Account for backing scale factor.

[sandinmyjoints]

On a high DPI display, like an Apple retina display, frame-char-width needs to be multiplied by the scale factor to accurately calculate width and height.

[pitkali]

I'll need to take a closer look. Good job for accounting for the possibility of there not being a backing scale factor to speak of. I find it interesting that frame-char-width would not already account for this, to be honest. Not surprising though — I have experience with projects terribly confused about different ways of measuring on-screen dimensions.

Which version of Emacs are you using on Mac? I mean, I can see in some patch I quickly googled that mac port should have this, but that doesn't tell me it needs this particular modification, and I'd prefer this to work for both mac port and stock version.

[sandinmyjoints]

I'm running https://bitbucket.org/mituharu/emacs-mac/ (Emacs Mac port), in fact, I built the latest release today. I was also surprised that it didn't seem to account for this, but it sure seems to be the case on my machine. Without this patch, I see about half the error message text. With it, I see all of it.

[pitkali]

I see. I just need to find some time in between travelling to test the stock version, then.

[pitkali]

Hey, so it took a while, but I finally got around to deal with this and I can't see any problems on my MBP with retina display. Could you explain how to reproduce the problem?

[pitkali]

By the way, did you upgrade to Catalina? I haven't done that yet and was wondering if it might be the reason I cannot see any issues here.

[sandinmyjoints]

I'm on Mojave, not Catalina.

I run emacs-mac in fullscreen, and this problem manifests as seeing a pos-tip tooltip showing text that is only about half of the text of a message from, for example, a linting error. I don't have time to try to put together a minimal repro, unfortunately.