12:59:48 #startmeeting 12:59:48 Meeting started Fri Jul 3 12:59:48 2015 UTC. The chair is zack. Information about MeetBot at http://wiki.debian.org/MeetBot. 12:59:48 Useful Commands: #action #agreed #help #info #idea #link #topic. 12:59:51 #topic roll call 12:59:55 who is around? 13:00:00 * orestis 13:00:10 matthieucan sent regrets, so he won't be 13:00:15 zack: I'm here, but might suddenly disappear 13:00:15 jpleau: clemux: ping 13:00:20 oh, hello :) 13:00:33 hi everyone :) 13:00:49 hello 13:00:55 #topic next meeting 13:00:59 next week, usual time? 13:01:08 Next week i won't be able to attend at this time 13:01:09 ok for me 13:01:12 won't work for me at 3pm 13:01:25 well friday is pretty difficult for me next week all day 13:01:31 2 down 13:01:36 how about thursday next week then? 13:01:47 works for me 13:01:50 same for me 13:02:13 matthieucan: clemux: ? 13:02:16 ok for me 13:02:41 3 ok for thu, and matthieucan can't make it on fri anyhow 13:02:44 so let's settle for thu 13:02:55 #agreed anticipate next week meeting by 1 day, to next Thursday 13:02:58 sorry 13:03:01 yep, works for me 13:03:04 awesome 13:03:09 #topic weekly review - orestis 13:03:17 so, I should start this section apologizing 13:03:31 as I haven't yet had time to review PR#22 13:03:37 zack: pong sorry 13:03:54 PR#22 looks great 13:03:55 so the first 3 items on orestis' plate are still pending, and blocked on me 13:03:58 but is blocked by the testdata 13:04:18 matthieucan: oh, ok 13:04:23 so it "just" need testdata update now? 13:04:36 anyway, I'm on it 13:04:41 PR#22 looked good to me too, I checked each commits and commented here and there 13:04:46 yeah, and I still need to review it deeplier 13:05:01 I'll try to update the testdata soon then -- unless someone wants to pick it up 13:05:03 I haven't tested it though :) 13:05:06 I can probably do it over the week-end 13:05:07 hehe, it is not that of a blocker. i mean i am basing future work on that but its ok for now 13:05:12 TBH I have no idea how it works 13:05:44 it's like: apply code; run the updater on your real system; recreate the DB dumps (there's a makefile for that); fix all broken tests 13:05:51 (e.g., stats and the like) 13:06:35 anyway: orestis: what's next after PR#22? 13:06:45 PR#21 got merged 13:06:52 the views for the file-by-file apoi 13:06:56 api* 13:07:00 *nod* 13:07:03 yep, that was great 13:07:28 license statistics is almost finished but i can't PR anyway before PR#22 13:07:34 indeed 13:07:42 orestis: all is clear on the expected behavior now? 13:07:55 yeap 13:08:00 cool 13:08:09 what else? 13:08:16 that's all this week 13:08:28 so the 2 items mentioning PR#21 can go under reviewed, right? 13:08:29 i didn't have time for the other cards 13:08:32 yes 13:08:53 good work! 13:08:57 #topic orestis - next week 13:09:03 how about next week? 13:09:30 orestis: what about the doc? any improvement on that? 13:09:42 i should finish the license stats, work on the things i picked up last week and probably spdx 13:09:52 (afk) 13:10:14 looks good to me 13:10:24 i didn't have time for the other three cards on this week yet.. i spend much time on license statistics 13:10:49 please assign to yourself the relevant tasks 13:11:00 yeap doing this now :) 13:11:03 (also, next week will be shorter, as we meet on thu) 13:11:22 #topic clemux - weekly review 13:11:27 clemux: your turn! 13:12:10 first task: add_package hook for hello world plugin 13:13:13 (clemux is having network issues, apparently) 13:14:52 since the plugin doesn't do much, it was easy, but now the call_hooks tasks is ready for other plugins 13:15:02 he's back \o/ 13:15:04 *nod* 13:15:36 next task, which I added during the week because I spent some time on it: how to make pluggable celery tasks 13:15:49 the documentation was merged