Hello all, it’s time for the fortnightly coding standards meeting.
| quietone |
Still at RTBC |
| Alex Skrypnyk |
@quietoneSorry, I'm a bit confuse in which point to add things, so will just choose this one :slightly_smiling_face: |
| Alex Skrypnyk |
[#2464123]this is marked ass RTBC, but I do not think it should be RTBC. It requires further discussion based on my last comment and a comment from pfrenssenI'm not sure if I should be changing that status or coding standards lead should do it.That proposed change is quite a big change - it involves disabling, rather then enhancing the rule, which will affect many projects and teams. |
| Jonathan1055 |
I agree with @Alex Skrypnyk that this issue is no longer RTBC. It was set to that status four years ago, and there have been 12 more comments since then. The proposed solution in the issue summary is no longer what might be agreed on. |
| gapple |
++ I attempted an update to the full proposed text from #25 with some changes for clarity based on further feedback.It looks like there's consensus that the rule should be changed, but the exact wording needs agreement before it's really RTBC (edited) |
| gapple |
The coder sniff could be moved to the DrupalPractice ruleset as a warning, since it's likely to give false positives |
| Jonathan1055 |
Regarding the Coder sniff, I don't see any benefit in moving it into DrupalPractice. Either we can fix the sniff to make it detect and enforce the updated standard, or, if the agreed new standard cannot be programmed (ie it involves the content/context of the type of comment which is not detectable via programming) then we disable and remove the sniff. |
| quietone |
There is an issue where people have put their hand up to be on the committee. #3378689: Add more coding standards maintainers, part II |
| quietone |
But we don't have a process for 'appointing' or even what the make up of the group is. |
| quietone |
Is there an issue that I have overlooked? |
| urvashi_vora |
This is the same issue, I also wanted to get a feedback on, as I am still interested to be a part. Please let me know If anything is to be done from my end |
| quietone |
What are you thoughts on things like how many people should be on the committer, what groups the represent etc. |
| quietone |
For example, currently all the members are core committers but the group should have representation from other parts of the community. |
| urvashi_vora |
It seems reasonable, but I believe diversifying the committee by adding members from different parts other than core committers could be valuable. As different people could bring different perspectives and fresh insights. To successfully include new maintainers who aren't core committers, it's important to approach the decision-making process thoughtfully. Confirming their genuine interest and commitment to active participation is essential, rather than just a temporary involvement. |
| urvashi_vora |
However, I would also value getting the opinions of different people within the committee before coming to a certain point. |
| catch |
@quietone I believe the process is that the TWG members (currently === the coding standard committee) can just add people to the coding standards committee. |
| quietone |
That work for me. |
| quietone |
What about what groups are represented on the committee? |
| quietone |
Do we want a representative for the Coder project? |
| Björn Brala (bbrala) |
I'm not sure if setting a benchmark for that is really something we should be worrying about. If anything I think that most valuable is 2 sets of people; some core maintainers (for connections, and effects on the framework), on the other side, community representation (but also preferably with experience in the php ecosystem outside drupal). |
| quietone |
So, that could be a core contributor or a project contributor but they do not need to be a project maintainer. |
| catch |
Yeah I also think formal representation isn't the issue, but just generally having active people who aren't core committers would be good. Especially it's about people who work on contrib more than we do. |
| quietone |
Then what about group size? |
| catch |
That I don't know. Enough to keep things running if people are busy/away but not too many that it results in lots of overhead. I would not really want to go above 12 I guess? |
| Björn Brala (bbrala) |
To keep it manageable i would aim at max 7 or something. |
| Björn Brala (bbrala) |
ha (edited) |
| catch |
I think we will be around 7 if we add the people on that issue, but then that is all the volunteers we have so far. Also I didn't actually count. |
| Björn Brala (bbrala) |
If we define a max, I would choose an uneven number btw |
| quietone |
Dare I ask. Why? |
| catch |
It's easy to contribute without being a maintainer, but we just don't want it to go back down to zero/one for years again and stall. |
| Björn Brala (bbrala) |
If there is a vote, you cannot have a deadlock then :stuck_out_tongue: |
| Björn Brala (bbrala) |
Might be a moot point though, but have seen that go wrong in some of the groups i have been in |
| quietone |
That assumes the group will use voting for decision making. I don't think we have made that decision. |
| Björn Brala (bbrala) |
Yeah i know, just thinking ahead if there is a lot of discussion ona topic you could end up with 'a majority is for' |
| catch |
More likely that stuff we disagree on gets indefinitely deferred :wink: |
