[00:10] <pmatulis> how do i remove a resource from an application (opposite of 'attach-resource')?
[00:13] <anastasiamac> pmatulis: there is no such thing
[00:14] <anastasiamac> pmatulis: why do u need it?
[00:14] <anastasiamac> pmatulis: generally, u'd use resource, u would not "un-need' it...
[00:22] <babbageclunk> thanks hpidcock
[00:42] <thumper> wallyworld: sorry, back on it now
[00:59] <thumper> wallyworld: it is all good and merging
[01:24] <thumper> wallyworld: merged
[02:18] <timClicks_> has anyone got a recommendation for a local DNS server that I can "snap install <dns>"? I would like to add an "external hostname" to things using a .local tld that resolve to my microk8s cluster
[02:18] <timClicks_> (without using xip.io. ideally everything would be locally hosted)
[02:26] <pmatulis> anastasiamac, what if things change and it's no longer required?
[02:27] <pmatulis> kind of like any of the other "undoing" commands
[03:09] <anastasiamac> pmatulis: we have not encountered "resource is no longer required" case. also 'attach-resource' command does attach but also update... this satisfies  cases we have heard of...
[03:10] <anastasiamac> pmatulis: if u have scenario where it is required, plz file a bug and we'll prioritize :)
[03:10] <pmatulis> anastasiamac, i imagine a user can also just make a mistake right? like attach a resource to the wrong application?
[03:11] <anastasiamac> pmatulis: u can only attach a resource if ur charm declares it..
[03:11] <anastasiamac> pmatulis: and u can always update
[03:12] <pmatulis> well, i do have a use case for wanting to remove resources. i will open a bug later. thanks for the context anastasiamac
[03:13] <anastasiamac> nws
[03:52] <lucidone> Is it possible to change a controllers metadata-source after bootstrapping? The docs suggest not. Is it the case that bootstrapping with `--metadata-source ~/simplestreams/images` means that I'm locked into that set of images unless I destroy and recreate the controller? What are my options here with only user permissions in an openstack deployment?
[06:35] <hpidcock> pr please https://github.com/juju/bundlechanges/pull/59
[09:29] <nammn_de1> morning  manadart, stickupkid and i were talking about possible implementations of multiple featureflags:  https://github.com/juju/utils/pull/305 do you have a preference? Thankso
[09:29] <stickupkid> nammn_de1, mine is number 2, let's see what manadart thinks
[09:29] <nammn_de1> stickupkid manadart: another small one, unrelated to above. https://github.com/juju/cmd/pull/70 some feedback for it?
[09:37] <manadart> stickupkid nammn_de1: I missed why we have 2 flags for the same feature...
[09:37] <stickupkid> manadart, rick wants users to be able to target them both in 2.7
[09:37] <stickupkid> manadart, i.e. generations and branches are the same thing
[09:38] <nammn_de1> manadart: he said he would like to ease user experience from someone going from 2.7 to 2.8
[09:38] <nammn_de1> e.g
[09:41] <manadart> stickupkid nammn_de1: I am probably with stickupkid on the alias thing. Mostly because:
[09:42] <manadart> 1) Checking whether to make some logic nascent based on these *multiple* flags is a poor question to pose. A feature:flag should be 1:1.
[09:43] <manadart> 2) It is ambiguous anyway. Is it "or" or is it "and" for example.
[09:59] <stickupkid> I'd love it if the feature flags was a type, then just make global functions
[10:03] <stickupkid> https://paste.ubuntu.com/p/S8NpRRRtTp/
[10:07] <zeestrat> Anyone have a rough ETA on next libjuju release? Is this side of new years realistic?
[10:09] <stickupkid> zeestrat, hopefully, we're just wrapping up on the last PR
[10:10] <stickupkid> manadart, forward port of 2.7 into develop https://github.com/juju/juju/pull/11007
[10:13] <zeestrat> stickupkid: thanks, looking forward to it! it will make my ci lights finally go green.
