[15:00] <bac> #startmeeting
[15:00] <MootBot> Meeting started at 09:00. The chair is bac.
[15:00] <MootBot> Commands Available: [TOPIC], [IDEA], [ACTION], [AGREED], [LINK], [VOTE]
[15:00] <jelmer> me
[15:00] <bac> hello, who is here today?
[15:00] <abentley> me
[15:00] <bigjools> me
[15:00] <sinzui> me
[15:00] <jcsackett> me
[15:00] <benji> me
[15:00] <mrevell> me
[15:00] <deryck> me
[15:00] <EdwinGrubbs> me
[15:00] <henninge> me
[15:00] <flacoste> me
[15:00]  * benji things it should be "I" :P
[15:00] <mars> me
[15:01] <mars> benji, "moo" is always fun
[15:01] <benji> heh
[15:01] <salgado> me
[15:02] <gary_poster> me
[15:02] <danilos> me
[15:02] <gmb> Me
[15:02] <allenap> me
[15:02] <danilos> oh, capital "Me" is here as well
[15:03] <bac> great, let's start.  pretty light agenda today
[15:03] <bac> [topic] agenda
[15:03] <MootBot> New Topic:  agenda
[15:03] <bac> * Roll call
[15:03] <bac>  * Agenda
[15:03] <bac>  * Outstanding actions
[15:03] <bac>  * Mentat update.
[15:03] <bac>    * Salgado (ui)
[15:03] <bac>    * StevenK (code)
[15:03] <bac>    * MRevell (ui)
[15:03] <bac>  * New items
[15:03] <bac>    * Should reviewers expect the review template to be used? - abentley
[15:03] <bac>    * Integrating test timing into reviews, recap.  -bac
[15:03] <bac>  * Peanut gallery
[15:03] <bac> [topic] mentat update
[15:03] <MootBot> New Topic:  mentat update
[15:03] <bac> salgado, getting any more UI reviews?
[15:03] <gary_poster> bac, benji should be in mentat update
[15:04] <salgado> nope
[15:04] <bac> mrevell, getting any?
[15:04] <bac> thanks, gary_poster
[15:04] <mrevell> bac, No.
[15:04] <salgado> haven't done any in a long while
[15:04] <jelmer> I will have one tomorrow
[15:04] <jcsackett> bac: i'm in the mentoring process as well.
[15:04] <danilos> I've also asked henninge to be my UI mentor as well, so I'll be starting as UI reviewer one of these days as well
[15:04] <sinzui> salgado, me too
[15:04] <bac>  sorry benji and jcsackett.  updated on the wiki now
[15:04] <benji> np
[15:04] <danilos> perhaps there's no need for UI reviewers anymore
[15:05] <flacoste> danilos: why?
[15:05] <sinzui> salgado, may be should agree that mentoring ui <= 3 month
[15:05] <danilos> flacoste, why would I want to be, or why there might not be a need?
[15:05] <sinzui> salgado, maybe you should graduate next week because you have all the experience you are ever going to get
[15:05] <danilos> flacoste, it's just that people haven't been getting any UI branches for review
[15:05] <abentley> danilos, there's certainly a need for UI review, but maybe not one so frequent as to demand specialized reviewers.
[15:05] <bac> sinzui: +1
[15:06] <EdwinGrubbs> benji has been doing a good job, but we haven't been getting very many requests for reviews on Wednesday
[15:06] <salgado> sinzui, well, I'm not sure a time limit is a good idea as in some periods (like now) we may stay a long time without doing any reviews
[15:06] <danilos> flacoste, i.e. mrevell and salgado hasn't gotten any in a week, so I am wondering if it's smart to try to achieve that specialization with unused workforce we've got
[15:06] <jcsackett> EdwinGrubbs: i've found the same thing on thursdays--i think maybe it's an end of year slow down?
[15:06] <flacoste> jcsackett: probably is
[15:07] <henninge> I have not had that many, either.
[15:07] <bac> jcsackett, flacoste: but with the BugJam on perhaps it'll pick up?
[15:07] <salgado> danilos, fwiw, I was doing 3 or so reviews a week when I started.  just lately there doesn't seem to have been many people working on UI
[15:07] <abentley> fwiw, monday OCR is rarely busy.
[15:07] <flacoste> bac: number of reviews should yes
[15:07] <flacoste> salgado, danilos: yeah, UI reviews usually come in burst
[15:07] <flacoste> when a new feature is developped
[15:08] <danilos> salgado, right, thanks; I don't want to become a mentat as well and thus distribute the low UI activity over more people
[15:08] <bac> danilos: perhaps you have a point that we don't need new UI reviewers in the pipeline if we can't get enough to get the current mentats trained.
