| Odd_Bloke | blackboxsw: Ugh, this Actions PR stuff is a real PITA. | 14:14 |
|---|---|---|
| meena | Odd_Bloke: so you should look at something more pleasant like… how to refactor cloudinit/net | 15:00 |
| * meena has a call with a customer in minus two minutes, andhopefully iwon't fall asleep utnil it actually happens | 15:03 | |
| Odd_Bloke | meena: Oh, I thought you'd got feedback from Ryan? | 15:04 |
| Odd_Bloke | I can still take a look, of course. :) | 15:04 |
| meena | Odd_Bloke: he added > [] — and that's all i've seen | 15:06 |
| Odd_Bloke | What more could you want? ;) | 15:06 |
| meena | Odd_Bloke: yeah, i dunno either, sometimes i just feel like i have too high expectations which i don't know how to communicate | 15:07 |
| Odd_Bloke | meena: So I'll take a look at that doc today. And let's be explicit: what feedback are you seeking? | 15:16 |
| Odd_Bloke | blackboxsw: Just dropped a wall of text regarding next steps for CLA actions: https://github.com/canonical/cloud-init/pull/164#issuecomment-573075057 | 15:16 |
| rharper | meena: I hope all of the comments made it | 15:46 |
| rharper | meena: in view mode, on the right, there is 💬 mark with a number to indicate comments, clicking that will show them on the right; I added comments to sections: Background, Code duplication | 15:48 |
| meena | rharper: aahh. so I discovered @goner's comment, and you appear to have answered it | 17:32 |
| rharper | meena: yes | 17:33 |
| rharper | meena: it took me a little bit to realize I could comment without making changes to the content; that's why you saw my [] edit ... | 17:33 |
| blackboxsw | Odd_Bloke: ok I responded to the wall of text and refactored the CLA signing PR to still get triggered on pull_request, and fail the status check with the CLA failure comment. If we still think we need to add a comment or label to PRs, we could leave in a top-level PR grooming action that'd walk open PRs and add labels. | 18:38 |
| Odd_Bloke | Ugh, not even being able to comment _sucks_. | 18:40 |
| Odd_Bloke | Really feel like "you won't be able to use this for the most basic GitHub action" should be more prominent in their docs somewhere. ¬.¬ | 18:41 |
| meena | 😬 | 19:18 |
| meena | rharper: i'll try to clarify re your comments. | 19:29 |
| rharper | meena: great; generally I think Odd_Bloke 's suggestion is the direction to go; but I wanted to make sure we understand the end goal of reuse; maybe "more portable to other BSDs" is what you mean vs. cross distro (which to me refers within the linux family variants) | 19:31 |
| meena | rharper: nah, pretty much all BSDs most of these things will work the same… except DHCP. | 19:39 |
| meena | s/BSDs/Linuxes/ | 19:39 |
| meena | i hope i have clarified. it a bit now: i reckon we can remove the code duplication section? | 19:40 |
| meena | rharper: my issue with Odd_Bloke's suggestion is how that if we put the functions into the Distro classes, then, uh, how do i get a distro object in cloudinit/net? | 19:41 |
| rharper | meena: re: code dupe; yes; though maybe it's worth a bug/PR to comment/change the naming so it's not as confusing ? | 19:43 |
| meena | rharper: they are private functions, so, yah | 19:44 |
| meena | rharper: if i started with *that* PR, that would be the ULTIMATE YAK SHAVE. | 19:45 |
| Odd_Bloke | meena: https://github.com/canonical/cloud-init/blob/master/cloudinit/distros/__init__.py#L85-L94 <-- this is where the Renderer is instantiated, so I think you could just pass `self` in there? | 19:47 |
| Odd_Bloke | (I'll defer to rharper on if that makes sense though.) | 19:47 |
| rharper | Odd_Bloke: for renderers, yes; for cloudinit.net ; which is the bulk of the porting work; I don't think we need Distro there; net module could likely get by with use of is_freebsd() variants; | 19:48 |
| rharper | with some import log, we likely could have 'from cloudinit import net' determine linux vs. bsd and pull in named methods | 19:50 |
| meena | rharper: https://github.com/canonical/cloud-init/pull/147/files | 19:50 |
| rharper | s/log/logic | 19:50 |
| meena | rharper: i did that … well, i started that a year ago, but i sucked at python / cloud-init. and so i, very naïvely, tried to translate all functions instead all public functions | 19:51 |
| meena | but from the layout of cloudinit/net it's not clear what's public and what's not | 19:52 |
| meena | perhaps a good first step, as i mentioned, it would be a good idea to start with marking the private functions first. | 19:53 |
| rharper | meena: everything is public that's not named with leading _ | 19:55 |
| meena | rharper: i know | 19:55 |
| meena | we need to mark the things that aren't used anywhere other than cloudinit/net with a _ | 19:56 |
| meena | i don't suck that bad at python! | 19:56 |
| meena | i literlly cannot remember what i was trying to do | 20:55 |
| meena | (that is usually a good place to quit, and sleep instead) | 21:20 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!