=== cprov is now known as cprov-ZzZ
=== cprov-ZzZ is now known as cprov
=== mrevell is now known as mrevell-lunch
=== mrevell is now known as mrevell-lunch
=== mrevell-lunch is now known as mrevell
bacme 215:01
* sinzui says me15:01
barrywelcome to the ameu launchpad reviewers meeting, for the next 45 minutes or less15:01
barry== Agenda ==15:01
barry * Roll call15:01
barry * Next meeting15:01
barry * Action items15:01
barry * Queue status15:01
barry  * Leftover branches - confirm that the on-call reviewer should allocate all remaining branches on the general queue at the end of his shift (or discuss a different solution).15:01
barry * Mentoring update15:01
barry * Review process changes15:01
barry   * Tool update15:01
* barry goes a reviewer-rustlin'15:02
barrygreat, welcome to our first meeting of 2008!15:04
barry * Next meeting15:04
barrytoday + 1week?  any objections?  any known absences?15:04
barry * Action items15:05
barryActions from the last meeting:15:05
barry * barry to edit him some wikis about on-call procedures15:05
barry * intellectronica to work on a cover letter template15:05
* barry sucks; he will edit him some wikis15:05
* intellectronica sucks too15:05
barryi really will try to do that this week15:05
barry* Queue status15:06
barry  * Leftover branches - confirm that the on-call reviewer should allocate all remaining branches on the general queue at the end of his shift (or discuss a different solution).15:06
barry * Mentoring update15:06
intellectronicai'll try to do this promptly and talk to mwh about inclusion in his tool15:06
barry(oops c&p error there)15:06
barryintellectronica: excellent, thanks15:06
barryso the general queue got pretty big15:06
sinzuiintellectronica: mwh is out for the next 3.5 weeks15:07
barryand intellectronica reminded me this morning that we'd decided to have on-call reviewers do a general queue assignment at the end of his shift15:07
intellectronicasinzui: well, i won't talk to him, then15:07
barryintellectronica: we can keep this on the agenda until mwh gets back15:08
barryi don't think it's urgent15:08
barryfwiw, i gave lifeless the opportunity to shed load so he won't be doing PR assignments for now15:09
barryand i know we all want to just kill off PR anyway15:09
barrybut for now, intellectronica can you do a general queue assignment when you're done for the day?15:09
intellectronicawill do. if ppl /don't/ want branches for review tonight, drop by in #launchpad-reviews and say yo15:10
salgadoI'd stopped doing the allocations because then on call reviewers could take the unassigned ones when there wasn't any branches for them to review15:10
intellectronicasalgado: but today, for example, by the time i finish the queue will, in all likelihood, still have quite a few branches in it15:11
barrysalgado: i wonder if falling behind was mostly an aberration around the holidays/new years?15:11
salgadoyeah, I was wondering that as well15:12
sinzuibarry: That and the fact that reviewers are still out15:12
intellectronicabarry: from what i can tell the queue filled up significantly after your shift yesterday15:12
intellectronicabarry: so it's something we can expect to happen every now and then15:12
barryintellectronica: it was pretty full when i checked toward the latter part of my shift, but by then i was already booked on reviews15:12
intellectronicaeither way - allocations are not going anywhere, even with on call reviews :-/15:13
barrywe need to clear the backlog, so let's do this: intellectronica do an assignment at the end of today and let's all do our reviews the old way15:13
barrythat should clear the backlog and let us get back to our super-efficent on-call process :)15:14
salgadosounds good to me15:15
barryso maybe we should do general assignments on an as-needed basis only, and not as s.o.p. for on-callers?15:15
flacostei have some problems with that strategy15:16
flacosteproblem being that branches on the general queue still should be serviced in the <48H SLA15:17
flacosteso they sould be assigned ASAP for that15:17
flacosteif the on-call cannot do them15:17
flacosteotherwise, there is a risk that they'll be unassigned for a complete shift15:17
flacostebecause the on-call is prompted for new branches15:17
flacosteand then branches on general reviews will linger15:18
flacosteso i think the process should be15:18
flacosteon-call reviewer does allocation15:18
flacostemany times during the day15:19
flacosteif he has slack he assigns some to himself15:19
flacosteput assign others15:19
flacosteno keeping in general queue for the case - where the on-call "might" have time15:19
flacostereviewers should still check PendingReviews daily to see if they have assigned branches15:19
barryok, that's fine by me15:20
flacostei would suggest that after each review, the on-call does an allocation run15:20
intellectronicaflacoste: you've got a point there15:20
flacosteand it's ok for new branches to pre-empt stuff in the general queue15:20
flacosteiow, even if there are unallocated reviews in the general queue, the on-call should still accept an interactive request for a review15:21
flacostethe 'general queue' stuff will be assigned to other reviews15:22
