[wp-trac] [WordPress Trac] #61175: Integrate PHPStan into the core development workflow
WordPress Trac
noreply at wordpress.org
Mon Jul 21 17:31:54 UTC 2025
#61175: Integrate PHPStan into the core development workflow
--------------------------------------+-------------------------
Reporter: westonruter | Owner: justlevine
Type: enhancement | Status: assigned
Priority: normal | Milestone: 6.9
Component: General | Version:
Severity: normal | Resolution:
Keywords: has-patch has-unit-tests | Focuses:
--------------------------------------+-------------------------
Comment (by dmsnell):
> It's a shame that GitHub doesn't provide a good developer experience for
checks that are passing with warnings.
Agree with you on this, @johnbillion, and that’s part of why I think the
burden should fall on us before we proactively reject work. I’ve seen some
very good integrations that rely on interaction via comments, both on the
PR front-roll and via line-level comments.
> we'd still need a way for a committer to ensure nothing gets missed when
they come to commit the change
I wonder how far we could go simply listing these issues. Supposedly we
are all already reviewing code, and if there are errors noted on the PR
then it would fall within the review obligations to make sure they aren’t
overlooked. This feels no different than if you came to one of my PRs and
pointed out a subtle type-casting bug in my use of `in_array()`. Maybe I
respond with some explanation of why it looks wrong but isn’t, or maybe I
say thanks for catching my mistake before I push it to the world.
Having that interaction to me is part of embracing the humanity of code
review. If there’s no way to discuss it then it’s just a gate-keeper.
Maybe people prefer this, and I won’t stand in their way; just wanted to
share since the technologies and automations we build have real impacts on
the people who have no choice but to appease them or find workarounds.
I’m still a bit worried that if we provide only an approve/reject mode
without discussion, then people will try to find awkward workarounds
before the underlying issues can be identified, and then it becomes that
much harder to find and deal with them later.
----
Has anyone collected data on which changes //would// have been rejected
had we been employing this for the past year or so? It’d be informative to
know the breakdown of //caught errors that were missed in review// against
//rejected valid work//.
--
Ticket URL: <https://core.trac.wordpress.org/ticket/61175#comment:39>
WordPress Trac <https://core.trac.wordpress.org/>
WordPress publishing platform
More information about the wp-trac
mailing list