[15:08] <danilos> bac, right, that was my point
[15:08] <jcsackett> bac: maybe get a queue of people willing to be UI and phase them in when the next feature rush begins?
[15:08] <abentley> danilos, Sorry, I thought you meant the whole concept was outdated.
[15:09] <bac> abentley: yeah, me too
[15:09] <danilos> abentley, oh no, sorry for confusing you guys :)
[15:09] <bac> [topic] Should reviewers expect the review template to be used? - abentley
[15:09] <MootBot> New Topic:  Should reviewers expect the review template to be used? - abentley
[15:10] <abentley> Back when I  started, there was a template that was expected to be used for code review.
[15:10] <abentley> It's basically preserved in the lpreview_body plugin.
[15:10] <abentley> It expects a summary, pre-implementation notes, implementation details, lint, etc.
[15:11] <abentley> I use it, but I find that basically no one else does.
[15:11] <bac> abentley: i always use it and really like it when others do.
[15:11] <abentley> Okay, maybe that's too strong.
[15:11] <jcsackett> abentley: that template was provided to me when i started. i've seen it from a few others (though sometimes not with all sections).
[15:11] <bigjools> I use it sometimes.  It's massive overkill for simple branches.
[15:11] <abentley> There are a bunch of people not using it.
[15:11] <bac> personally i like it because i'm both lazy and forgetful.  it helps with both.
[15:11] <deryck> I don't use it.  there I owned up to it. ;)
[15:12]  * deryck looks around at the rest of the room
[15:12] <deryck> I feel it's too prescriptive.
[15:12] <bac> bigjools: the sections that are overkill are easily deleted or marked 'n/a'
[15:12] <abentley> bac, +1
[15:12] <jcsackett> deryck: could you unpack that a bit? not sure what you mean.
[15:12] <bigjools> that's extra hassle, particularly if I use the web ui
[15:12] <flacoste> actually, web ui is the major problem there
[15:12] <bac> bigjools: oh.  i *never* use the web ui...
[15:12] <flacoste> i most often use the web ui to submit branches
[15:13] <flacoste> (ok, i don't submit that many anymore...)
[15:13] <jcsackett> i use the web ui, and find that pasting in the template as a starting point isn't that big a deal.
[15:13] <gary_poster> more people may be using it soon, if the switch to using tarmac is successful
[15:13] <flacoste> i'd use the template
[15:13] <gary_poster> it == the web
[15:13] <flacoste> if it was easy to get at
[15:13] <abentley> gary_poster, why?
[15:13] <flacoste> gary_poster: why?
[15:13] <bigjools> if the branch is cleaning up, or a trivial bug, I find that template too prescriptive, annoying to edit and a waste of time.
[15:13] <gary_poster> because you'll be able to submit without using a commandline if you want
[15:13] <bigjools> however, it's useful for a more complex change
[15:14] <abentley> gary_poster, I think you are talking about lp-land, not lp-propose.
[15:14] <jelmer> It'd be nice if it the template was available in the web interface as well, not just in "bzr send".
[15:14] <gary_poster> oh, you are right, abentley.  thanks, sorry
[15:14] <jelmer> I'm also guilty of not using the template now that I've switched to using the web interface primarily for proposing merges.
[15:15] <abentley> jelmer, it's not just in bzr send, it's also in lp-propose, and that's preferred.
[15:15] <gary_poster> not using template: me too.
[15:15] <abentley> jelmer, I don't see how the web site is going to run lint on your local machine :-)
[15:15] <bigjools> I stopped using lint when it kept coming up with a million* false positives
[15:15] <abentley> Personally, I think it's a useful reminder of key things.
[15:16] <abentley> Like who the pre-implementation call was with.
[15:16] <abentley> I agree it's overkill for trivial bugs.
[15:17] <bigjools> pre-imp details are the most useful thing on that template
[15:17] <abentley> bigjools, it's pretty good for me now.
[15:17] <abentley> bigjools, lint is pretty good for me, I mean.
[15:17] <benji> abentley: indeed; I use it as a checklist; I make sure that I've considered every item on the template, even if I don't include it
[15:17] <bigjools> ok I'll try it again, thanks
[15:17] <bac> i guess the bigger issue is whether reviewers think the merge proposals are providing all of the expected information, whether people use the available tools, or not.
[15:18] <flacoste> abentley: does lp-propose submit the template on the web UI?
[15:18] <abentley> bac, also, if we're going to include test execution times, it would be sensible to add them to the template.
