[00:01] <wallyworld> babbageclunk: standup?
[00:49] <thumper> wallyworld: https://github.com/juju/juju/pull/8220
[00:49] <thumper> I'm going to make coffee
[00:49] <wallyworld> ok, looking
[00:49] <thumper> wallyworld: I'm asking for a review because it isn't just a backport
[00:49] <thumper> but it is sufficient for what we need
[00:49] <wallyworld> righto
[00:49]  * thumper -> coffee
[00:53] <wallyworld> thumper: it looks like just testing chnges, did the PR include a functional change also?
[00:57] <thumper> nope
[00:57] <thumper> wallyworld: interface one up next
[00:57] <thumper> then the functional change
[00:58] <thumper> I'm trying to reduce the scope of the final review
[00:58] <wallyworld> ok
[00:58] <wallyworld> testing one lgtm
[00:59] <thumper> wallyworld: this one has the testing commit and the interface https://github.com/juju/juju/pull/8221
[01:00] <wallyworld> righto
[01:44] <babbageclunk> wallyworld: https://github.com/juju/juju/pull/8222 plz?
[01:44] <wallyworld> sure
[01:54] <wallyworld> babbageclunk: i made a suggestion
[02:00] <babbageclunk> wallyworld: yeah, thanks - tweaking it now
[02:01] <wallyworld> awesome
[02:07] <wallyworld> thumper: ? 1:1
[02:08] <thumper> yeah, just finishing up with IS call
[03:30] <thumper> wallyworld, jam: review up https://github.com/juju/juju/pull/8223
[03:30] <wallyworld> ok
[03:41] <wallyworld> thumper: has the hub watcher stuff been reviewed elsewhere? i think so?
[04:37] <jam> wallyworld: I reviewed it on a target for 2.3, so I'll give Tim's branch a review here.
[06:33] <wallyworld> jam: i have a small PR to add caas application-config to the applocation entity in the juju/description repo (if you have a chance) https://github.com/juju/description/pull/32
[08:31] <thumper> wallyworld, jam: I'm probably going to miss the team standup tomorrow as my daughter has a school end of year assembly where she is getting a prize
[08:31] <thumper> so I should be there :)
[19:22] <thumper> morning
[19:22] <thumper> balloons: I realised that I probably won't be around for the release call due to a school assembly that I need to be at
[19:22] <thumper> balloons: got time to chat now?
[19:23] <balloons> thumper, ohh, morning
[19:23] <balloons> give me 5
[19:24] <thumper> ack
[19:28] <balloons> ok, hopping into release call thumper
[19:28] <thumper> ok
[20:33] <babbageclunk> wallyworld: ping?
[20:43] <thumper> balloons: managed to reproduce failure...
[20:43] <thumper> and it happens on both 2.2.7 and 2.3.2
[20:43] <thumper> which makes me happier
[20:44] <balloons> oO?
[20:49] <wallyworld> babbageclunk: hey
[20:51] <babbageclunk> wallyworld: sorry, had you been waiting for me to do that description review?
[20:52] <wallyworld> not really, i only just added it at my EOD
[20:52] <babbageclunk> ok cool
[20:52] <wallyworld> i need it for today, so all good, ty
[20:53] <wallyworld> babbageclunk: thanks for the version pickup in setApplications, i missed that
[20:53] <wallyworld> there should be a test for that stuff
[20:53] <babbageclunk> wallyworld: no worries. Can you look at this audit log errors one? https://github.com/juju/juju/pull/8225
[20:53] <wallyworld> to ensure that the model has all the latest versions set
[20:54] <babbageclunk> wallyworld: yeah, true
[20:54] <wallyworld> sure, i'll need 30 mins - just woke up
[20:57] <babbageclunk> no hurry.
[20:57] <thumper> balloons: a subordinate charm with no relations is being confused for a principal
[20:58] <thumper> balloons: so we are asking for constraints
[20:58] <thumper> however juju status doesn't show any difference between a subordinate with no relations, and a normal charm with no units
[20:58] <balloons> thumper, ahh.. I was confused why that made you happy, but I get wanting consistency
[20:59] <thumper> I ended up bootstrapping two controllers
[20:59] <thumper> and doing the exact same thing on both
[20:59] <thumper> and was getting the same results
[20:59] <thumper> finally worked out what it might have been
[20:59] <thumper> and it failed on both
[20:59] <balloons> wallyworld, babbageclunk we still on to talk upgrade testing in 30 mins?
[20:59] <thumper> so, yes, consistence
[20:59] <babbageclunk> balloons: yup yup
[20:59] <balloons> brillant, ty
[20:59] <thumper> ok, I'm about to head off, back after lunch
[21:00] <babbageclunk> gah, thumper - can I quickly get your opinion?
[21:01] <babbageclunk> or wallyworld, since he really is away?
[21:02] <babbageclunk> wallyworld: re exclude list for audit logging it occurred to me that we don't want empty conversations (where the requests and responses are all for uninteresting API methods) clogging up the log.
[21:05] <babbageclunk> so what I'm planning on doing is buffering up the conversation (with any uninteresting API requests/responses) in the recorder until we see an interesting request, at which point the buffered conversation and req/resps will also be written. Does that sound sensible or crazy?
[21:08] <babbageclunk> anyway, I'll talk to you about it at the end of the meeting.
[22:22] <wallyworld> babbageclunk: did you want to  talk now?
[22:22] <babbageclunk> wallyworld: sorry give me 5?
[22:22] <wallyworld> sure
[22:26] <babbageclunk> wallyworld: ok now's good
[22:26] <babbageclunk> in 1:1?
[22:28] <wallyworld> ok
[22:58] <wallyworld> babbageclunk: lgtm with a suggestion
[23:04] <babbageclunk> wallyworld: thanks!
[23:55] <wallyworld> babbageclunk: i'm having trouble joining hte call