/srv/irclogs.ubuntu.com/2010/08/11/#launchpad-meeting.txt

=== mwhudson_ is now known as mwhudson
=== Ursinha is now known as Ursinha-afk
=== mrevell is now known as mrevell-lunch
=== Ursinha-afk is now known as Ursinha
=== mrevell-lunch is now known as mrevell
bac#startmeeting15:01
MootBotMeeting started at 09:01. The chair is bac.15:01
MootBotCommands Available: [TOPIC], [IDEA], [ACTION], [AGREED], [LINK], [VOTE]15:01
bachi who is here?15:01
abentleyme15:01
adeuringme15:01
deryckme15:01
bacsinzui just wandered off15:01
bacdanilos:  ping15:02
danilosme15:02
bacgmb, EdwinGrubbs, leonardr, jelmer ping15:02
leonardrme15:02
bacbigjools:  yo15:02
EdwinGrubbsme15:02
henningeme, too15:03
bigjoolsme ish15:03
bacflacoste, noodles775, mars: ping15:03
noodles775here15:03
marsme15:03
jelmerme15:04
sinzuime15:04
bac[topic] agenda15:04
MootBotNew Topic:  agenda15:04
bac * Roll call15:04
bac * Agenda15:04
bac * Outstanding actions15:04
bac * New topics15:04
gmbme15:04
bac  * Format import lists so they merge well and are clear to review. RobertCollins15:04
bac  * Call for new reviewers.15:04
gmbSorry15:04
bac  * Removing enums and errors out of interfaces.  TimPenhey15:04
bac  * Policy decisions: Reviewers Meeting or mailing list?  RobertCollins15:04
bac * Peanut gallery15:04
bac[topic] outstanding issue15:04
MootBotNew Topic:  outstanding issue15:04
bacthere is just the one, coop vs shared vs inter.  that discussion has been happening on the mailing list so i don't see the need to bring it up here too.  so i won't.15:05
bac[topic] Format import lists so they merge well and are clear to review. RobertCollins15:05
MootBotNew Topic:  Format import lists so they merge well and are clear to review. RobertCollins15:05
baclifeless brought this up on the list but i was quite confused by the discussion15:06
bigjoolsthis is the issue that came up a year (?) ago15:06
bacwe discussed it in the asiapac meeting last week and it turns out, he is advocating for the one-line-per-item in lists (as we do now) plus extending that to imports15:06
bacwhich we discussed and rejected a year ago as bigjools said15:06
bacrobert and the code team are still very much in favor in this change.15:07
abentleybac, true, dat.15:07
* henninge doesn't mind either way15:08
henningeActually, I see both formatings in our code.15:08
henningeformats, formations, formatations ???15:09
henninge;-)15:09
bigjoolsI'm not sure if I care which way it is, but I do care that the files are all the same15:09
gmbbigjools, +115:09
bacbigjools +115:09
henningeOuch, that is a lot of work.15:09
gmbThat's the biggest bugbear.15:09
flacosteme15:09
bigjoolssurely we can write a tool to reformat them?15:10
bacif we do decide to make the change it'll be a big, mechanical branch.15:10
sinzuiwe may have such a tool15:10
bigjoolsJFDI15:10
bacpeople who do big apocalyptic branches report seeing lots of merge conflicts over imports...i personally don't see it often15:10
sinzuithe migrater has a script that rewrite imports. I think we need to update its formatter15:10
bacbigjools:  before we JFDI we have to decide if that's what we really want.15:11
baclet's vote then.  are there any strong arguments for or against first?15:11
bigjoolsif we have a tool, we can try it out and revert it if we hate it15:11
flacostei don't like it15:11
flacostebut only esthetics argument15:11
henningeReadability is surely on the one-liner side.15:12
bigjoolsflacoste: what you gain on the editor you lose in the diff.15:12
henningewith lists being alphabetical!15:12
abentleybac, the arguments pro are fewer conflicts and clearer patches.15:12
flacosteyep15:12
abentleybac, clearer patches because only the thing added/removed will show, not the whole import block for that module.15:12
bacwould a single import stay on one line but two or more become a list?15:12
henningebac: I'd like that.15:13
daniloswe should just fix the very long "one-line" imports as an alternative solution (i.e. be more specific about what you are importing from)15:13
abentleybac, that would be fine by me.15:13
abentleydanilos, I'm not sure what you mean by 'very long "one-line" imports'15:14