[15:18]  * flacoste doesn't know about lp-propose
[15:18] <bigjools> s/expected/useful and pertinent/
[15:18] <bac> those that i see that use the template tend to cover all of the bases.  doesn't do a thing about the quality of the prose, though.
[15:19] <abentley> flacoste, it opens up your editor to edit the description, then loads the proposal in the browser when it's done.
[15:19] <flacoste> bac: dev writes in code :-p
[15:19] <flacoste> abentley: then I should be using that!
[15:19] <jcsackett> abentley: that sounds like the coolest thing ever.
[15:19] <flacoste> i think it might just be that people don't know about lp-propose
[15:19] <danilos> I find it's hard to find out about it
[15:19]  * jcsackett never heard about it.
[15:19] <deryck> bac: that's part of my issue with the template, sometime those who use it, just list a bunch of info, rather than writing a couple paragraphs explaining what is happening in the code, which is often more useful to me.
[15:19] <danilos> I used to have one of previous submit plugins with the template, but now I type most of the relevant sections out of my head
[15:20] <bac> abentley: i think part of the problem is new people don't know how to use your plug-in and some experienced folks forgot.  could you send out a reminder email or a pointer to the wiki?
[15:20] <jelmer> I was vaguely aware of it, but didn't know it was the proper way to propose merges instead of the web UI.
[15:20] <danilos> it's simply hard to find what the latest and best way to submit MPs is (i.e. appropriate plugin and such: I knew nothing about lp-propose either)
[15:20] <flacoste> jelmer: it sounds like it's a wrapper around the uI, which is exactly what we need
[15:20] <abentley> jelmer, I meant that it's preferred over "bzr send", not necessarily the web UI.
[15:21] <mars> danilos, yes, that is odd - we used to do that just fine (years ago, when I joined)
[15:21] <abentley> danilos, it ships as part of bzr :-P
[15:21] <jelmer> abentley: don't you still need lpreview_body to actually get the template though?
[15:21] <abentley> jelmer, yes, you do.
[15:22] <danilos> abentley, heh, right, that's probably why it's harder for people to find out about it: if you are not actively looking for it and you've been using something like lpreview_body or whatever in the past, you wonder why it doesn't work as well anymore
[15:22] <abentley> jelmer, since it's packaged, we could add it to lp-developer-dependencies, if it's not already.
[15:22] <danilos> (and I was actually stuck on whatever was before lpreview_body with my "lpsend" as the alias)
[15:23] <jelmer> abentley: I think that's a good idea
[15:23] <bac> abentley: can you send that reminder email and pursue getting it added to lp-d-d?
[15:23] <abentley> danilos, I bear some blame, since I wrote it and didn't promote it.
[15:23] <abentley> bac, Sure.
[15:23] <bac> abentley: thanks for bringing up the topic...and for writing the tool.
[15:23] <bac> moving on
[15:23] <bac> [topic]  Integrating test timing into reviews.  --bac
[15:23] <MootBot> New Topic:   Integrating test timing into reviews.  --bac
[15:24] <bac> last week we started the discussion about paying attention to test timing.  we've had a lot of discussioin on the mailing list about what that means.
[15:24] <bac> has anyone tried and have successes or failures to report?
[15:25] <mars> Aaron added two tests yesterday, timing was 2 seconds
[15:25] <bigjools> timing info on its own means nothing to me
[15:25] <mars> TBH, I didn't know how far to pursue it - how much was setup, how much existing tests, what it ok?
[15:25] <mars> 'what is ok'
[15:26] <flacoste> bac: given that we are planning on rewriting the persistence layer and that there is controversy on the metrics side
[15:26] <flacoste> why don't we move on to another aspect?
[15:26] <flacoste> and revisit this later, once the story around persistence and tests is more clear
[15:26] <bac> francis, sure we can do that
[15:26] <flacoste> well, that's not an edict!
[15:26] <flacoste> just proposing
[15:26] <bac> i unwisely thought this would be an easy one to start with
[15:26] <bigjools> I concur :)
[15:27] <bac> i'll propose something next week
[15:27] <bac> [topic] peanuts
[15:27] <MootBot> New Topic:  peanuts
[15:27] <bigjools> o/
[15:28] <bac> yes bigjools?
[15:28] <abentley> Who's seen A Charlie Brown Christmas this year? :-)
[15:28] <bac> are you really left handed?
[15:28] <bigjools> ...
