Micha Reiser
Micha Reiser
> Yes, I was thinking about that. I feel it goes against the spirit of Black's AST safety check if we modify comments (it doesn't actually violate the check, of...
It could make sense to run ty locally with `TY_MEMORY_REPORT=full` between main/this branch. One reason why memory usage might be reduced is that this PR deduplicates some salsa metadata because...
> After rebasing on main the sphinx memory difference went away and now there's a diff on prefect, feeling wary. We round the the memory usage numbers to get more...
It looks "correct" in the sense that it fixes the bug in my PR
> Oh does it? It looked like it caused some new problems too :) It does. So I think you might be on to something but I honestly don't know...
> @MichaReiser CI is green now 🤔 Unfortunately, it doesn't remove the prefect diagnostics anymore.
It seems the main issue is that Django migration insists on the empty line. Do you know why Django migration adds the empty line? Is this something inherent to the...
> I think what I'd really appreciate is a clear signal from the ruff team whether this is something that will be supported or not. If it is, we'll gladly...
I'll add this as a sub issue to https://github.com/astral-sh/ty/issues/175 as it is about the same functionality -- some way to share configurations. @aeroyorch could you say a bit more about...
One major concern with this approach is that memory usage is likely to grow significantly (~projects*files) because we parse and analyze the same file multiple time, especially because we start...