flacostei think he means multi-line import15:14
danilosabentley, well, single-statement imports with lots of things being imported15:14
flacostefrom canonical.launchpad.interfaces import A, B, C, D... over 3 lines15:14
flacosteif you import for more specific location most of them will fit on one line15:14
flacostebut it doesn't solve the conflict problem15:15
flacosteif you need to add a 3rd symbol15:15
flacosteor a second one15:15
flacosteor the readability of the patch15:15
danilosright, it does reduce the chances of it happening though (you are going to get conflicts with the proposal to split them up as well)15:15
abentleyflacoste, I haven't noticed a lot of too-general imports lately, but maybe I'm not looking for them.15:15
flacostewe don't add them15:15
bacok, let's vote15:15
sinzuiThere are still a lot from c.l15:15
bac[vote] Change our coding standards to require one line per item for imports15:16
MootBotPlease vote on:  Change our coding standards to require one line per item for imports.15:16
MootBotPublic votes can be registered by saying +1/-1/+0 in the channel, private votes by messaging the channel followed by +1/-1/+0  to MootBot15:16
MootBotE.g. /msg MootBot +1 #launchpad-meeting15:16
danilos-115:16
MootBot-1 received from danilos. 0 for, 1 against. 0 have abstained. Count is now -115:16
abentley+115:16
MootBot+1 received from abentley. 1 for, 1 against. 0 have abstained. Count is now 015:16
bigjools+1 because I fancy a change :)15:16
MootBot+1 received from bigjools. 2 for, 1 against. 0 have abstained. Count is now 115:16
adeuring-115:16
MootBot-1 received from adeuring. 2 for, 2 against. 0 have abstained. Count is now 015:16
gmb+115:16
MootBot+1 received from gmb. 3 for, 2 against. 0 have abstained. Count is now 115:16
henninge+1 for readability15:16
MootBot+1 received from henninge. 4 for, 2 against. 0 have abstained. Count is now 215:16
noodles775+015:16
MootBotAbstention received from noodles775. 4 for, 2 against. 1 have abstained. Count is now 215:16
sinzui+115:16
MootBot+1 received from sinzui. 5 for, 2 against. 1 have abstained. Count is now 315:16
mars+015:16
MootBotAbstention received from mars. 5 for, 2 against. 2 have abstained. Count is now 315:16
bac-015:16
* bigjools spies a fence-sitter15:16
danilos:)15:16
henningebac: -0 doesn't work15:17
flacoste-115:17
MootBot-1 received from flacoste. 5 for, 3 against. 2 have abstained. Count is now 215:17
bac-115:17
MootBot-1 received from bac. 5 for, 4 against. 2 have abstained. Count is now 115:17
jelmer+115:17
MootBot+1 received from jelmer. 6 for, 4 against. 2 have abstained. Count is now 215:17
marsfascinating :)15:17
danilos(this seems like the most interesting vote so far :)15:17
bacvoting to close in ten seconds15:17
deryck+115:17
MootBot+1 received from deryck. 7 for, 4 against. 2 have abstained. Count is now 315:17
EdwinGrubbs-115:17
MootBot-1 received from EdwinGrubbs. 7 for, 5 against. 2 have abstained. Count is now 215:17
gmbDemocracy in action, folks.15:18
bac#endvote15:18
bac[endvote]15:18
MootBotFinal result is 7 for, 5 against. 2 abstained. Total: 215:18
henningeplus the code team and Rob15:18
bacok, so the voting in AsiaPac will be perfunctory.  the motion will carry as they all favor it.15:18
danilosthey haven't had the chance to listen to our wonderfully founded arguments, though :)15:19
bacdanilos:  yeah,i don't think that's going to sway them15:19
bac:)15:19
baci'll seek an asiapac volunteer to update the wikis15:19
danilosbac, perhaps some beating or bribing then? :) anyway, thanks all for the vote15:19
daniloss/vote/voting/15:20
bac[topic] call for new reviewers15:20
MootBotNew Topic:  call for new reviewers15:20
bacAny team leads want to nominate new mentats?15:20
bacbigjools:  ???15:20
* bigjools got distracted15:20
* bigjools is nominating SteveK15:21
bacyou have mentor?15:21
bachey bigjools, how about tim?15:22
bigjoolsyes, thumper agreed to mentor him15:22
baci'll send an email to steve and invite him to the asiapac meeting15:22
bigjoolsthanks bac15:22
bacthanks bigjools for nominating him15:22
bigjoolsmy pleasure :)15:22
bac[topic] Removing enums and errors out of interfaces.  TimPenhey15:23