| Björn Brala (bbrala) |
hahahaha |
| Björn Brala (bbrala) |
Unfortunately very very true |
| quietone |
Well, that is why I advocate for consensus decision making along with a time frame for decisions. |
| catch |
But for me I am more interested in getting stuff through that's uncontroversial, just that has been incredibly hard so far. |
| Björn Brala (bbrala) |
yeah |
| quietone |
Yea, there is enough to do with getting a system restarted. |
| Björn Brala (bbrala) |
We could just adopt some psr's or symfony standards with a few differences and have to think less way less of cs rules :wink: |
| Björn Brala (bbrala) |
(80% kidding there) |
| quietone |
I am not sure where to document agreements on this. The project page seems the best option right now. |
| catch |
Someone suggested that but both we and Symfony left PHP-FIG, so for me not happening. |
| catch |
Also we have infinitely more standards than either afaik. |
| Björn Brala (bbrala) |
Not sure that the fact we have more standards is a good thing. And sure, the PHP-FIG debacle makes those things a little harder. Im sorry to derail this one a little, this discussion i started here is offtopic for this thread. |
| catch |
fwiw I think it's good that the current RTBC issues are loosening standards a bit. |
| catch |
Some of them made sense for procedural code, but not for classes and especially not PHP 8. |
| Björn Brala (bbrala) |
yes |
| Björn Brala (bbrala) |
Things will get better also because of this group working again so we should be less stuck in the past going forwards. |
| Björn Brala (bbrala) |
I have this feeling that we are duplicating a lot of effort to have docs that fully reflect the standard when a lot of those are covered by automation. Why do we need to document those things twice? That feels like a way to make maintaining the standards harder. |
| quietone |
Interesting idea. But I don't see how it is documenting twice, in that documentation is not code. |
| Björn Brala (bbrala) |
Sure documentation is not code, and there is merit in documenting things outside rules that verify standards.In the long run though, it would be great if the drupal coding standard could shrink. I would be extremely happy if we work towards needing less and less documentation.If you look at frameworks like Symfony and Laravel, those coding standards are way more standard with only some specifics. (https://symfony.com/doc/current/contributing/code/standards.html https://laravel.com/docs/5.0/contributions#:~:text=be%20promptly%20addre....)I think it would be a really really good long term goal to find ways to minimize the load on the community to get coding standards right and moving and not describing things that have been automated in docs could be a way to work towards that. |
| Björn Brala (bbrala) |
To be very honest, I would love to see Drupal at some point adopt more established coding standard used in the community, with a relatively small set of exceptions where we feel that is really needed. But I know that is not realisitic in the short term. |
| quietone |
Points taken. But it is challenging when we don't follow PSRs. |
| quietone |
Certainly some long term goals here. |
| Björn Brala (bbrala) |
Yeah i know we cannot do this right now, and there is not even the governace ready to discuss this properly |
| Björn Brala (bbrala) |
So no worries, but i will be repeating this once in a while and see what would be an oppertune time to have this discussion proper. |
| Alex Skrypnyk |
@Björn Brala (bbrala)sorry, I do not understand what exactly this point is about
As a developer, I want to be able to see documentation for rules of every language we are using in DrupalI also want to see code examples that follow those rules.
A page like this is very valuable in teams, because people can reference specific parts when communicating.https://www.drupal.org/docs/develop/standards/php/php-coding-standards
I agree that manually supporting documentation while we have rules in Coder feels like a duplication of efforts. I would suggest to automate documentation building from the comments of the rules defined in Coder. this would allow to have consistent docs in Coder's codebase and public documentation built out of it and no more manual documentation updates.Drupal API site is built out the Drupal codebase. And it works :slightly_smiling_face: |
| Alex Skrypnyk |
As for "ours" vs "theirs" standards - this is a bit subjective - I like "ours" more :slightly_smiling_face: |
| gapple |
++
As a developer, I want to be able to see documentation for rules of every language we are using in DrupalI also want to see code examples that follow those rules.
|
| gapple |
There's also instances where the coder rules have had bugs or been stricter than the written code styles, and having the styles documented (and the discussion where they were implemented) was necessary to correctly adjust the automated coder rules. |
Comments
Comment #7
quietone commentedComment #8
quietone commentedComment #9
quietone commentedComment #10
quietone commentedCopy conversation '5' again.
Comment #11
urvashi_vora commentedVerified. I think all points are covered now. Thanks @quietone.
Comment #12
quietone commented@urvashi_vora, pointed out in Slack that I missed giving them credit. Doing so now.
Comment #13
urvashi_vora commentedThanks @quietone.
The meeting notes match the conversation in Slack.
Hence marking it as Fixed.