=== salgado-afk is now known as salgado === salgado_ is now known as salgado [15:00] #startmeeting [15:00] Meeting started at 09:00. The chair is barry. [15:00] Commands Available: [TOPIC], [IDEA], [ACTION], [AGREED], [LINK], [VOTE] [15:00] hello and welcome to this week's ameu reviewers meeting. who's here today? [15:00] me [15:00] ni! [15:00] me [15:00] me [15:00] me [15:00] me [15:00] me [15:00] yup [15:00] me [15:00] me [15:01] me [15:01] intellectronica sends his apologies [15:01] adeuring: ping [15:01] me [15:01] me [15:01] cprov, danilos ping [15:02] me [15:02] gary_poster: ping [15:02] salgado: ping [15:02] barry: thanks [15:02] me [15:02] me! [15:02] [TOPIC] agenda [15:02] New Topic: agenda [15:02] * Roll call [15:02] * Wednesday AMEU hard stop, barry [15:02] * Action items [15:02] * Mentoring update [15:02] * Peanut gallery (anything not on the agenda) [15:03] i have a hard stop at 14:45 today. hopefully we won't go that long anyway :) [15:03] [TOPIC] action items [15:03] New Topic: action items [15:03] * adeuring to update wiki pages regarding `except Exception`, `KeyboardInterrupt` and `SystemExit` [15:03] done [15:03] adeuring: thanks! [15:04] * gary_poster to take importfascist and rSP() discussion to ml [15:04] nope. I will bring new energy to the task. ;-) [15:04] gary_poster: cool. keep it on the list then? [15:04] y [15:04] +1 [15:04] me [15:04] intellectronica's not here today, so... [15:05] [TOPIC] * Mentoring update [15:05] New Topic: * Mentoring update [15:05] lessee, i think we're missing both deryck and leonardr [15:05] any word from their mentors? how are things going? [15:06] barry, things is going well. [15:06] leonardr_: hi! [15:06] Very well. [15:06] hi barry [15:06] Speaking of which... [15:06] rockstar: great, thanks for mentoring. leonardr_ any comments, problems, feedback on reviewing? [15:07] gmb: what's up? [15:07] barry: Oh, nothing, I just realised deryck wasn't here :) [15:07] :) [15:07] deryck: how's the reviewing going? [15:08] barry: FTR, deryck's doing a fine job. He'll graduate in no time. [15:08] barry, I think going well, if I could remember to show up for things. :) [15:08] gmb, deryck excellent [15:08] deryck: isn't it almost your eod? :) [15:08] barry: i was hoping for some more interactive mentoring. on thursday i was reviewing a complicated branch and both rockstar and the person who submitted the branch were afk. i did the review but i don't think i learned as much from it as i could have [15:08] barry, not quite. :) [15:09] leonardr_: you can always ping me if needed [15:09] barry, ok, cool [15:09] anything else on mentoring? [15:09] leonardr_, I was afk? Hm. I don't remember that. I'll make sure to be around. [15:09] leonardr_: if the submitter is not around, drop the review if you have questions [15:10] I don't think dropping the review is right. I think you can respond without voting. [15:10] or just ask all those questions in the review! [15:10] if it was requested as an OCR, it goes to the back of the queue until that person shows up [15:10] that's what i did. i did other reviews which were easier, and then i did this one [15:11] these are the rules we agreed on a long time ago [15:11] bigjools: +1 [15:11] and i asked my questions in the review [15:11] but leave a question in the review and mark it needs-reply [15:11] the reason it was difficult is it took me a long time to understand the code well enough to formulate the questions [15:11] bigjools, hm, I guess you have different rules than I, but my team spans more timezones than yours. [15:11] leonardr_, was this cprov's branch? [15:12] rockstar, yeah [15:12] we should reiterate that people requesting an OCR need to be around or negotiate their absense with the reviewer [15:12] leonardr_: that'll happen at first, don't worry about it too much [15:12] * bigjools is shocked that cprov was not around :) [15:12] bac, why do they need to be around? [15:12] leonardr_: yeah, don't feel too bad about doing another review [15:12] instead of that one [15:13] rockstar: OCR is supposed to be interactive. [15:13] leonardr_, I was around last week. Remember, we got swamped? [15:13] rockstar: so the review can be interactive. that was one of the principles of having OCR [15:13] leonardr_: I thought it was perfectly fine to jump to another review when I was out [15:13] rockstar: they need to negotiate their absence with the reviewer, who might not want to take a particular branch if the submitter isn't around to answer questions in realtime [15:13] this is written up on the OCR page I thought [15:13] barry, okay, I think "negotiate with the reviewer" is better than "drop it until they come back" [15:14] leonardr_: there was enough time for finishing the review while I was around, but you had problems with your account, remember ? [15:14] cprov: i did [15:14] rockstar: right. but if the submitter just disappears, the reviewer has every right to move on [15:14] exactly [15:14] i don't blame anyone for not being around at my convenience. and i did move on--i reviewed the other two branches in the queue before coming back [15:14] barry, yes, because it wasn't negotiated. I'm happy with that. [15:14] leonardr_: right, I think when something like this happens there isn't much you can do, abstain and move on, no