MootBotNew Topic:  Removing enums and errors out of interfaces.  TimPenhey15:23
bactim sent out email about this topic last week but there was no discussion.  he asked me to bring it up here15:23
bigjools+1 million to that15:23
bac+115:23
flacosteyep, that's a JFDI case15:23
abentley+115:23
gmbA thousand times +115:23
danilos+1 here as well :)15:24
jelmer+1 too. Just to be sure, that means moving them to the model code?15:24
sinzui+115:24
bacso create lp/app/enums.py and errors.py15:24
bacerr, lp/<yourapphere>/enums.py i mean15:24
bac[topic] Policy decisions: Reviewers Meeting or mailing list?  RobertCollins15:25
MootBotNew Topic:  Policy decisions: Reviewers Meeting or mailing list?  RobertCollins15:25
henningemeeting!15:25
baclifeless brought this up last week and wanted to discuss it here15:25
bachis thought is having this meeting make policy as one of its goals drags out decisions15:26
bacML discussions can lead to quicker decision making, he argues15:26
bacthoughts?15:26
flacostemeeting are good for a sync point15:27
flacostemaking policy decision on mailing list is confusing15:27
abentleyHaving a meeting focuses discussion.  It may increase latency, but the actual decision-making is faster, and the results are clearer.15:27
henningeI don't agree. This meeting is prove of that re the decission on import formating.15:27
flacostewell, not confusing15:27
flacostebut requires art15:27
flacosteit's often not clear when the decision is reached15:27
danilosI strongly disagree with his assertion that meetings lead to slower decision making; if anything, they lead to faster decision making imho15:27
baci think we get more opinions in this meeting and it is easier to catch confusion and eliminate them15:28
henningedanilos: +115:28
flacostei think his impression might come from the AsiaPac/Ameu split15:28
henningeAlso, it is the only place left were we regularly meet as developers and discuss developing - sort of.15:28
flacostewhere the bullk of the devs are in Ameu15:28
abentleydanilos, agreed.  ML threads tend to wander, and don't necessarily reach a conclusion.15:28
bigjoolsdecisions in meetings, discussions about decisions on the ML, unless there's an obvious early consensus on the ML.15:28
danilosI15:28
deryckI favor the meeting, too.  Seems quicker to me, and you know when it's over. :-)15:29
danilosI did suggest him to move to Europe, was what I was trying to say :)15:29
bachenninge:  that is true and it makes this meeting sometimes be broader than the original intent.  i'm not saying that is bad, though.15:29
bacbigjools:  +115:29
danilosbigjools, +1, and it seems most others who voice their opinion are in agreement15:29
daniloshowever, flacoste does point a valid thing: how do we involve our dear AsiaPac developers in the decision making better15:30
abentleydanilos, hold their meeting before ours? :-)15:30
bigjoolsif it's a simple vote we just tally votes15:31
danilosbigjools, I seriously believe that the discussions we have are useful, so just tallying the votes is not always a true measure imho15:31
bacwell, i try to bring the items from here to their meeting.  and votes continue there.  really it's the best we can do.15:31
danilos(i.e. I've been swayed before to a different opinion)15:31
flacostelike we did today for a couple of items15:31
danilosbac, right15:31
bacgood feedback.  thanks.15:32
danilosanyway, I think we should keep the format, but look into other opportunities to improve it15:32
bigjoolsdanilos: that's why I said long discussions need to be on the ML15:32
bac[topic] other items?15:32
MootBotNew Topic:  other items?15:32
baci guess not.15:33
danilosbigjools, sometimes they are short discussions ;)15:33
bacthanks for coming15:33
danilosthanks all15:33
bac#endmeeting15:33
MootBotMeeting finished at 09:33.15:33
bigjoolsthanks bac15:33
bigjoolsdanilos: short is fine for meetings, long needs to go to ML.  Simple :)15:33
danilosbac, sinzui: btw, is it worthwhile to do import lines migration per-app? we've got a big branch in progress and would like to avoid having conflicts until that lands15:34
bacdanilos:  that is a good idea15:34
