15:05:43 <bcoca> #startmeeting core public irc meeting 15:05:43 <zodbot> Meeting started Thu May 13 15:05:43 2021 UTC. 15:05:43 <zodbot> This meeting is logged and archived in a public location. 15:05:43 <zodbot> The chair is bcoca. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:05:43 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic. 15:05:43 <zodbot> The meeting name has been set to 'core_public_irc_meeting' 15:06:07 <Shrews> o/ 15:06:13 <mkrizek> \o 15:06:36 <bcoca> #topic https://github.com/ansible/ansible/issues/74558 15:07:34 <bcoca> in 'preview' discussion, most of core seemed to be of the mind 'this should not be a std core test' 15:08:01 <sdoran> o/ 15:08:29 <bcoca> that if we wanted this restriction, it should be in runtime code 15:09:05 * gundalow waves 15:09:28 <sdoran> It should be a configurable option in one way or another. 15:09:44 <sdoran> Core allows all the things, collection authors can choose to restrict that further. 15:09:49 <cyberpear> otherwise, folks would want backport of new plugin types 15:09:49 * gundalow waves 15:10:20 <gundalow> mattclay felixfontein: FYI we are talking about https://github.com/ansible/ansible/issues/74558 ansible-test: Error if non-standard dirs exist under plugins/ 15:11:04 <bcoca> its a bit early for matt, PDT 15:11:19 <gundalow> ah, yup, hadn't looked at the clock 15:12:23 <gundalow> bcoca: could you please expand on `Also, if we are doing a 'extraneous dir check' i would extend this to verifying the full collection schema and 'warn' about those directories, just not error.` 15:13:30 <bcoca> yes, your check only affects 'plugins' subdir, if we are going to verify 'proper collection' i would verify ALL dirs for 'extraneous' .. but still not make it an error, just a warning 15:14:24 <gundalow> ah so https://docs.ansible.com/ansible/latest/dev_guide/developing_collections.html#collection-structure 15:14:24 <gundalow> though + `scripts` 15:14:45 <gundalow> I think `scripts` is where we put oldskool inventory scripts 15:15:05 <bcoca> still 'meaningless' to ansbile itself 15:15:23 <gundalow> nod 15:16:21 <gundalow> AFAIK this requirement (extra dirs under plugins) was because some collections made a new subdir that wasn't currently used by Ansible, and we were concerned that those dirs might conflict with a future thing in Ansible 15:17:31 <bcoca> well, not that worried about that .. also ended up with 2 "standards" 15:20:16 <bcoca> im tempted to punty this to a linting tool 15:20:30 <gundalow> OK, maybe we throw this on the backlog for a while, till there are more thoughts around the collection-only testing 15:20:33 <gundalow> bcoca: bingo :) 15:20:41 <gundalow> +1 15:21:30 <gundalow> If we have time, dericcrago and I would be interested in talking about about https://github.com/ansible-collections/community.general/issues/176 15:21:59 * dericcrago waves 15:22:32 <gundalow> comment added on 74558 15:25:39 <bcoca> #topic open floor 15:26:08 <gundalow> #topic https://github.com/ansible-collections/community.general/issues/176 15:26:11 <gundalow> :( 15:26:22 <gundalow> woops, wrong link 15:26:44 <gundalow> https://github.com/ansible/ansible/issues/74684 15:26:57 <bcoca> i was going to punt to galaxy team 15:27:12 <gundalow> Does Core have any thoughts/ 15:27:38 <gundalow> or any gotchas? 15:28:49 <bcoca> well, i would say that they serve diff purposes, core one is 'general' (just uses layout) while your template is for community collections (has licenses, code of conduct) 15:29:52 <gundalow> That's a fair comment 15:30:02 <dericcrago> off the top of my head I was thinking you could give it either a --bare or --opinionated to get one of the other 15:30:41 <bcoca> also, mulitple temlpates, --template community_collection 15:31:06 <bcoca> the design was always 'overridable' 15:31:49 <gundalow> dericcrago: so maybe do a table in #74684 showing the existing files (in both collection_template and ansible-core) and list which template they could apply to 15:32:16 <bcoca> add 'this is how you init this type of collection' with a-g init .. 15:32:22 <bcoca> to readme 15:33:37 <Shrews> heh, TIL about GH repo templates 15:33:58 <shertel> ^ me too 15:34:11 <tadeboro> Sorry for being late to the parts, but I think we decided last time that someone (me) will write the docs about this and call it a day. The extra check should go somewhere else. 15:34:13 <dericcrago> is `--template` already available? 15:34:27 <bcoca> gh templates have improved A LOT 15:34:57 <bcoca> --role-skeleton .. not sure equiv for collections 15:35:18 <shertel> --collection-skeleton 15:35:37 <gundalow> Shrews: shertel I think GH Collection Templates are ~ year old, 15:35:46 <tadeboro> My plan is to document the "these dorectories are safe to use" once I start teacing our internal teams some collection development next week. 15:37:19 <gundalow> tadeboro: cool, is that something that could end up on https://docs.ansible.com/ansible/latest/dev_guide/developing_collections.html#collection-structure 15:37:44 <tadeboro> gundalow: This is where I wanted to put it, yes. 15:37:45 <gundalow> dericcrago: Any other questions or thoughts you want to share at this point? 15:37:49 <gundalow> tadeboro: thanks :) 15:38:03 <dericcrago> ok, with `--collection-skeleton` is there a specific directory we can use so that these skeletons would be considered 'builtin'? 15:38:48 <bcoca> yes, easy to add logic that w/o fullpath/unfound relative to look into 'data dir' which is where we keep em 15:38:55 <tadeboro> https://github.com/ansible-community/community-topics/issues/15 contains a discussin topic related to the topic of linting. 15:39:11 <bcoca> also wil lprobably rewrite to --skeleton as common option 15:40:17 <dericcrago> ok, I'll start poking around at it, thanks! 15:41:49 <gundalow> Thank all 15:41:52 <gundalow> Nothing else from me 15:43:12 <bcoca> k, if nothing else ending in 3mins 15:44:55 * gundalow -> afk 15:47:05 <bcoca> #endmeeting