in PR #23 13:16:09 and I updated the hello world plugin, which does not depend on the updater's celery app anymore, in my branch 13:16:18 ok, cool 13:16:59 anything else? 13:17:03 next task was supposed to be "update tests for testing async updater", but first I had to make the tasks pass with my changes 13:17:15 yeah, the fight with tests :) 13:17:26 I should have done that earlier, when I made the changes to db_storage and fs_storage, but I didn't 13:17:40 so, I fought with the test suite yesterday and this morning 13:17:51 did you win? 13:17:58 :P 13:17:59 I still couldn't run everything, but I worked around it and disabled the webapp tests 13:18:36 and I fixed the tests, the sync updater and db/fs_storage (trello task "make updater pass unit tests (changes to db_storage, fs_storage, debmirror) 13:18:41 ") 13:18:46 cool 13:18:59 so what stages / plugins still need porting to async now? 13:19:08 so I've started working on creating a test suite for the sync updater 13:19:20 that's it for the week 13:19:27 (you mean "async", right?) 13:19:34 async, sorry 13:19:38 ok, cool 13:19:57 #topic clemux - next week 13:20:01 what's next now? 13:20:37 1) finish the test suite for the async updater 13:21:01 2) debug/fix a problem in the extract_new stage: the file_table retrieved from the DB is always None 13:21:20 oh, ok, that's bad indeed 13:21:47 back to my previous question: how many stages / plugins still need porting to async? 13:21:50 3, 4, 5, 6) port the 4 plugins to celery tasks 13:22:05 all the other stages are done already? 13:22:10 zack: no :( 13:22:19 I only worked on the extract_new stage 13:22:23 ok 13:22:36 so you prefer working on the plugins than on the other stages, right? 13:22:42 (it's to understand the status, no worries :)) 13:22:50 yes 13:22:55 fair enough 13:23:06 though I'll update the rm_package hooks at the same time, since it's mostly the same thing 13:23:19 (well, make them celery tasks, not test them in depth) 13:23:22 looks good to me 13:23:41 please fiddle with the tasks on trello as needed 13:23:42 I think the update_suite and maybe the gc stages could be done next week 13:23:49 but I've always been too optimistic until now 13:24:00 indeed, I was about to say: let's avoid overcommitting 13:24:25 anything else before we move to misc? 13:24:39 I should add one task though: make the async updater use the desources configuration 13:24:52 uh? what does it use now? 13:24:54 I've starting preparing for that, using structures similary to the configuration 13:24:59 similar 13:25:08 but it's still hard-coded 13:25:20 oh, and that's was part of the test case failures? 13:25:35 but yes, definitely, it should use the same conf 13:25:41 maybe you should pass it as a parameter of the jobs 13:25:48 so that the central updater sends it to the workers 13:25:49 yup 13:25:57 that way we know they're all on the same page 13:26:16 I suggest prioritizing this task then 13:26:21 it looks pretty important to me 13:26:25 got it 13:26:36 ok, let's move on 13:26:37 #topic misc 13:26:50 free for all! 13:26:55 nothing from me 13:27:04 hmm 13:27:26 jpleau: what's brewing? :) 13:27:37 For the rendering of a file's license, was it planned to rework the rendering (the design) 13:28:16 jpleau: if there is something you don't like, certainly ;-) 13:28:17 orestis: ? 13:28:50 It works great, I wasn't sure about how the folding was laid out particularly, I can make a list of suggestions and make a github issue 13:28:53 jpleau: i am not sure i understand. i plan on redoing the links for the license synopsis 13:29:08 that's all i am aware of 13:29:19 jpleau: by all means, if you've improvements to suggest, please report them 13:29:51 sure :) 13:30:04 anything else? 13:30:47 I'm fine 13:31:01 let's adjourn then 13:31:33 #endmeeting