 <zack> IIRC, matthieucan said he will be back, and can chair for those 2 meetings
16:00:47 <zack> (right? :))
16:00:51 <orestis> i think so
16:01:05 <zack> in any case, I don't think it makes sense to schedule a meeting right now
16:01:14 <zack> I'll mail all of us reminding matthieu to do so
16:01:29 <zack> but I'm confident it will be next week, back to the usual day/hour
16:01:42 <zack> #topic orestis - weekly review
16:01:56 <zack> orestis: floor is yours
16:02:04 <orestis> i pushed some use stories in PR#30
16:02:11 <orestis> i listed target users as well there
16:02:23 * zack opens the PR
16:02:25 <orestis> i ll push the use cases maybe tomorow
16:02:48 <zack> you haven't yet asked for review on that stuff, right?
16:02:56 <zack> or have I missed it?
16:03:15 <orestis> hm maybe not.. i thought you would check the old functionalities
16:03:30 <orestis> so i guessed that you would see that as well since it is on the same PR
16:03:43 <zack> I was actually waiting for an explicit request, I guess we didn't coordinate properly
16:03:46 <zack> but that's ok
16:03:51 <zack> here is what I suggest:
16:04:09 <zack> if you can submit a review request in the next ~20 hours
16:04:16 <zack> I'll be happy to go through it on my flight back
16:04:27 <zack> so that you've feedback before I leave on VAC
16:04:42 <zack> how does that sound?
16:04:43 <orestis> i have nothing to add on the use stories. so you can review those already
16:05:03 <orestis> i ll try to have the use cases as well for tonight so you can review all of them
16:05:04 <zack> ok, can you mail me asking to do so? (it works better with me as a TODO list ;))
16:05:15 <orestis> yeap i ll do that :)
16:05:27 <zack> bonus point: you include the .txt file as attachment to your email ;)
16:05:42 <orestis> in the meantime i have listed what i thought the most important tasks for the patch tracker on trello
16:05:45 <orestis> sure :)
16:05:58 <zack> ok, in which list on trello?
16:06:05 <orestis> on the patch tracker
16:06:10 <orestis> todo patches
16:06:12 <zack> ah, saw it
16:06:35 <zack> (i'm quickly going through them)
16:06:56 <zack> looks pretty good
16:07:05 <zack> there is only one I don't understand: the patch-track plugin for the updater
16:07:12 <zack> why is that needed?
16:07:44 <orestis> i thought we d have another table for the patch tracker so we would hjave to fill it on update time no?
16:07:56 <zack> so
16:08:00 <zack> it depends
16:08:19 <zack> using the DB makes sense only if we cannot sensibly extract all the info we need on the fly
16:08:26 <zack> when patches.d.n is used on the web
16:08:42 <orestis> with the old you could search by maintainers email for example.. if we choose to do that then we will need it
16:08:59 <zack> ok
16:09:07 <zack> but I think such a feature should not be specific to the patch tracker
16:09:18 <zack> if we want to search by email, I think it should be a general feature for all debsources blueprints
16:09:34 <zack> so I'd rather *not* put it in a patch-tracker-specific plugin
16:09:41 <orestis> ok i see
16:09:42 <zack> what else would need a DB table?
16:09:56 <orestis> i think nothing else for the moment
16:10:01 <zack> ok
16:10:09 <zack> so I think we should ignore the need of a plugin for patch tracker for now
16:10:21 <zack> and maybe add to backlog a more generic "search by package metadata" feature
16:10:27 <zack> (low priority)
16:10:34 <zack> what do you think?
16:10:54 <orestis> yeap sounds good. i ll update trello
16:11:00 <zack> cool, thanks
16:11:07 <zack> anything else to report about for this week?
16:11:10 <orestis> nope
16:11:18 <zack> next topic then
16:11:23 <zack> #topic orestis - next week
16:11:39 <zack> I guess you'll want to pick from the tasks you've created
16:11:50 <orestis> yeap,
16:12:05 <orestis> maybe the skeleton vlueprint, browsing by package prefixes etc..
16:12:11 <zack> ok
16:12:17 <zack> it should be fairly easy, right?
