[10:13] <teknico> uh oh, uistage is broken
[10:14] <teknico> I just landed a branch, but it's only doc changes, it shouldn't have broken it
[10:48] <frankban> teknico: working on it, it's not your branch 
[10:48] <teknico> frankban, great, thanks
[11:30] <teknico> frankban, it worked :-)
[11:31] <frankban> teknico: yes indeed, it was just a missing reference in modules.js
[11:32] <teknico> yep
[13:00] <gary_poster> thank you for fixing uistage et al, frankban 
[13:04] <frankban> gary_poster: welcome
[13:13] <frankban> gary_poster: is "needs retrospective" the right tag for when you want to discuss some details of that card during our Friday's call?
[13:13] <frankban> I mean: "class of service" -> "needs retrospective"
[13:13] <gary_poster> goodspud, hi.  I have a local lunch meeting at the time of the unit and service view meeting.  mm, benji or bac, you have experience with those pages.  Are you willing and able to attend a meeting at 1630-1700UTC and then report back to the group with minutes as I describe in "Engineer responsibilities" at the end of https://docs.google.com/a/canonical.com/document/d/1GevzPVClAJaGX7YHG6SmdX2geGvR7uXc4xnuc7VY3t4/edit
[13:13] <gary_poster> # ?
[13:14] <gary_poster> Specifically:
[13:14] <gary_poster> Engineer keeps minutes of the questions received, the input given, the decisions made, and the outstanding engineer-related questions and action items.
[13:14] <gary_poster> Engineer sends minutes to Juju GUI team and design team.
[13:14] <gary_poster> Engineer is responsible for getting the answers to each outstanding question from the meeting, or explicitly getting someone else to take over for that question.  If someone else takes responsibility, this hand-off is sent to the Juju GUI team and the design team.  The same goes for action items.
[13:14] <gary_poster> If an engineer is still on the hook for an answer or action item when a new meeting is called, they should consider themselves volunteered for the next meeting.
[13:14] <gary_poster> frankban, yes
[13:14] <frankban> gary_poster: cool thanks
[13:15] <bac> gary_poster: yes
[13:15] <gary_poster> welcome
[13:15] <gary_poster> thanks bac, I'll add you to the invitation
[13:15] <bac> i like the coercion of that last point.
[13:15] <bac> way to light a fire
[13:15] <benji> heh
[13:16] <goodspud> I'll bring snacks along to the meeting
[13:16] <goodspud> :)
[13:16] <gary_poster> :-)
[13:23] <gary_poster> benji, out of pur nosy curiosity, any progress on getting the production tests to pass?  I'm very curious as to what the resolution will be
[13:24] <benji> yep, I am down from approximately eight billion failures to 2
[13:24] <gary_poster> benji, lol, excellent
[13:24] <benji> they both appear to be of different classes of failures, so they will probably require different fixes, but I feel pretty good about it
[13:26] <gary_poster> cool, benji, thanks for overview.  once you get it working I'm thing that the default test target ought to test the production story, so we all are testing the end result by default.  we'd keep the previous test target as a way to debug.  wdyt?
[13:27] <gary_poster> teknico, approved Friday, no worries
[13:27] <benji> that's reasonable; if the tests were not so slow I would suggest running them both (or all three?), but that is not the case
[13:27] <gary_poster> right
[13:27] <gary_poster> I know!  we could parallelize them with lxc!
[13:27] <gary_poster> wait, no...
[13:29] <frankban> lol
[13:38] <gary_poster> bcsaller, could you update kanban board to reflect your current progress please?
[13:38] <bcsaller> gary_poster: done
[13:39] <gary_poster> thanks bcsaller 
[13:41] <hazmat> bcsaller, what's next after ultra-mega?
[13:42] <bcsaller> hazmat: panzoom is in progress, put it on the board
[13:42] <bcsaller> one of the smaller ones to sort out remaining issues
[13:42] <bcsaller> part way through it after yesterday
[14:14] <teknico> gary_poster, thanks
[14:14] <gary_poster> welcome
[14:18] <teknico> in the code we have README and HACKING files, while in the charm we have README.txt , does anyone care about the extension?
[14:23] <gary_poster> I usually add it, but don't care a lot.  teknico, you available for a quick call on another topic (Makefile fun)?
[14:24] <teknico> gary_poster, in five minutes?
[14:24] <gary_poster> teknico, I have another meeting in 5 minutes :-) can do it in 35
[14:24] <teknico> gary_poster, ok, I'll hop on right away then
[14:24] <gary_poster> ok thanks
[14:39] <teknico> bac, what happens if the kanban active lanes are full and lp2kanban needs to put a card there?
[14:40] <bac> teknico: i'd be really surprised if that is handled gracefully.
[14:40] <teknico> bac, I guess we'll find out soon enough ;-)
[14:40] <bac> teknico: yes, should be interesting.  put on your eye protection!
