13:44 <dnkl> ngo

13:45 <dnkl> ngortheone: neovim enables the kitty keyboard protocol in foot, but not in alacritty (last time I checked)

13:45 <dnkl> might be a decode bug in neovim

13:47 <dnkl> latest alacritty also has kitty kbd support. Not sure if neovim enables that or not, but fwiw, I see the exact same behavior in alacritty; Shift+F6 is interpreted as F18

13:48 <dnkl> alacritty 0.13.1, neovim 0.9.5

14:50 <ngortheone> dnkl this happens in midnight commander too

14:51 <ngortheone> Shift-F6 is to rename a file in place

14:51 <ngortheone> when I do this is in foot+mc it prompts me to delete the file instead

14:52 <ngortheone> so I think it is unrelated to the application, but it is related to how foot emits events

15:03 <dnkl> nah, I doubt it

15:13 <dnkl> show me the escape sequence emitted by foot *inside* neovim

15:24 <ngortheone> how can I do that?

15:27 <ahesford> width = round(scale * round(width / scale)) kind of bugs me

15:28 <ahesford> I'm thinking about the ramifications of capturing the integral numerator provided by fractional-scale and doing width = round((width * scale) / 120.0) to avoid the double-round

15:43 <dnkl> ngortheone: no idea...

15:44 <dnkl> ahesford: the point of the double round is to ensure the width we pick can be converted back and forth between logical and physical coordinates.

15:44 <ahesford> that should also work using the rational scale and a single round after the fact

15:53 <ngortheone> I used showkey -a. Here is what shift+f6 emits in foot <ESC>[17;2~

15:53 <ngortheone> and in alacritty: <ESC>[17;2~

15:53 <ngortheone> hm...

15:53 <ngortheone> why do apps behave differently?

15:55 <dnkl> ngortheone: I told you: neovim uses the kitty keyboard protocol. When you run showkey, you see legacy key codes

15:57 <dnkl> ahesford: right, but the scale factor is used in so many places that I don't think it's worth changing.

15:58 <ngortheone> does midnight commander support kitty protocol too?

16:00 <dnkl> ngortheone: probably not. But here's the thing; in the legacy keyboard protcol, the escape codes for F18 and Shift+F6 *are the same*

16:02 <ngortheone> it still does not make any sense. Shift+F6 works fine in alacritty+mc, but doesn't in foot+mc

16:03 <ngortheone> if mc does not support kitty protocol - it receives the same old legacy escape code from the terminal

16:03 <ahesford> dnkl: makes sense

16:04 <dnkl> ngortheone: for me, alacritty+mc behaves the same as foot+mc - it prompts for deletion

16:05 <ahesford> for all logical sizes between 0 and 3840, and all integer scale numerators between 60 and 360 (scales 0.5 -> 3), there are 1844 pairs of width and scale where the current implementation differs from the alternative

16:06 <ahesford> in those cases, the current double round is always one pixel smaller

16:08 <ngortheone> dnkl: think I found a clue. My alacritty has TERM=xterm-256color, but foot has TERM=foot

16:09 <ngortheone> when I set TERM=xterm-256color in foot - Shift+F6 works fine

16:11 <dnkl> ngortheone: guessing they check the *name* of the terminfo and change their behavior...

16:12 <ngortheone> Yeah. My guess is that there are is no terminfo for foot or alacirtty

16:12 <ngortheone> on my system at least

16:12 <ngortheone> ok, does foot have its terminfo file? can I somehow point apps to it explicitly?

18:35 <dnkl> ngortheone: you're already using foot's terminfo

21:12 <ngortheone> dnkl: are there any downsides to setting TERM=xterm-256color ? What features will I loose?