| *** gvincent has joined #ara | 08:45 | |
| *** gvincent has quit IRC | 08:46 | |
| *** gvincent has joined #ara | 08:47 | |
| *** gvincent has quit IRC | 09:11 | |
| *** sshnaidm|afk is now known as sshnaidm| | 09:16 | |
| *** sshnaidm| is now known as sshnaidm | 09:16 | |
| *** gvincent has joined #ara | 09:24 | |
| *** odyssey4me has joined #ara | 09:42 | |
| larsks | dmsimard: would it be useful if ansible objects (playbooks, tasks, etc) had a "description" field that could be used like a python docstring? I figure ara could use that in the ui to provide additional information about plays, tasks, etc. | 12:44 |
|---|---|---|
| ara-slack | chief_wrench: Isn’t that what the name: property is for? | 12:45 |
| dmsimard | larsks: there's a thing called "labels" that I'd like to be a thing in 1.0 that isn't exactly what you're asking for but I was asking myself it it would be relevant to apply to other objects than just playbooks | 12:46 |
| dmsimard | labels are generic and would be used to create arbitrary groups of playbooks (or hosts, or tasks, or...) | 12:46 |
| larsks | dmsimard: I'm asking because I've always wanted documentation like that to be a first-class object in playbooks rather than comments so that it can be machine-accessible, and it turns out the change is basically a single line of code to ansible. | 12:47 |
| larsks | Curious if other folks would find it useful or if it's just me. | 12:47 |
| dmsimard | larsks: how would you see that working in practice ? | 12:47 |
| dmsimard | would it be an ansible variable ? | 12:47 |
| larsks | No, just a new attribute on plays, tasks, etc. In addition to 'name', you could have e.g. 'description'. Like this: https://gist.github.com/larsks/e6e2ff29873716fc2aaa04a129904ebd | 12:48 |
| larsks | I figured a tool like Ara would be able to use that information to (optionally?) display play/task/etc documentation in the UI. | 12:49 |
| dmsimard | I agree with @chief_wrench that it overlaps a bit with name but we can argue that the name would be succint while the description would be verbose | 12:49 |
| dmsimard | if description gets added to ansible, ara can definitely pick it up and display it | 12:49 |
| larsks | Right, but the question I am asking is more along the lines of "do you think this is a useful idea or am I barmy?" | 12:50 |
| dmsimard | ah, apologies, was looking at it from an ara perspective | 12:51 |
| larsks | Well, that's part of the question, right? Do you think having access to extended documentation like this be useful to Ara. | 12:52 |
| dmsimard | I make a distinction between what would be useful to ara and what would be useful to users :p | 12:54 |
| dmsimard | I mean, the point of the description field would be to allow users to better understand what the play or task was about | 12:54 |
| *** irclogbot_3 has quit IRC | 12:55 | |
| *** irclogbot_2 has joined #ara | 12:55 | |
| larsks | Right, but arguably you can get that from comments as well. I see the primary benefit being making this information available in a machine-readable fashion. | 12:56 |
| *** altlogbot_2 has quit IRC | 12:57 | |
| dmsimard | what you linked is just a string, though | 12:57 |
| dmsimard | would you see description becoming something more than a string ? | 12:57 |
| *** altlogbot_2 has joined #ara | 12:57 | |
| dmsimard | for example, zuul has some documentation concepts to describe different ansible roles and jobs | 12:58 |
| larsks | I was not assuming it would be any more structured than a string, no, other than that the string could be markdown. | 12:58 |
| dmsimard | ok, just to give you an idea of what I meant with Zuul: https://opendev.org/zuul/zuul-jobs/raw/branch/master/roles/build-docker-image/common.rst which gets rendered as https://zuul-ci.org/docs/zuul-jobs/container-roles.html#role-build-docker-image | 13:01 |
| larsks | Sure, but that's just like dropping a README.md next to your python code: you probably still use docstrings in your classes and functions. | 13:04 |
| larsks | Anyway, it sounds like your opinion is mostly "meh", which is fine. I was really just looking for an opinion other than my own :) | 13:05 |
| dmsimard | larsks: Over time I have disciplined myself to try and give good names to plays and tasks (even if they end up being a bit verbose) | 13:06 |
| dmsimard | I see the use case for the description field. I would use it when I need it (where I would have otherwise put a comment, perhaps) but not for every task. | 13:06 |
| dmsimard | If it ends up landing, there's no problem to make it available in ara | 13:09 |
| *** dbpiv has joined #ara | 13:38 | |
| *** bwatson has quit IRC | 13:55 | |
| *** etienne has quit IRC | 14:29 | |
| *** spiette has quit IRC | 14:52 | |
| *** spiette has joined #ara | 14:54 | |
| *** etienne has joined #ara | 15:12 | |
| *** sshnaidm is now known as sshnaidm|off | 15:23 | |
| *** dbpiv has quit IRC | 15:35 | |
| *** altlogbot_2 has quit IRC | 16:09 | |
| *** altlogbot_3 has joined #ara | 16:11 | |
| *** altlogbot_3 has quit IRC | 16:43 | |
| *** altlogbot_1 has joined #ara | 16:43 | |
| *** altlogbot_1 has quit IRC | 16:53 | |
| *** altlogbot_1 has joined #ara | 16:53 | |
| *** gvincent has quit IRC | 17:11 | |
| *** gvincent has joined #ara | 18:31 | |
| *** gvincent has quit IRC | 18:36 | |
| *** gvincent has joined #ara | 18:36 | |
| dmsimard | apollo13: have you seen this kind of error before ? tox is suddenly confused about pyproject.toml and setup.py: https://github.com/ansible-community/ara/issues/33 | 21:45 |
| *** etienne has quit IRC | 21:52 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!