[15:28] <danilos> bac, that was him with his back turned on us
[15:28] <gary_poster> :-)
[15:28]  * bigjools is speechless for the first time in ages
[15:28] <bigjools> anyway
[15:28] <danilos> but he's also a slow typist (especially with only one hand)
[15:29]  * gary_poster laughs
[15:29]  * bigjools sees the gutter approaching
[15:29] <mars> lol
[15:29] <bigjools> I want to talk about the mailing list thread that jelmer brought up
[15:29] <bigjools> regarding api only functions in model classes
[15:29] <jelmer> bigjools: thanks, I forgot about that
[15:30] <bigjools> I think it's a good idea to prepend api_ in front of any method that's only used in the api
[15:30]  * danilos is still behind on his mail
[15:30] <bigjools> anyone got any comments?
[15:30] <abentley> bigjools, I think it's a good idea.
[15:30] <bigjools> (with liberal use of export_as of course)
[15:30] <danilos> other than that we should have made API a separate layer in the first place? no :)
[15:31] <bigjools> I assume that's coming
[15:31] <danilos> anyway, "api_" as the prefix for API-only methods is probably good
[15:31] <bigjools> this is a stopgap
[15:31] <abentley> bigjools, it makes me sad that the API has the zope naming convetion, though.
[15:31] <jelmer> abentley: is that documented somewhere?
[15:31] <bigjools> we should change that *now* if we can
[15:31] <bigjools> but it might be too late
[15:31] <abentley> jelmer, Not as a special thing.  All our code has the zope naming convention.
[15:31] <mars> danilos, makes sense: model -> views (HTML), model -> api-layer (JSON)
[15:31] <jelmer> abentley: I found a couple of methods which explicitly used "export_method_as" and used names with underscores and lowercase characters.
[15:31] <gary_poster> I think it's too late myself
[15:31] <leonardr> bigjools: a while ago we decided it was too late
[15:32] <gary_poster> consistency is more valuable IMO
[15:32] <leonardr> and now it's even later
[15:32] <bigjools> :(
[15:32] <abentley> gary_poster, definitely too late for 1.0 :-(
[15:32] <gary_poster> yup
[15:32] <bigjools> maybe on the next version bump?
[15:32] <bigjools> then we can make everything consistent
[15:32] <bigjools> it's a bit of a mess right now
[15:33] <danilos> E_TOOMUCHWORK
[15:33] <danilos> at least imo
[15:33]  * gary_poster thinks that the webservice will get attention separately
[15:33] <bigjools> I think it's valuable work - we don't have anyone looking at our whole api, other than the people who use it
[15:33] <danilos> of course, we can choose when the next version bump will be
[15:33] <gary_poster> i.e., this is the worng forum
[15:33] <danilos> gary_poster, +1
[15:33] <jcsackett> gary_poster + 1
[15:33] <gary_poster> looking at the whole api: that was to have been what leonardr did soon :-)
[15:33] <bac> gary_poster: yep
[15:33] <bigjools> anyway, votes for api_ ?
[15:33] <abentley> bigjools, +1
[15:34] <danilos> bigjools, +1
[15:34] <gary_poster> sure, _1
[15:34] <bac> +1
[15:34] <gary_poster> heh
[15:34] <deryck> +1
[15:34] <gary_poster> +
[15:34] <jcsackett> +1
[15:34] <danilos> :)
[15:34] <jelmer> +1
[15:34] <deryck> is underbar 1 even less than -1? ;)
[15:34] <bigjools> it's a wunderbar
[15:34] <benji> +1
[15:34] <gary_poster> heh, yeah, maybe so :-)
[15:34] <leonardr> +0
[15:34] <bac> bigjools: looks like you have a winner
[15:34] <bigjools> motion carried
[15:34] <bac> bigjools: will you update the style guide?
[15:35] <bigjools> if I can remember where it is
[15:35] <bigjools> :)
[15:35] <jelmer> Alternatively, I'd be happy to update it
[15:35] <danilos> bigjools, I am guessing dev.launchpad.net/StyleGuide :)
[15:35] <danilos> nope, but it does give useful hints :)
[15:35] <bigjools> what!  it's in an obvious place?  I'd never have thought to look there.
[15:35] <bac> any other topics?
[15:37] <bac> i'll look at the list of people who will be around next wednesday and cancel this meeting if it looks too low.
[15:37] <bac> thanks for coming everyone
[15:37] <flacoste> thanks bac
[15:37] <bac> #endmeeting
[15:37] <MootBot> Meeting finished at 09:37.
[15:37] <danilos> cheers bac
[15:37] <mars> thanks bac
[15:37] <bigjools> cheers
[15:37] <gary_poster> thank you
[15:37] <gmb> Ta
[15:37] <abentley> thanks, bac.
[15:38] <jcsackett> thanks, bac.
[15:39] <benji> leonardr: your "and now it's even later" comment made me think of TMBG's "Older" (http://www.youtube.com/watch?v=2ltJ8kK4G90&feature=related)