16:12:27 <orestis> yes those are going to be easy
16:12:56 <orestis> i think that next its probably good to start the summary of a package/version in the patch tracker
16:13:10 <zack> can you expand on what is a "summary" for you?
16:13:36 <orestis> general info of the package and then list the patches along with some info on the patch (description, files changed)
16:13:56 <zack> so, the package summary too seems to be a more generic feature
16:14:03 <zack> that we might want (one day) for all blueprints
16:14:05 <zack> doesn't it?
16:14:42 <zack> if so, I'd rather focus on patch tracker specific functionalities for this week
16:14:43 <orestis> i think its more patch tracker specific.. i mena why would we need that in the sources one?
16:15:04 <zack> which informaiton of a package do you plan to show?
16:15:24 <orestis> i found a [patch tracker running based on the old one.. i was talking about this page http://patches.osdyson.org/package/gdata-sharp/2.2.0.0-2+dyson1
16:15:31 <zack> oh, I think I understand what you've in mind
16:15:42 <zack> the browsing of the patch tracker will "stop" at the package/version level
16:15:52 <zack> while debsources goes down to the source file level
16:15:58 <zack> ok, fair enough
16:16:11 <zack> but I recommend making the code/template that shows the package information easy to reuse
16:16:18 <zack> because we *might* end up using it elsewhere
16:16:36 <orestis> yes sure.. more patch specific uses would be downloading or viewing a patch
16:16:41 <zack> ok
16:16:42 <orestis> if you prefer i can start with those
16:16:46 <zack> so
16:16:58 <zack> I think that to list the patches, you will already need to decide which patch formats to support
16:17:16 <orestis> true!
16:17:31 <zack> I think we should begin with "3.0 (quilt)" for now
16:17:41 <zack> as it is the most popular in *current* development suites
16:18:06 <orestis> yeap ok
16:18:08 <zack> don't bother supporting other formats for now, but keep in mind they *might* be supported in the future
16:18:13 <zack> so prepare the needed abstractions
16:18:39 <orestis> ok sounds good for me
16:18:43 <zack> looks enough for the week, right?
16:19:09 <orestis> yes.. if i have time i might do the view of a patch
16:19:33 <zack> sounds good
16:19:42 * orestis goes to update trello
16:19:45 <zack> k
16:19:50 <zack> #topic clemux - weekly review
16:19:54 <zack> clemux: you're up!
16:20:12 <clemux> so, I'm still trying to write tests for the celery tasks I've already written
16:20:26 <zack> any specific blocker?
16:20:37 <clemux> I'm kind of stuck... I think I need help, and should have for help earlier :/
16:20:49 <zack> uhm
16:21:00 <zack> the problem is that I'm about to leave for ~2 weeks
16:21:03 <clemux> yes :(
16:21:09 <zack> so,yes, if you're stuck you should've definitely asked earlier
16:21:14 <zack> why didn't you
16:21:15 <zack> ?
16:21:48 <zack> anyway, I still have *some* bandwidth before leaving
16:22:01 <zack> I'll be back in Europe on Sunday, but will be busy the whole day
16:22:10 <clemux> I thought I'd manage, then I got stressed out when I couldn't
16:22:14 <zack> on Monday I'll basically work
16:22:27 <clemux> i don't have any excuse / good reason
16:22:32 <zack> so, if you mail me before I leave (in ~20 hours) I can give you feedback while flying back
16:22:45 <zack> and then again I can give feedback on Monday, before leaving on Tuesday
16:22:50 <clemux> will do
16:23:03 <zack> so let's avoid losing these opportunities, as they will be the last before debconf
16:23:10 <clemux> do you have like10 min after the meeting?
16:23:21 <zack> I've to run to a talk
16:23:23 <zack> *but*
16:23:28 <zack> feel free to leave a /query
16:23:33 <clemux> yes
16:23:36 <clemux> ok
16:23:37 <zack> I''ll check IRC periodically and try to give feedback
16:23:44 <zack> it will just not be (really) synchronous
16:23:58 <zack> anything else?
16:24:36 <clemux> I've also looked at the update_suites and gc stages, but did not produce anything because I wanted to have the test suite ready first...
