[20:19] <fwereade__> niemeyer, ping
[21:37] <thumper> fwereade__: you are working late
[21:39] <fwereade__> thumper, everybody's asleep :)
[21:39] <thumper> fwereade__: luckily it seems it was just my laptop that was screwed, not all raring
[21:39] <thumper> due to an old staging ppa I had enabled
[21:39] <thumper> which someone broke :(
[21:40] <fwereade__> thumper, good and bad news then ;)
[21:40] <thumper> fwereade__: so I now have a half functioning raring install, and I'm getting my quantal vm up to a state where I can run the go tests
[21:40] <thumper> fwereade__: I am a little confused though...
[21:41] <thumper> fwereade__: last week I was sure I merged one of my branches
[21:41] <thumper> but it doesn't seem to be the case
[21:41] <thumper> perhaps the lbox submit failed and I didn't notice
[21:41] <fwereade__> thumper, huh, strange
[21:42] <thumper> I'm just going to run all the tests again, make sure it has a recent trunk merged, and try again
[21:42] <thumper> the juju-core tests seem to be running fine inside the virtualbox quantal image I have
[21:42] <thumper> albiet slowly
[21:44] <fwereade__> thumper, I definitely can shave a few seconds off the uniter tests, at least, but I still haven't got around to it :(
[21:45]  * thumper shrugs
[21:45] <thumper> it is what it is
[22:01] <thumper> fwereade__: still around?
[22:02] <thumper> fwereade__: https://codereview.appspot.com/7370043/diff/3007/cmd/juju/constraints.go#newcode21
[22:02] <fwereade__> thumper, heyhey
[22:02] <thumper> I think we need [<service> [<constraints> ...]]
[22:02] <thumper> but does that make sense?
[22:02] <thumper> the constraints are for the service
[22:03] <fwereade__> thumper, I still think that environment constraints make sense, and I think the tool handles them..?
[22:04] <thumper> [<service>] [<constraints> ...]
[22:04] <thumper> is what you proposed
[22:04] <thumper> I'm wondering which is more normal?
[22:04] <fwereade__> thumper, wait, *get* constraints is just [<service>]
[22:04] <thumper> fwereade__: for ones where we don't set a service?
[22:04] <thumper> fwereade__: well... I'm pretty sure the command treats the first param as a service name only
[22:04]  * thumper double checks
[22:05]  * fwereade__ looks in bafflement at that code, and is sure it made sense when he wrote it
[22:05] <thumper> yes, it seems that get is just <service>
[22:05] <thumper> I should change that
[22:06] <fwereade__> thumper, incidentally I vaguely recall being a bit wishy-washy in the review -- but I'm very much +1 on init
[22:06] <thumper> fwereade__: yes, get takes zero or one params
[22:06] <thumper> fwereade__: but we should change the params for set to be... different
[22:07] <thumper> ah set requires --service
[22:07] <thumper> fwereade__: which makes me think that we have a mis-fit here
[22:07] <thumper> juju get wordpress
[22:08] <thumper> but juju set --service=wordpress
[22:08] <thumper> add -constraints to that if you like
[22:08] <fwereade__> thumper, yeah, I think you're right
[22:08] <thumper> but.. not for this branch
[22:08] <thumper> I'll fix the help text
[22:08] <thumper> but we should fix at some stage
[22:09] <fwereade__> thumper, the only source of inertia is a vague desire to avoid churn against the python version when it's ...merely icky, as opposed to wrong, if you like
[22:10] <thumper> fwereade__: oh, is this a direct copy of the python version?
[22:10] <thumper> if so, lets just leave it for now
[22:10] <thumper> and perhaps file a bug with a nice tag, like "cli-cleanup" :-)
[22:11]  * fwereade__ knows another bug that could use a tag like that :/
[22:12] <fwereade__> (the hook tools all accept --format, even if they aren't meant to produce output)
[22:12] <thumper> fwereade__: so, I can at least run all the tests in an quantal environment to confirm prior to committing
[22:12] <fwereade__> thumper, great
[22:20] <thumper> fwereade__: with the thought of aiming for constency
[22:20] <thumper> fwereade__: we have the args for "set-constratins" to be [key=[value] ....]
[22:21] <fwereade__> thumper, yep, sgtm
[22:21] <thumper> fwereade__: shoudl we change the args for "set" to be "<service> key=[value] ..." ?
[22:21] <fwereade__> thumper, ah: sorry... yes, I think that makes sense
[22:22] <thumper> kk
[22:32] <thumper> gah
[22:32] <thumper> typed lbox propose instead of submit
[22:32] <thumper> fwereade__: I'm going to try to change the account that lbox is using... to the work one, so email gets through
[22:33] <thumper> otherwise launchpad drops the email on the floow
[22:33] <thumper> floor
[22:36] <fwereade__> thumper, heh, unhelpful :(
[22:37] <thumper> fwereade__: well, launchpad isn't going to trust just anyone :)
[22:37] <fwereade__> thumper, quite so ;)
[22:37] <thumper> how long does the submit normally take?
[22:39] <thumper> fwereade__: it seems to be taking ages...
[22:40] <fwereade__> thumper, I think it still does all the propose work of generating the diff for the final rietveld push
[22:41] <thumper> hmm...
[22:41] <thumper> it has been going for minutes and minutes
[22:41] <thumper> with no feedback after the identification of the rietveld link
[22:41] <thumper> oh just done
[22:41] <thumper> geez
[22:42]  * thumper double checks launchpad to check
[22:42] <thumper> yep, in trunk now...
[22:42] <thumper> ok, back to the next pipe