01:45 <llyyr> is it expected that animated cursors are stuck on the first frame when hovering foot?

01:46 <llyyr> I'd have thought the cursor-shape protocol would allow the compositor to handle it

02:58 <yonson> Is there a standard pattern to using systemd user socket and sway on startup? I seem to only be able to use the new socket file after manually pullin in the WAYLAND_DISPLAY env var for systemd and calling start on the socket. Ideally, the socket would begin on startup after the env is ready, but struggling with that part. Realted PR: https://codeberg.org/dnkl/foot/pulls/1281

03:06 <yonson> I guess i could call "systemctl --user start foot-server.socket" in my sway config, just not sure if that is a footgun or something

04:32 <Arnavion> You'd call `systemctl --user start graphical-session.target`, and any unit that is part of that target and which you had previously `enable`d will start automatically

04:33 <dnkl> llyyr: yes, assuming the compositor implements animations... don't think sway does

04:34 <llyyr> sway does now

04:34 <llyyr> it works on swaybg at least, but not on foot (or mpv for that matter, which is also using the cursor-shape protocol)

04:35 <llyyr> https://github.com/swaywm/sway/pull/7571

04:36 <dnkl> llyyr: have you checked foot is actually using the new protocol? I.e have you checked the log output? or foot --version?

04:36 <llyyr> foot version: 1.15.1 -pgo +ime +graphemes +fractional-scaling +cursor-shape -assertions

04:37 <llyyr> I doubt it's a foot issue since it doesn't work with mpv either

04:37 <llyyr> I'll just ask in #sway

05:41 <p00f1> llyr: that sway PR isn't merged yet

05:41 <p00f1> llyyr *

05:41 <llyyr> oh...

05:41 <llyyr> (why did I think it was?)

05:54 <dnkl> if that's the case, foot should warn about it (on stderr, not in the foot window)

10:40 <Morteza> Hey, does anybody know why I can't disable cursor blinking even tough i have it disabled in config

10:40 <Morteza> corsponding line in config

10:41 <Morteza> [cursor]\nstyle=beam\nblink=no

11:10 <dnkl> Morteza: guessing your shell overrides it

12:14 <ahmadraniri[m]> Hello since foot version 0.15.x came, my compositor won't launch foot, 0.14.x works fine. I saw an issue about it. When I run foot using gdb, the bt told me about wl_proxy_get_version stuff. My compositor is https://gitlab.com/lidgl/tinyway. Any clue to fix it ? Should I make adjustment to my compositor ? Thanks.

12:21 <dnkl> ahmadraniri[m]: doesn't foot log any errors?

12:28 <ahmadraniri[m]> <dnkl> "ahmadraniri: doesn't foot log..." <- Foot gave me some warnings about no XDG activation, fractional scaling, and IME disabled before segmentation fault.

12:28 <dnkl> please don't use matrix replies here; this is an IRC channel

12:29 <dnkl> anyway, absolutely sounds like a compositor issue. Try #wlroots

12:29 <ahmadraniri[m]> Oh my bad. Thanks dnkl.

12:29 <dnkl> regardless of channel, you'll have to link to the actual backtrace

12:30 <dnkl> no one will be able to help you without it

13:07 <dnkl> TIL emacs can use the kitty keyboard protocol: https://github.com/benjaminor/kkp

15:07 <yonson> Arnavion: that makes sense, thanks!

16:30 <p00f1> dnkl: how do you even use terminal emacs without it

16:31 <p00f1> even things like the default keybinding for xref-find-apropos don't work

16:34 <p00f1> also is #footerm:matrix.org yours

16:42 <dnkl> no, that's not mine

16:42 <dnkl> you use xterm's modifyOtherKeys protocol

16:43 <dnkl> before emacs 29, you had to:

16:44 <dnkl> (add-to-list 'term-file-aliases '("foot" . "xterm"))

16:50 <dnkl> fwiw, I've been using terminal-only emacs for years now