12:59:54 #startmeeting 12:59:54 Meeting started Fri Jun 26 12:59:54 2015 UTC. The chair is zack. Information about MeetBot at http://wiki.debian.org/MeetBot. 12:59:54 Useful Commands: #action #agreed #help #info #idea #link #topic. 12:59:58 #topic roll call 13:00:01 wh ois here? 13:00:06 hi everybody! 13:00:12 jpleau sent regrets, so won't be around 13:00:15 hello 13:00:23 I'll be here in ~1min 13:00:37 matthieucan: ack 13:00:40 * orestis 13:00:40 orestis: ping 13:01:29 ok 13:01:34 #topic next meeting 13:01:45 next meeting next week at the usual time is good for everyone? 13:01:56 good for me 13:02:01 good for me 13:02:04 for me, it should be good ~90% 13:02:05 yep 13:02:14 matthieucan: if I can't make it, can you chair? 13:02:24 zack: sure 13:02:27 ok then 13:02:34 #agreed next meeting next friday, usual time 13:02:41 #topic weekly review - orestis 13:02:49 orestis: table is yours :) 13:03:28 pr#12 and pr#19 were merged.. (doc for the batch api) 13:03:38 ack 13:04:08 batch api was merged as well. 13:04:13 PR#18 13:04:23 ack too 13:04:27 (i'm moving to reviewed as we speak) 13:04:35 ah ok cool 13:04:52 so the views are under review i think 13:04:55 there were 2 tasks in pr#12, right? 13:05:20 yes i had to update the file-by-file url schema 13:05:27 ok, cool 13:05:33 the views are in which PR? 13:05:42 PR#21 13:05:54 matthieucan: are you on it? 13:06:11 zack: yep, in my todo list 13:06:15 cool 13:06:16 sorry, didn;t have much time this week 13:06:22 no worries, same here :) 13:06:37 np it is not blocking anything anyway 13:06:40 orestis: then we discussed the DB structure for license stuff this morning 13:06:49 but I guess there was no task associated to it 13:06:50 orestis: ok great :) 13:06:58 ah, no, sorry 13:07:03 there is 13:07:04 it was the DB schema 13:07:15 but you were waiting for my feedback, right 13:07:34 so we can already move to next week planning, right? 13:07:44 nope just a sec 13:07:48 sure 13:08:34 in this branch https://github.com/oorestisime/debsources/tree/copyright-db i have te changes in the model, the migration script and the license hook. if i open a PR travis will fail. 13:08:59 right, because the DB tests will fail, is that it? 13:09:03 yes 13:09:11 should i open a PR anyway to review the code? 13:09:13 so, just leave a note in the PR text about that 13:09:16 yes please 13:09:22 ok works 13:09:25 I'll review, keeping in mind that it's normal that travis fails 13:09:51 anything else? 13:09:55 and last thing is the authors.. we said today that we won't return the authors 13:10:00 correct 13:10:06 what should i do with that card? 13:10:22 put it in the backburner 13:10:27 we don't want to do it for now 13:10:30 but in the future ... maybe 13:10:38 ok works 13:10:46 and yes that's all for this week 13:10:48 cool 13:10:56 #topic next week - orestis 13:11:02 what do you want to pick up for next week? 13:12:01 since we will have the new table maybe i should refactor the API to use these entries instead of parisng the d/copyright. 13:12:24 sounds good 13:12:36 then i think ninka is interesting to do and the license statistics 13:12:47 except if spdx is more important 13:12:54 the plugin part is done already? 13:13:06 because first we will need to get the data into the DB 13:14:33 i thought plugin = hook like the sloccount or ctag hook 13:14:39 yes, that's correct 13:14:42 there's more to it? 13:14:45 nope 13:14:53 is that in the branch you mentioned above? 13:14:57 yes 13:15:21 ok 13:15:46 so, I'd rather do the stats first 13:15:49 not that ninka is not interesting 13:15:50 bug 13:15:51 but 13:16:02 I've just completed running ninka on all the archive with a collegue for research reasons 13:16:09 and I'd like to first have a look at those data 13:16:14 and understand the needed db structure 13:16:20 before implementing something similar in debsources 13:16:28 so let's postpone ninka for now 13:16:31 license stats look good 13:16:36 ok then 13:16:45 and if you think you will have extra time, yes, next step would be spdx 13:16:49 (up to you) 13:17:17 well there also small tasks to be done, like the base templates, the documentation for the copyright BP 13:17:26 