flacosteone gotcha though15:22
barrysounds good.  the only risk is that the on-caller will not have anything to do while there are still needs-review branches in other people's queue.  i'm okay with that15:22
flacosteallocation of reviews to somebody who is fed up with reviews because he was on-called the previous day15:22
flacosteor something like that15:22
barryflacoste: that reviewer can always move the branch to the rejected queue15:23
flacostegood point!15:23
flacosteand we should emphasis that!15:23
flacostedon't be afraid of rejecting reviews because you had a full on-call day!15:23
flacosteof course, allocating to the next on-call is probably a good idea15:24
bacor on-call reviewers can mark their queue with /!\ if they really don't want to accept other reviews.  20% is a lot for of time for reviewing15:24
barryhey, there ya go.  we should just wipe all the review queues and round-robin all needs-reviews to the next on-caller :)15:24
flacostebarry: i fear for the SLA with that strategy15:25
barryflacoste: i was kidding15:25
barrywell, this is never going to be perfect until we get rid of PR anyway <wink> so let's go with flacoste's plan and see how it goes15:26
* barry will edit him some wikis15:26
barryanything else on this topic?15:27
barry * Mentoring update15:27
gmbI've had no reviews for salgado or mwhudson to mentor since last cycle.15:28
barrygmb: cool.  you're next on-call is coming up right?15:28
gmbA combination of sickness and super-efficient on-call revfiews, I think.15:28
gmbbarry: I think so. salgado, are you o/c tomorrow?15:29
salgadogmb, yep15:29
gmbsalgado: Mind if I join you?15:29
salgadogmb, not at all!15:29
flacostehey, another worthwhile allocation strategy: gives them to mentoree!15:29
gmbbarry: There's your answer then :)15:29
barrygmb: excellent!15:29
barryflacoste: you're the perfect straight man.  which leads me to my next question...15:30
barryis it time for more recruits?15:30
barryright now we have just gmb, ddaa, and jvt15:30
flacostewith mwhudson off, i think it's a good idea15:30
barryi think jvt will never graduate (by his own choice iiuc)15:31
flacostesome people just like staying in school all their life!15:31
barryany nominations?15:31
flacostebarry: i think you had volunteers?15:31
barryflacoste: i did15:32
barryjsk was one :/15:32
barryschwuk was another15:32
flacostewe should15:33
flacostesend an request for candidates to launchpad15:33
barrythere aren't too many others left are there? :)  allenap, edwin, bigjools (actually i think he volunteered too), leondardr15:33
bigjoolsI did indeed15:33
barryhow many can we reasonably digest now?  we need mentors for each15:34
cprovintellectronica: here15:34
sinzuiallenap attended our reviewer's meeting at all hands. I thought he was interested15:34
intellectronicacprov: i was just suggesting that you might want to become a reviewer15:35
flacostei suggest a "call for candidacy"15:35
cprovintellectronica: wait, you are nominating me to become a reviewer ...15:35
flacosteand collect candidates and try to find mentors next week15:35
barryflacoste: i say a volunteer candidate has to rustle himself up a mentor.  :)15:35
flacostethat's also a good strategy15:35
barryand we'll accept whoever 1) volunteers; 2) has an agreed mentor15:36
flacosteand we can talk about people who have 1) but not 2) in next meeting15:36
* barry feels like we're getting very close to "everyone is a reviewer"15:36
barryflacoste: +115:36
barryoh yeah, abel too15:37
barry(abel is not a reviewer yet)15:37
barrygreat.  anthing else on this topic?15:37
flacostecarlos and danilos are two others15:37
barryflacoste: right!15:37
cprovintellectronica: maybe after 1.2.5, I feel busy enough until Soyuz 2.0 features get established, but thanks for considering it15:38
flacosteas is Edwin15:38
danilosI should start the procedure soon enough15:38
danilos(as discussed with kiko last year :)15:38
barrydanilos: it's never to early to find a mentor :)15:38
barry* Review process changes15:38
barry   * Tool update15:38
danilosbarry: will look into finding one, thanks for the heads up15:38
barrydanilos: i will send a "call for candidates" later today15:39
danilosbarry: great, thanks15:39
barrygmb: anything to report on the tool?  i had some problems with the plugin yesterday, but i think it was because my rf was messed up.  i re-branched and it worked great15:39
gmbbarry: No, nothing new. With sickness etc I haven't touched on it much, though allenap has volunteered to help with it when he gets time.15:40
sinzuiI always use the -d option to pass the diff I made. I sometimes annotate the diff15:40
barrygmb, allenap_ great!15:41
barryokay, that's it from me.  we have 3 minutes left.  anybody have anything not on the agenda?15:42
=== allenap_ is now known as allenap
barrysounds like we're done then15:43
barryMEETING ENDS15:43
barrythanks everyone!15:43
bacthanks barry15:43
mwhudsonnotes from off: there's a branch on the bazaar mailing list that will make the diff computation review-submit does much faster15:43
barrymwhudson: cool15:44
=== cprov is now known as cprov-out

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