[11:00] <manadart> stickupkid: Back. Looking at your patch.
[11:01] <stickupkid> nice
[11:27] <nammn_de1> stickupkid: coming back to yaml/json output: https://github.com/juju/cmd/pull/69
[11:27] <nammn_de1> added some tests and make it to return {}
[11:38] <stickupkid> nammn_de1, gave some feedback, we need to work out what happens with an invalid format input error...
[11:41] <nammn_de1> stickupkid:  afaict that is not possible as we register the command beforehand and this   happens to me:  juju models --format=yafdsfs
[11:41] <nammn_de1> ERROR invalid value "yafdsfs" for option --format: unknown format "yafdsfs". But we still can look more into error handling just to be safe
[11:43] <stickupkid> nammn_de1, let's replicate that error handling.
[11:44] <stickupkid> nammn_de1, I'd like to write a guide of how to deal with errors, so others can follow
[11:44] <nammn_de1> stickupkid: ho?
[11:44] <stickupkid> sure
[11:50] <nammn_de1> stickupkid: ahh ho for 2 min?
[11:51] <stickupkid> sure
[11:59] <nammn_de1> stickupkid : thanks for the feedback. Added it https://github.com/juju/cmd/pull/69
[11:59] <stickupkid> nammn_de1, approved
[12:00] <stickupkid> gr8 don't have green tick, sigh - nammn_de1 might be worth speaking to rick_h or jam about my lack of green tick POWER in juju/cmd
[12:00] <jam> stickupkid: green tick POWER! :)
[12:02] <jam> stickupkid: I added hackers to Write on juju/cmd
[12:07] <jam> stickupkid: does that give you the GTP ?
[12:30] <manadart> stickupkid: Patch for emitting redirects: https://github.com/juju/juju/pull/10989
[13:25] <stickupkid> jam, yaep
[13:25] <stickupkid> *yeap
[16:06] <stickupkid> manadart, sorry it took so long, was busy in other stuff
[16:06] <stickupkid> manadart, approved
[16:06] <manadart> stickupkid: Giddyup. Ta.
[16:08] <stickupkid> manadart, https://github.com/juju/juju/pull/10989#discussion_r356125073 this would be good at some point
[16:08] <stickupkid> help with traceability etc
[16:09] <manadart> stickupkid: Yep. Saw it. Let's talk tomorrow. About to finish up for today.
[16:09] <stickupkid> manadart, peeeeeeeeeeeeeeeeerfect
[16:10] <stickupkid> nammn_de1, CR'd
[17:32] <nammn_de1> stickupkid: thanks! Lol, looks like I overlooked that some commands fail on init as well. Should I add them as well?
[17:33] <stickupkid> erm, probably, propose a fix, let's see
[17:33] <nammn_de1> stickupkid: ohhh man haha
[17:33] <nammn_de1> :D
[17:48] <nammn_de1> stickupkid: damn on cmd/main it is harder as we do not have access to supercommand, looks like i need to parse it myself
[17:50] <nammn_de1> stickupkid: no access to commonflags, but then we are back to the discussion whether we want to add a ParseFlag(name, flags) function
[17:51] <stickupkid> nammn_de1, let's look tomorrow, brain is fried today
[17:51] <nammn_de1> stickupkid: haha no worries, i will open a pr tomrrow and we can talk about then
=== timClicks_ is now known as timClicks
[21:57] <lucidone> Is it possible to have relations between charms in different clouds? Use case is postgres in openstack related to an app in k8s
[21:59] <rick_h> lucidone:  definitely, they're called cross model relations
[21:59] <rick_h> https://jaas.ai/docs/cross-model-relations
[22:00] <lucidone> Magic thanks, these are the docs I was looking for :)
[22:03] <rick_h> lucidone:  and for the video inclined https://www.youtube.com/watch?v=NUx6kYE60Mc
[22:04] <lucidone> Neat! Cheers
[23:46] <hpidcock> wallyworld: https://github.com/juju/bundlechanges/pull/59