[01:00] lifeless, thumper, rockstar, mwhudson, StevenK: ping [01:00] #startmeeting [01:00] Meeting started at 19:00. The chair is bac. [01:00] Commands Available: [TOPIC], [IDEA], [ACTION], [AGREED], [LINK], [VOTE] [01:00] here [01:00] * rockstar [01:00] hi guys [01:01] hi [01:01] thumper: did you see that article about the dunedin psycho murderer? [01:01] only here vaguely [01:01] bac: ah... no [01:01] when? [01:01] it's in the New Yorker magazin [01:01] i threw a link up on FB [01:02] hi [01:02] [topic] agenda [01:02] New Topic: agenda [01:02] * Roll call [01:02] * Agenda [01:02] * Outstanding actions [01:02] * New topics [01:02] * Mentat update. [01:02] * Some of the UI reviewers have graduated, but ReviewerSchedule indicates that no one has graduated. Same for UI/Reviews, which still lists people who've left the team. [01:02] * https://dev.launchpad.net/ArchitectureGuide as per the epic - will reviewers please start discussing the values and metrics during reviews? [01:02] * Assertion errors and the webservice (gary) [01:02] * Peanut gallery [01:02] i guess we're all here but steven [01:03] [topic] mentat update [01:03] New Topic: mentat update [01:03] thumper: has steven been getting any more reviews? [01:03] not that I've noticed [01:03] but it has been week 3 [01:03] i made an appeal to the AMEU people to direct some towards him if they weren't pressed for time [01:03] * thumper nods [01:03] we'll see if that happens. i'm not optimistic it will. [01:04] heh [01:04] everything is urgent [01:04] especially with team leads breathing down your neck to move those kanban cards... [01:04] thumper: well, do what you can... [01:05] i mentioned lifeless had thoughts about changing the metric for branch size. i only meant to introduce it but the idea generated lots of reaction [01:05] so, lifeless, i apologize for not getting up with you last week to form a proposal as we agreed. [01:06] bac, what kinds of reactions? [01:06] * rockstar COULD read the backchat, but is lazy. [01:07] "too vague", "not a metric" [01:07] nothing terribly hostile [01:08] bac, the backchat indicates no interest in change at all, which is disheartening to me. [01:08] i tried to introduce robert's idea: https://dev.launchpad.net/ArchitectureGuide as per the epic - will reviewers please start discussing the values and metrics during reviews? [01:08] rockstar: well, it was not presented properly and i think there were just some knee jerk reactions [01:08] bac, okay. [01:08] * bac hopes he didn't poison the well [01:09] bac, as for me and my house, I will keep my diffs small. [01:09] lifeless: i think we'll need an email about what you'd like to see wrt the ArchitectureGuide [01:09] :) [01:09] rockstar: your wife starting to hack launchpad? [01:10] bac: no, that is choco [01:10] finally gary raised a point about assertion errors and the API [01:10] he has his own facebook account don't you know [01:10] rs=bac for any work by choco [01:10] bac, assertion errors and the API is also something I wonder about. [01:10] choco a fine LSD if ever there was one [01:11] bac: huh, why? [01:11] gary and benji have been looking at exception handling and the API. they want to put together a white list of acceptable exceptions for the API. [01:11] bac: what I wrote is what I want [01:11] work to be forthcoming [01:11] bac: I can send a mail, but its all there already [01:12] in the meeting log from last meeting, which I saw you edit on the wiki :P [01:12] bac, a whitelist? Really? I don't like that idea. [01:12] lifeless: i think an email hitting the hightlights will be more effective than just posting a URL for people to read. [01:13] bac, for instance, we raise BranchMergeProposalExists in the case where someone proposes a merge against a branch where one already exists. [01:13] I think that's much more helpful than say, an assertion error. [01:14] lifeless: i like what you're encouraging, but i think a lot of reviewers have a fixed way they work, certain things they look for, etc. IMO it'll take some effort to change habits. [01:14] rockstar: yes, i think that's the direction gary is headed. [01:14] rockstar: that's much better than raising an AssertionError [01:15] bac, okay. My concern is that, for a whitelist, when we add an exception, we have to add it to the whitelist. [01:15] Currently, to make it API-compatible, we just annotate the exception with webservice_error() [01:16] rockstar: valid points you can raise when gary makes his proposal. this was just a head up, i think. [01:16] bac, ah, okay. [01:16] bac: perhaps we can refine it here then [01:17] bac: the idea is that they - check the metrics; discuss the patch in the context of the values & goals on that page. [01:17] rockstar: gary and benji also made the reminder that assert() shouldn't be used as it can be optimized away... [01:17] bac, yeah. assert() is less helpful than an AssertionError [01:18] FWIW assert and -O aren't relevant for us, though I agree in principle (and we don't use assert() for that very reason in bzr) [01:18] it is (or at least was) also surprisingly common for people to break lines by doing [01:18] assert (condition, [01:18] message) [01:19] mwhudson, ew. [01:19] anyway, there seems to be a lot of talk here about partially presented ideas :-) [01:19] mwhudson: yeah, i used to see syntactically correct but completely wrong asserts a lot [01:21] bac, it looks like we're summing up here. I have a proposal. [01:22] lifeless: i agree with your ideas regarding the values and goals. [01:22] (a full proposal) [01:22] lifeless: i just don't know how to make it happen, other than what you're already doing, which is very helpful. [01:22] rockstar: sure [01:23] bac: ok, so I'll send a short message asking for this. [01:23] bac, I propose that we abolish the "javascript" review, and just make it a regular code review. [01:23] bac: and perhaps you could follow up in the review team meetings asking whether folk are finding this to work well? [01:23] lifeless: certainly [01:24] bac, I'll be updating the JavascriptReviewGuidelines a bit more, as I've learned a bunch about writing high performance javascript recently, and our code makes me cringe. [01:24] I think we could make everyone a javascripter if they had to review javascript and have things explained to them. [01:25] +1 to rockstar's proposal [01:25] rockstar: i look forward to hearing what you've learned. [01:26] bac, there will be an email today or tomorrow. [01:26] becoming more competent in JS is one of my goals... [01:26] (and I'll be participating in performance Tuesday from now on) [01:27] is there anything else you'd like to talk about? [01:27] Not from me. [01:28] thumper: when you see StevenK would you remind him of the meeting time? [01:28] bac: ack [01:28] thanks guys. [01:28] #endmeeting [01:28] Meeting finished at 19:28. === mrevell is now known as mrevell-lunch === Ursinha-afk is now known as Ursinha === mrevell-lunch is now known as mrevell === benji is now known as benji-lunch === Ursinha is now known as Ursinha-lunch === benji-lunch is now known as benji === Ursinha-lunch is now known as Ursinha [22:09] OOPS-1713EB2282 [22:09] https://lp-oops.canonical.com/oops.py/?oopsid=1713EB2282 === Ursinha is now known as Ursinha-afk