bacdanilos:  we didn't really talk about the D part of JFDI.15:35
bacit is more tech-debt15:35
danilosbac, I can imagine sinzui already hacking away at a script to do it, which is why I am saying that :)15:35
bacdanilos:  he is not.  his fingers are idle at the moment.15:36
danilosbac, hopefully we don't end up with JFI :)15:36
bacthat's not to say he isn't thinking about it...15:36
danilosbac, heh, thanks :)15:36
sinzuidanilos, I have my own agenda of removing imports from c.l and any tool that make it easier for me to do will probably help you.15:36
sinzuidanilos, I am hacking on codehosting at the moment. rosetta will be next week maybe15:37
danilossinzui, rosetta should be completely glob-import free, so I expect it to be a much easier job :) anyway, thanks for working on that15:38
sinzuiyes, code is clean too. answers and blueprints had a lot but, they were small apps15:38
sinzuiI dread the registry-soyuz apps15:38
bigjoolsdanilos: JFI - lol15:39
=== Ursinha is now known as Ursinha-lunch
=== salgado is now known as salgado-lunch
=== Ursinha-lunch is now known as Ursinha-brb
=== salgado-lunch is now known as salgado
=== Ursinha-brb is now known as Ursinha
=== EdwinGrubbs is now known as Edwin-lunch
=== Edwin-lunch is now known as EdwinGrubbs
lifelessoh hai?22:29
thumperhey22:30
thumperbac: are we doing this now?22:30
thumperbac: or do you want to wait for StevenK?22:30
sinzuibac: is trying to talk22:31
bacthumper:  i cannot tonight22:31
bac#startmeeting22:31
MootBotMeeting started at 16:31. The chair is bac.22:31
MootBotCommands Available: [TOPIC], [IDEA], [ACTION], [AGREED], [LINK], [VOTE]22:31
baci'm having connectivity problems too22:31
bacthumper, mwhudson, rockstar, lifeless :ping22:32
rockstarhi22:32
mwhudsonhi22:32
thumperhi22:32
bacmwhudson:  i didn't know if you wanted to show up for these.  you're welcome, of course!22:32
mwhudsonbac: o22:32
mwhudsoni'm still doing launchpad reviews occasionally, so i think it makes sense22:33
bacok, great22:33
lifelessyeah, Ii'm here22:33
lifelessalso in a meeting with emo :P22:33
lifelesselmo22:33
bacso we had a lively meeting this morning.22:34
thumperelmo the emo?22:34
thumperthat is an interesting picture22:34
mwhudson:)22:34
bacwe discussed the one-import-per-line issue22:34
bacand voted22:34
bacso let's vote here and make it official22:35
bac[vote] Change our coding standards to require one line per item for imports22:35
MootBotPlease vote on:  Change our coding standards to require one line per item for imports.22:35
MootBotPublic votes can be registered by saying +1/-1/+0 in the channel, private votes by messaging the channel followed by +1/-1/+0  to MootBot22:35
MootBotE.g. /msg MootBot +1 #launchpad-meeting22:35
mwhudsonbac: one line per item means "like bzr" ?22:35
bacmwhudson:  yep22:35
rockstar+122:35
MootBot+1 received from rockstar. 1 for, 0 against. 0 have abstained. Count is now 122:35
thumperbenji had an interesting email22:35
mwhudson+122:35
MootBot+1 received from mwhudson. 2 for, 0 against. 0 have abstained. Count is now 222:35
thumper+122:35
MootBot+1 received from thumper. 3 for, 0 against. 0 have abstained. Count is now 322:35
baci wonder how lifeless would vote...22:36
bac[endvote]22:36
MootBotFinal result is 3 for, 0 against. 0 abstained. Total: 322:36
rockstarWait, is wgrant around?  Maybe he should have had a say too...22:36
bactotal:  10 for,5 against, 2 abstained22:36
bacrockstar:  you don't need him22:37
rockstarNice.22:37
bacso we decided it makes sense to make the change one app at a time22:37
lifelesshmm +122:37
rockstarIt really wasn't about need, but more that he contributes as well.22:37
bactotal:  11 for, 5 against, 2 abstain22:38
wgrant+1, anyway.22:38
bac12 for22:38
bacnow you're all just rubbing it in22:38
=== salgado is now known as salgado-afk
thumperI'll start with code22:38
rockstarbac, wait, the chad!  It made my vote unreadable!  :)22:38
bacyeah, it's another tech-debt issue.  so update your code as you have time.22:39
rockstarthumper, of course code is going to be the first to make the change...  :)22:39
bacthanks for championing it lifeless and thumper22:39
mwhudsonhooray for a definitively coloured bikeshed22:39
