12:01:10 #startmeeting 12:01:10 Meeting started Fri Feb 6 12:01:10 2015 UTC. The chair is zack. Information about MeetBot at http://wiki.debian.org/MeetBot. 12:01:10 Useful Commands: #action #agreed #help #info #idea #link #topic. 12:01:18 #chair matthieucan 12:01:18 Current chairs: matthieucan zack 12:01:29 #topic opw debsources - next meeting 12:01:32 next week, same time? 12:01:43 I am ok. 12:01:52 yes 12:02:19 #agreed next meeting next Friday, usual time 12:02:24 #topic opw debsources - weekly review 12:02:38 sophiejjj: your turn :) 12:02:43 I'm a bit worried about thef act we' 12:11:15 I think the refactoring is big enough already, without the mainlib/utils/db rewriting 12:11:15 yes. 12:11:15 what's your plan to address that problem? 12:11:16 hmmm 12:11:16 are you going to participate in gsoc? 12:11:17 well, maybe, but that's not an answer 12:11:18 your internship work should be done by the end of the internship. full stop 12:11:18 especially because if it is not, no one else could contribute to the code base 12:11:18 or it should take the risk of huge divergences 12:11:23 yes. that's my major concern. 12:11:30 so, here is what I think you should do 12:11:32 so I am hurrying on it these days. 12:11:50 split the refactoring work you're doing into (relative) small sub-parts 12:12:02 and push those sub-parts to us one by one, so that they're mergeable on their own 12:12:29 zack: I don't think it's feasible. 12:12:50 then you should reduce the aim of your refactoring work 12:13:41 it is IMHO very unwise to start a 1-month long refactoring work, in the last month of a 3 month internship 12:13:47 the risk is simply too high 12:13:52 first step, blueprints. second step, move methods from views. whynot third step, mainlib refactoring, after discussion, but small priority today 12:14:11 matthieucan: right, that's a nice partitioning of the work 12:14:58 I am now working on the blueprint though. 12:15:11 the thing matters is, the code is hard to merge in, I think. 12:15:18 unless it's totally done. 12:15:34 it's hard to merge, I think, because you're trying to do too many things at once 12:15:49 yes. that's right. 12:15:52 why couldn't the mainlib refactoring wait until the blueprint refactoring is done? 12:16:01 no. 12:16:07 sophiejjj: what do you mean? can't you add s blueprint system without touching views.py? I'm not talking of sharing methods with c.d.n yet 12:16:42 hmm. 12:16:50 but anyway the mainlib is not the major concern. 12:17:04 right, mine was just an example 12:17:05 the big workload is coverting methods to view and handle the related html. 12:17:27 to summarize: we cannot afford, at this stage, to wait several week before seeing a patch that is proposed for merge 12:17:45 we should have code that is potentially mergeable on a weekly basis, possibly more frequently with that 12:17:53 not because we're on a hurry, that's not the problem 12:18:05 but because we want to give you feedback as quickly as possible 12:18:15 to avoid that you're working on stuff that risk not being merged 12:18:20 does that make sense? 12:18:24 yes. 12:19:26 so, what do you think could be reasonably quickly mergeable into master of the current work? 12:20:05 no. 12:20:18 refactoring is a large work. 12:20:43 uhm 12:20:52 current code is based on the assumption of a single webapp 12:21:02 directly registered with the routes of debsources. 12:21:44 ok 12:21:48 finding a balance between "merging" and refactoring is kinda a dilemma. 12:22:12 by refacotring. 12:22:20 let's say that we're encouraging you to push that "balance" toward "the sooner we have something reviewable for merge, the better" 12:22:20 I am maintaing the logic of the current debsources. 12:22:27 if that means giving up on some side-goals, so be it 12:23:18 to be clear, I'm pushing in that direction because I do not want your work to go wasted, I really want to be able to integrate it into the codebase before the end of OPW 12:23:52 and sometime I've the feeling you're taking a path that is making it less likely to happen, rather than more likely to happen 12:24:09 zack: I totally understand you. 12:24:33 so, what do you want to set as a goal for next week? 12:25:07 or what about a mid-week goal? 12:25:16 the goal is hard to set, if I could think of one, I would tell you. 12:25:52 what about being able to plug a second blueprint? nothing about shared library, but only a flask blueprint 12:26:44 with this done, it'll be simpler to refactor, because we'll directly see what needs to be shared/hacked 12:27:08 matthieucan: I've done that on my code at gh. but considering merging into the current codebase is like dancing with the shackles. 12:27:24 nothing cherry-pickable? 12:27:57 hmm. 12:28:36 ok, so let me propose another way of approaching this 12:28:52 - please mail us a short summary of the goals of the current refactoring work 12:29:08 - and an estimation of how much you think of that work is already done, and how much still needs to be done 12:29:17 got it. 12:29:37 - then, during next week, maybe daily, we keep in touch (mail or IRC) to see how well the work is progressing 12:30:02 in the meantime we can at least review the work done thus far, and try to figure out if there are choices we need to change 12:30:13 matthieucan: I guess that'll be mostly up to you for the webapp part though ;) 12:30:36 yeah, sure 12:30:51 I can't commit on providing a daily feedback, but I'll do my best 12:31:23 great, thanks 12:31:38 sophiejjj: what do you think? would that "progress tracking" help you? 12:31:59 sure. 12:32:16 sophiejjj: anything else you'd like to add to next week objectives? 12:32:42 (or anything else you'd like to discuss before closing the meeting?) 12:33:05 currently no. 12:33:18 ok, let's adjourn then 12:33:21 #endmeeting