Aaron Jensen

Reported upstream: bug#41422 https://debbugs.gnu.org/cgi/bugreport.cgi?bug=41422

@sergeyklay I simply remove the `lower-frame` call in my local copy of `company-posframe`. I'm not even sure why it's there as everything seems to work without it. Perhaps a patch...

@tumashu thank you for the fix. It looks like it was fixed upstream as well, so it will be fixed in Emacs 28: https://debbugs.gnu.org/cgi/bugreport.cgi?bug=41422

I can reproduce this as well on Emacs 27.1 on macOS AFAICT, `fit-frame-to-buffer` is resizing too narrowly by one (or two, if there's meant to be a space to the...

Hi @jmuheim, I wasn't able to successfully use it, no. I had more luck with this pull request by @Kriechi, but it still caused several of my tests to fail....

From my limited debugging, it looks like it may be a problem with the way that `dired-noselect` works. It appear to trigger vertico to reopen.

That is **much** better. This is a profile from several helm projectiles, which is the test I used before. As you can see, `get-frame-persp` is no longer getting top billing...

By the way, @Bad-ptr could I trouble you to explain what was happening here? I do not understand what would have caused this predicate to get called as a result...

I see, thank you very much for the explanation and the fix!

That seems like an interesting idea. Would you propose that as a change to emacs proper? It doesn't redefining it here would be good, yeah?