lifelessPink22:39
bac\o/22:39
lifelessbtw22:39
lifelesscoop22:39
bacgah, not yet22:40
lifeless:)(22:40
lifelessthat was :)22:40
bacwe also got the good news that bigjools has asked stevek to become a reviewer and that thumper will be mentoring22:40
* thumper ndos22:40
thumpernods22:41
bacas mentioned earlier it probably makes sense for him to go to the ameu meeting22:41
thumperENOCAFFEINE22:41
bacif not, we may need to move the time here to a few hours later22:41
thumperbac: when is that?22:41
bacer, 14Z22:41
thumperbac: because AM of the AMEU meeting meens he won't be up22:41
thumperthat is a midnight meeting for him22:41
bacyuk22:41
thumperor 1 am22:41
thumpernot really practical22:42
bacok.  let me discuss with him22:42
bacif we moved this meeting two hours later it would be his 930a and my 730p22:42
bacnot a problem, usually, as long as i don't forget22:42
thumperthat might be better...22:42
mwhudsonfine for me, obviously...22:43
bacthumper:  i brought up your idea about moving enums and errors and it was met with universal praise.22:43
thumpercool22:43
thumperI'm happy with that22:43
baclifeless:  we also discussed your idea of moving policy decisions to the mailing list rather than discussion in the reviewers meeting22:44
bacthe feeling was that the meeting allows for decisions to be clarified and decided much more quickly than on mailing lists, where things tend to drag on and get muddled22:45
rockstarYes.22:45
lifelesssure, I did mention it wasn't baked :)22:45
lifelessalso btw, policy decisions is overly broard22:45
lifeless*broad*22:45
rockstarThere's value in IRC meetings, methinks.22:45
baci think we get more participation "in person" than on the ML22:45
lifelessthere are policy decisions about doing revisions22:45
lifelesss/revisions/reviews/22:45
baclifeless:  sorry if i misrepresented your idea22:45
lifelessand there are policy decisions about what the code should look like22:46
lifelessI was only speculating about the second set22:46
lifelessbac: I don't know if you did or didn't; I'm just being clear here-and-now22:46
rockstarlifeless, well, yeah, that stuff contains bike shedding in both places, so it's better on the mailing list where I can ignore it if I want.22:46
lifelessanyhow22:47
lifelesslike I said - unbaked; I will bring it up if I start to feel its baked22:47
lifelessbut please do think about how we can:22:48
bacin reality, discussions on IRC need to be limited to about ten minutes.  so if the topic has not been introduced before, some topics can be decided quickly while others will need to go to the ML22:48
lifeless - bring in more community folk (not just wgrant, there are others :P) into code & design discussions22:48
lifeless - get things that are easy to happen without waiting for a meeting22:48
lifelessbac: asgreed22:49
baclifeless:  thanks for bringing up the idea.  you have fresh eyes and can see where we've become stodgy22:49
* wgrant likes IRC for discussions, since it means you can't keep things on the iinternal list :P22:49
lifelesswgrant: FTR I only use the internal list for actually confidential discussions, or if someone starts one there.22:50
mwhudsonwgrant: we have internal channels too, you know :-)22:50
bacwgrant:  we've tried to be better about that22:50
wgrantmwhudson: Shh.22:50
bacany other topics for today?22:51
lifelessponies22:51
lifelesslamont wants one22:51
bacwhat happened to woody?22:51
thumpersend him a picture of the storm trooper my little pony22:51
lifelesshe moved on, I don't remember the details22:51
bacwe're sprinting this week and the conference room came with a 3' inflatable tux as a mascot22:52
lifeless\o/22:52
lifelesswhere are you?22:52
bacmy town22:52
lifelessnoice22:52
baccarolina del norte22:52
baccarrboro, specifically22:52
bacok, so if we've devolved into pony discussions i think we should call it a meeting22:53
bacthanks for participating22:53
mwhudson+122:53
mwhudsonthanks bac22:53
thumperthanks22:53
baci'll let you know if we have a time change22:53
bac#endmeeting22:53
MootBotMeeting finished at 16:53.22:53
=== Ursinha is now known as Ursinha-bbl

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!