16:24:38 <clemux> so no
16:24:48 <zack> ok
16:24:56 <zack> #topic clemux - next week
16:25:13 <zack> so, same items of last week?
16:25:18 <clemux> ... and so I still got the same tasks to do :/
16:25:21 <zack> ok
16:25:32 <zack> I won't stress the point too much, because you're clearly well aware of it
16:25:41 <zack> but we're in dangerous territory with the async part
16:25:44 <zack> I think the design is great
16:25:50 <zack> and that you've coded quite a bit of it
16:25:56 <zack> but we're still far away from merging
16:26:11 <zack> and we should absolutely try to get to that point *earlier* than the end of gsoc
16:26:17 <clemux> yes
16:26:20 <zack> because we cannot risk to merge (and deploy) the last day
16:26:31 <zack> it's definitely too much of a dangerous change to do last minute
16:26:49 <zack> so, relax, face one small issue at a time, and we will get there in time :)
16:27:39 <zack> ok, misc stuff
16:27:54 <zack> #topic misc
16:28:02 <zack> other topics?
16:28:11 <clemux> nope
16:28:14 <orestis> i have one thing to mention.. travis fails on the PRs due to some cloning issue
16:28:25 <zack> "cloning" ?
16:28:33 <clemux> oh, yes
16:28:36 <orestis> it doesnt clone the repo
16:28:44 <zack> do you have a URL handy?
16:28:51 <zack> (to the travis failure)
16:28:56 <orestis> https://travis-ci.org/Debian/debsources/builds/72067704
16:29:03 <zack> let's see
16:29:18 <orestis> and https://travis-ci.org/Debian/debsources/builds/71402313
16:29:31 <zack> uhm
16:29:42 <zack> I'm trying it locally
16:29:56 <clemux> I did it locally recently
16:30:01 <zack> it looks to work
16:30:06 <clemux> had no problem
16:30:08 <orestis> maybe it was just these two tries and travis had some network issues
16:30:13 <zack> maybe they've just reduced the travis bandwidth so that it takes too long?
16:30:20 <zack> yeah, something like that
16:30:23 <zack> can you trigger a retry?
16:30:37 <zack> (worked fine on my laptop, and also very quick < 1 minute)
16:30:42 <orestis> i don't know how. ill check later
16:30:59 <zack> orestis: if there is no explicit way to do that, you can do an empty commit and push
16:31:00 <orestis> and the other thing is about debconf.. we are supposed to send a description of a small talk
16:31:09 <orestis> zack: yes sure i ll do that
16:31:10 <zack> git commit --allow-empty
16:31:23 <zack> orestis: uh?
16:31:47 <zack> a small talk that you should give?
16:31:50 <orestis> but as we said with olasd you are already going to talk about debsources so I am not sure what we should do
16:32:08 <orestis> gsoc stusents will talk for about 15 minutes about their work on debsources
16:32:12 <zack> so, please discuss this with olasd (on this channel, so that we have backlog)
16:32:35 <orestis> arggggg gsoc students will talk about their work
16:32:38 <zack> I think we can either pass, pointing to the other talk
16:32:40 <zack> *or*
16:32:47 <zack> you can do a brief presentation of your specific work
16:33:00 <zack> which might be too detailed to fit in the more general debsources talk by myself and matthieucan
16:33:05 <zack> I'm open to both options
16:33:19 <orestis> ok i ll discuss this with olasd later here
16:33:22 <zack> I mean, our talk will be general
16:33:36 <zack> so copyright.d.n, patces.d.n, and the async stuff will basically only mentioned in passing
16:33:37 <olasd> I've had exactly zero answers for people to talk for now, so both options are open for me
16:33:54 <zack> so it might be nice to have dedicated small talks to your work
16:34:08 <zack> that will allow *you* to do the talk :-), and also present more specific details if you want to
16:34:12 <zack> olasd: ok, thanks
16:34:21 <zack> olasd: thanks!
16:34:30 <zack> (crappy network)
16:34:34 <zack> anything else?
16:34:41 <orestis> that's all for me :)
16:34:44 <orestis> from*
16:34:55 <zack> cool, so let's adjourn
16:34:55 <zack> #endmeeting