UbuntuIRC / 2020 /05 /19 /#smooth-operator.txt
niansa
Initial commit
4aa5fce
[06:25] <jam> morning all
[06:35] <MarkMaglana> morning jam !
[08:42] * Chipaca pokes the internal irc
[08:42] * Chipaca presents footage of said event: https://www.youtube.com/watch?v=JMJXvsCLu6s
[08:45] <jam> morning Chipaca
[08:48] <Chipaca> jam: how's things?
[08:58] <jam> going alright here.
[08:58] <jam> Chipaca, they just called to tell me my coffee machine should be fixed by tomorrow, whi
[08:58] <jam> which is pretty good news.
[08:58] <jam> Though the repair cost is ~ the same as an aeropress :)
[09:00] <Chipaca> I'm probably going to jinx it, but my aeropress is 6+ years old and still working just fine :)
[09:53] <Chipaca> niemeyer: hiya! two questions about 'ops.lib.use': one is if you have a strong preference for LIB vs LIB_. The other is whether API/PATCH are the right names for what semantic versioning (which we encourage) is calling MAJOR and MINOR
[09:54] <Chipaca> patch in particular seems strange as it's the name of the third element (the z in x.y.z)
[09:54] * Chipaca is not blocked by these questions and goes back to writing tests
[09:57] <niemeyer> Chipaca: major and minor mean nothing to most people.. API and patch versions has a stronger semantic
[09:59] <niemeyer> Chipaca: My preference is obviously for the existing names otherwise I'd not have used them, but it's certainly subjective
[09:59] <niemeyer> (in terms of the underline, that is)
[10:01] <Chipaca> niemeyer: ok
[10:01] <Chipaca> niemeyer: sometimes it feels like it's too mashed together without the _, but sometimes it's fine
[10:01] <Chipaca> it's *very* subjective :)
[10:02] <niemeyer> These names will be always seem next to each other..
[10:02] <niemeyer> Just type them out and look at them for a moment
[10:02] <Chipaca> niemeyer: yeah, i think they feel weird when writing the tests, but not too weird in the init file itself, which is probably the right way around for weird
[10:03] <niemeyer> Looked better without, since the whole prefix is mostly noise in that case.. but I won't argue too much in either direction :)
[10:41] <Chipaca> niemeyer: another q: what about API or PATCH being 0?
[10:41] <Chipaca> that's ok, right?
[10:44] <niemeyer> Chipaca: Yeah, sounds fine
[11:12] <facubatista> Muy buenos días a todos!
[11:27] <jam> morning facubatista
[11:27] <facubatista> hola jam
[13:21] <Chipaca> jam: we should add a test for dispatch + independent action
[13:23] <jam> Chipaca, indeed. Though it should follow the same logic as the hooks/ independent script (AFAICT)
[13:25] <vgrevtsev> hi team, a quick q: do we have a way to alter the "Ports" in the `juju status` output after the charm is deployed?
[13:28] <jam> vgrevtsev, that is generally populated from calling "open-port", IIRC
[13:30] <vgrevtsev> I can't find any reference to the `open-port` in the framework code, so I'm wondering how it's getting done and how can I invoke it within the charm code
[13:30] <vgrevtsev> I doubt that the `open-port` is available using the k8s charms.. at least I tried to run it inside the operator pod and got nothing
[13:46] <Chipaca> vgrevtsev: https://github.com/canonical/operator/issues/179
[13:47] <Chipaca> so no surprise you don't find it yet :)
[13:47] <Chipaca> i don't know about the k8s side of your question though
[13:48] <vgrevtsev> Ok, so at least something! ;D
[13:50] * facubatista brb
[18:29] <facubatista> jam, which is the minimum set we need? (as the OF itself adds other hooks, right?)
[18:49] <jam> So a minimum to be forward and backward compatible is probably dispatch + start + upgrade-charm
[18:49] <jam> I think those are all the entry points that you could hit
[19:23] <facubatista> jam, no "install"?
[20:01] <Chipaca> facubatista: yes, also install
[20:01] <facubatista> I included it just in case
[20:02] <Chipaca> :)
[20:03] <facubatista> So, First version of the build command: https://github.com/canonical/charmcraft/pull/4
[20:03] <facubatista> and with that, /me eods
[20:04] <Chipaca> facubatista: huzzah
[20:05] <Chipaca> facubatista: have a good'un!