The news is a bit inaccurate. It's more about client process creating a file that server process does not have access to. It is permission issue, but there are many kinds of permission issues and they are not all fixed. Could you be a little bit more accurate on the news?

Also cc @godlygeek and @pablogsal . I think the fix is fine but maybe you have other suggestions.

maybe it's just a wording thing? I think the news is accurate but perhaps too broad. the script from pdb can't be read in the process that's being debugged

I think it's a wording thing. I think the issue is that, we did not fix it from the reading side. It's not like we fixed something so the process can read that file. We granted the read permission after we wrote it - that's what's fixed.

do you have a suggestion for the contents of the news file?

I would do something like

Grant read permission to the attach script for all processes for :func:`pdb.attach` so it can be read by :func:`sys.remote_exec`.

I'm also not the expert in news and I'll probably wait for someone else to see if this is a better description.

Perhaps

Ensure :func:`pdb.attach` script is readable by remote Python process.

Fix looks reasonable to me. For the news entry, maybe "Fix PermissionError when a superuser attaches PDB to another user's process" or something like that.

@asottile you can sum the suggestions up and make it a bit clearer. news is not as strict as the real documentation. Good to have a clear one but won't be a serious blocker.

Thanks @asottile for the PR, and @gaogaotiantian for merging it 🌮🎉.. I'm working now to backport this PR to: 3.14.
🐍🍒⛏🤖

Thanks for the PR!

GH-134616 is a backport of this pull request to the 3.14 branch.