Open picnixz opened 2 months ago
Jelle suggested for "any standard library module that gets a significant issue volume" (emphasis added).
I agree with this.
But I'm not sure we'd want to add hundreds of labels for all the modules:
>>> import sys
>>> len(sys.stdlib_module_names)
288
It would increase the task of maintaining the labels: adding new ones, deciding what to do with removed modules -- remove or merge, and so on.
But I'm not sure we'd want to add hundreds of labels for all the modules:
Yes, it was more than what I expected :') But there are some that can require a topic I think, at least those that have projects! (like topic-unittest
or topic-logging
does not exist!)
Originally posted by @JelleZijlstra in https://github.com/python/core-workflow/issues/545#issuecomment-2286848045
When this was first suggested, I was like "ok why not?" and was a +0. But some time has passed again and I really think that it's worth it. It allows triagers to categorize issues way more simply when a title is cryptic and draw attention to the involved component.
I'm pretty sure that some labels will never be used or only used once because the module is forgotten (e.g., the
graphlib
or thenetrc
module). But I think we can make it once and for all. It's not really an issue for triagers to search through the labels and it would also let me feel a bit less sad when I can't find the label and can just putstdlib
on it.There are some modules that could be gathered under similar labels, e.g.,
annotationlib
could betopic-typing
unless Jelle wants something more precise for this specific typing part.On the other hand, what we could do as a first step, is to add labels for the existing projects. The rationale behind is that labels are visible on the issue list while projects are only visible if you open the issue (or am I mistaken because of refined github?)
I'll write a script to actually create the labels that need to be created (maybe not today) but I'd like to raise this question in a more visible way.