ok, perfect then 13:17:45 just note down those tasks in trello, ok? 13:18:07 ok so i ll focus first on using the db to retrive the licenses, then the statistics and then some minor tasks here and there 13:18:22 yeap ok i ll do that now 13:18:25 cool 13:18:28 let's move to clemux then 13:18:33 #topic weekly review - clemux 13:18:37 clemux: your turn :) 13:18:57 first thing: my tasks were not well defined this week 13:19:13 that's bad™! :) 13:19:18 "base: extract new stage" was not granular enough 13:19:33 oh, I though you meant the tasks in trello -- never mid 13:19:34 mind 13:19:44 *nod* then 13:20:10 and "update prototype according to specifications" at this point is basically "do all the other tasks" 13:20:48 so, I worked on two celery tasks this week 13:21:10 ok 13:21:25 extract_new: which I have updated to use chords (parallel tasks with a callback) 13:21:47 and calls the other task: add_package, which I've been stuck on for some time because of the dumb mistake we talked about yesterday 13:22:01 yup (that's fixed now, right?0 13:22:02 ) 13:22:21 it now extracts packages and fills the database 13:22:27 yay 13:22:40 I should have commited and made a pull request, but I was confused by my badly defined trello tasks 13:22:51 I'll do that after the meeting 13:23:06 fair enough 13:23:19 next problem: I was having trouble running a celery task on a specific worker, for calling hooks on the same machine 13:23:30 (http://stackoverflow.com/questions/31073110/celery-how-to-use-celery-utils-worker-direct) 13:23:34 I fixed that a few minutes ago 13:23:47 is that for testing purposes, or for the "shortcuts" when workers are local? 13:24:12 no, it's required for hooks, since they need to access the extracted sources 13:24:18 oh, I see 13:24:48 it basically solves the problem of sending large data set over the network, since we /must/ run them on the same machine anyway 13:24:59 right 13:25:08 ok 13:25:11 anything else to report about? 13:25:23 that's it for the week 13:25:36 #topic next week - clemux 13:25:39 so, next steps? 13:26:36 for next week: 13:26:45 - update plugins for use in celery tasks 13:27:33 (from what I've seen, it means removing the DB access from the plugins, and make them return data either as local files, or as celery messages) 13:27:43 so, am I understanding it right that you want first to do the add part completely, and only later look into GC ? 13:27:47 (I' 13:27:54 (I'm ok with that, it's too better understand the path here ) 13:28:03 yes, to be sure we the protocol is well designed 13:28:09 sounds reasonable, yes 13:28:16 since it's the more tricky part 13:28:59 next (not sure on the order) : "plugin_system: add package hook" (continue working on what I've started this week) 13:29:10 yep 13:29:19 having a working add package + plugins would be awesome 13:29:44 might be made more granular by adding a task "port hello word plugin to celery updater" or something 13:29:57 that would be fun :) 13:30:11 also: I will need to write tests from now on 13:30:37 so I need to update the tests to allow testing the async updater 13:30:48 *nod* <- vigorously 13:30:56 setting the celery configuration key for running all tasks synchronously, etc. 13:31:12 already added to the trello, I can move it to the "this week" list 13:31:19 ok, sounds about right for next week 13:31:28 anything else before we move to misc? 13:31:35 that's it 13:31:38 #topic misc 13:31:45 open grounds for any additional topic? 13:31:56 s/?/./ 13:32:16 nothing from me 13:32:16 oh, one question: when I'm stuck on something for a long span of time 13:32:26 should I create a new task for that specific problem, so it's trackable? 13:32:33 it's useful, yes 13:32:40 also, if it's too long, feel free to ping me/us 13:32:45 yup 13:33:02 it sucks if you (any of you) stays blocked on a given problem for ~1 day 13:33:11 that's already a good deal of your week, so... 13:33:21 but yeah, the task helps reviewing what you've done 13:33:35 and helps you not feeling you've haven't done anything (which is not true) 13:33:42 anything else? 13:33:47 that's it, thanks :) 13:34:08 ok then, let's adjourn 13:34:12 #endmeeting