worries. [15:15] cool [15:15] cprov: yep [15:15] ok, in the future i'll abstain and do other work rather than bang my head against a branch all day [15:15] leonardr_: thanks for bringing this up [15:15] np [15:16] [TOPIC] * Peanut gallery (anything not on the agenda) [15:16] New Topic: * Peanut gallery (anything not on the agenda) [15:16] does anybody have any other topics today? [15:17] that seems like a "no" :) [15:17] barry: just to mention: [15:17] henninge: go ahead [15:17] I changed my OCR slot to Monday Euro. [15:17] barry, do we have a policy on multiline list comprehensions? [15:17] * bigjools thought of something [15:17] henninge: thanks. did you update the wiki page? [15:17] barry: yes [15:18] rockstar: i believe we do, in our python style guide [15:18] henninge: thanks! [15:18] bigjools: go ahead [15:18] ok [15:18] leonardr_, another option is to use kiko's approach and ask tons of questions on the review. once the developer answers them you can do a proper review [15:18] in a fix I made this week I had a hellish time dealing with code that was a mix of unicode and ascii strings [15:18] Bjorn had some interesting post-review points [15:19] but the upshot is that the code should be using unicode strings throughout, up until the point where you have to encode it as utf [15:19] is this enforceable? [15:19] salgado, true [15:19] across LP I mean [15:19] salgado: yes, that would work too, but unnecessarily binds the reviewer to the branch. [15:19] bigjools, it'd be nice. [15:20] bigjools: In bzr, we often hold strings as utf-8-encoded bytes, for performance reasons. [15:20] when we're on python 2.6, i really want to add "from __future__ import unicode_literals" to the header of every file [15:20] abentley: interesting, how much of a performance hit is there? [15:20] (and absolute_import but that's a different issue) [15:20] bigjools: dunno. And I know that bzr manipulates a lot more data than your typical Launchpad page. [15:21] right [15:21] abentley: bzr is probably more byte-oriented than launchpad, which is more text oriented [15:21] it probably won't make much difference for us, we have bigger performance problems to worry about [15:21] barry: Lots of things are defined as unicode, like filenames and revision-ids. [15:22] abentley: good point about filenames, didn't know about revids [15:22] the thing I was fixing were error messages in our upload processor, as soon as an ascii string formatter pulls in some unicode from the database, you're in trouble [15:22] i do think enforcing unicode literals will help, but it will probably expose places where we are being sloppy [15:23] barry: won't that means a 100,000 line diffs? [15:23] it might need a larger concentrated cross-team effort on it at some point [15:23] barry: to update the expected output of doctests? [15:23] flacoste: not if people are printing string values! [15:23] barry: they are not [15:24] barry: just grep for u' in the tests [15:24] barry: Python's willingness to convert ascii-encoded str into unicode is convenient, but can mask problems later. [15:24] flacoste: it's our rule (now) but yeah there's legacy code [15:24] abentley: yep [15:25] anyway, we're not on python2.6 so it's kind of a moot discussion, but it should be a goal of ours to be explicit in our code whether we're talking strings or bytes [15:25] (where byte literals are defined with b'' prefix) [15:25] it won't be painless though ;) [15:26] anyway... [15:26] anything else on this or other topics for today? [15:26] my immediate question is, it might be worth trying to enforce unicode now in reviews to reduce the potential for more problems. What do you think? [15:27] bigjools: possibly. i hate to introduce more u'' prefixes when we can soon make them all go away [15:27] bigjools: otoh, we can probably convert module-by-module [15:28] I guarantee there will be lots of pain unless it's done all at once [15:28] but ok, let's move on [15:28] bigjools: which isn't to say you guarantee it will be painless if it is all done at once :) [15:28] there's relative levels of pain :) [15:29] :) [15:29] okay, anything else? [15:29] 5 [15:29] 4 [15:29] 3 [15:29] 2 [15:29] 1 [15:29] #endmeeting [15:29] Meeting finished at 09:29. [15:29] thanks everyone! [15:29] barry: thanks yous [15:30] Cheers barry [15:30] cheers === salgado_ is now known as salgado === salgado is now known as salgado-lunch === salgado-lunch is now known as salgado === salgado is now known as salgado-afk [23:34] barry: hello [23:35] #startmeeting [23:35] Meeting started at 17:35. The chair is barry. [23:35] Commands Available: [TOPIC], [IDEA], [ACTION], [AGREED], [LINK], [VOTE] [23:35] mwhudson: hi [23:35] thumper: hi [23:35] jml: hi [23:35] hi [23:35] well, not much to recap from ameu today [23:36] so i'll just open up the floor to anything you guys have [23:36] I don't have anything [23:36] oh [23:37] me neither [23:37] mwhudson: record breaking [23:37] i guess we're done then :) [23:37] w00t [23:37] #endmeeting [23:37] Meeting finished at 17:37. [23:37] * mwhudson wonders what the relationship between number of attendees and length of meeting is [23:37] i think it's not linear :) [23:37] exponential [23:38] you don't even want to know how long the meeting is when i'm the only one there [23:38] * barry -> more dinner