[14:42] <teknico> bac, and earplugs too, for good measure!
[15:17] <teknico> uhm, nothing happened
[15:29] <gary_poster> bac bcsaller benji frankban goodspud hazmat jovan2 Makyo teknico call in 1
[15:31] <gary_poster> bac goodspud starting
[15:31] <bac> coming right there
[16:28] <goodspud> hazmat, bac - use the team hangout for the meeting
[16:30] <bac> goodspud: ok
[17:46] <frankban> could anyone please review https://codereview.appspot.com/6929057 ? no rush, I am near my EOD. Anyway, charm tests work again.
[17:53] <Makyo> frankban, taking a look.
[17:54] <frankban> thanks Makyo 
[19:30] <hazmat> gary_poster, login stuff fwiw in lp:~hazmat/juju/rapi-login
[19:31]  * hazmat moves on to annotation support
[19:31] <gary_poster> hazmat awesome thanks.  we should switch to that from rollup?
[19:31] <hazmat> gary_poster, no.. i need to do some testing, i'll merge it into rollup after i get annotation support going
[19:31] <gary_poster> cool
[19:31] <hazmat> gary_poster, there are a couple of unit tests, but i'd like to do an end to end test with the gui & charm
[20:17] <gary_poster> sounds good
[21:28] <gary_poster> bac, benji, bcsaller, I'm reviewing all three cards up for review (only one left).  teknico's and Makyo's still need a second review, if you are willing and able.
[21:28] <benji> k
[21:29] <bcsaller> I can take one as well, tell me which you want benji and I'll take the other 
[21:29] <bac> gary_poster: ok.  i will in a bit.  i just put my shoes on so the dogwalk cannot wait.
[21:29] <gary_poster> benji bcsaller bac I suggest you coordinate using the tags on the kanban cards.  put your nick on the one you plan to do or are doing
[21:30] <benji> I don't especially want one (or any), bcsaller; have at it
[21:30] <gary_poster> thanks all
[21:30] <gary_poster> :-)
[21:47] <gary_poster> Makyo, I take it the target-specific variables didn't have the characteristics we were hoping for?  Maybe they were evaluated anyway?
[21:48] <gary_poster> (lloking at https://codereview.appspot.com/6873071/diff/1/Makefile)
[21:48] <gary_poster> looking
[21:58] <Makyo> gary_poster, Correct.  Additionally, the variable containing the target name contained the current target (which changed as make worked through the dependency tree)
[21:58] <gary_poster> Makyo, yeah I saw that second one.  Too bad target specific variables are not lazy, but I guess that makes sense
[21:59] <gary_poster> in the context of other make decisions, I mean
[22:00] <Makyo> gary_poster, yeah.  Additionally the env var solution fits with the other release patterns, i.e.: PROD=1 make tarball
[22:00] <gary_poster> yeah, I thought of that.  I'm mostly just going to be asking for comments explaining why we have to do this stuff.
[22:01] <Makyo> gary_poster, okay.
[22:01] <Makyo> gary_poster, additionally, found out that bzr ls without the -V option lists all files in the directory (but still contacts the parent branch to do so), so the later instances of that translated neatly to find commands.
[22:02] <gary_poster> Makyo, cool, though we need to exclude editor effluvia from them at the least
[22:02] <Makyo> gary_poster, okay.  Wasn't before, but that's a good addition.
[22:02] <gary_poster> Makyo, it did, because we were using bzr ls
[22:03] <gary_poster> and we don't commit editor effluvia
[22:03] <Makyo> gary_poster, bzr ls without the -V option would include those files.  -V wasn't used in those latter two instances.
[22:03] <gary_poster> ah!
[22:03] <gary_poster> suck
[22:03] <Makyo> But that is a good addition for omitting.
[22:03] <gary_poster> well, then yes, please make it better :-)
[22:03] <Makyo> \o/
[22:03] <gary_poster> :-)
[22:09] <gary_poster> benji, for the charm, I'd like to be able to make a release file without uploading it.  you removed the phony target to do so.  Am I correct in assuming that you don't object to adding it back as long as we have a use case?
[22:12] <benji> gary_poster: ah, I hadn't reallized that was a use case; sure add it back
[22:13] <gary_poster> cool thanks
[22:13] <benji> I would love for us to have tests of the Makefile; I claim that we need a card to that end
[22:15] <gary_poster> benji, make a slack card to show us how we coudl do it please
[22:16] <gary_poster> we could do that in the charm too fwiw
[22:16] <gary_poster> though granted it would be nicer and more appropriate in the code itself
[22:20] <gary_poster> Makyo, reviewed: "Land with changes" :-)
[22:20] <gary_poster> running to get son from acting class...
[22:20] <Makyo> gary_poster, thanks :)
[22:25] <bac> gary_poster: i'm back but looks like all the reviews are taken
[22:48] <Makyo> Going to lay down for a bit.