dnkl changed the topic of #foot to: Foot - fast, lightweight and minimalistic Wayland terminal emulator || 1.8.2 || https://codeberg.org/dnkl/foot || channel logs: https://libera.irclog.whitequark.org/foot
themainman has quit [Quit: WeeChat 3.1]
lilblacky has joined #foot
novakane has joined #foot
gustafla has joined #foot
<gustafla> Hi! I really like this terminal. Are there any plans to get it into Arch Linux official repositories? After using gentoo for a while, I'm a bit allergic to compiling updated
<gustafla> updates
gustafla has quit [Quit: Client closed]
gustafla has joined #foot
<dnkl> In short, this isn't up to us
<gustafla> I see. Glad to hear it's possible. Meanwhile someone (me if I really wanted to bother) could set up a foot-bin AUR package. Unless your optimizations also include -march=native?
<gustafla> But yeah thanks for your work!
<dnkl> gustafla: no -match-native. It uses makepgs CFLAGS, except it forces -O3, mostly due to a regression in GCC where -O2 and PGO doesn't produce better result than regular -O2
<dnkl> gustafla: you're welcome :)
gustafla has quit [Quit: Client closed]
<sosodank> you pretty much have to enlist an Arch TU to move your package into community
<sosodank> and they're then going to have to maintain it
<rcf> Evangelize on the Arch channels and maybe someone will start using it and package it
<rcf> That's my hope anyway.
<sosodank> have you seen -march=native improve performance for foot?
<testuser[m]> march=native doesn't have much difference than setting it to an x86 v2/v3/v4 march
<rcf> I've never actually looked into it, mainly because I've never run into any performance issues
<rcf> Hence no need to expend effort on proper benchmarks.
<sosodank> right, i meant that more for gustafla
<dnkl> Anyone know a terminal application that has support for mouse wheel *left/right* events? This corresponds to mouse buttons 6 and 7 in X.
<ifreund> my mouse doesn't even have those buttons :D
<dnkl> ifreund: I had to borrow one :D
* dnkl is, obviously, adding support to foot
<sosodank> i've seen 8 and 9, but no 6 and 7
themainman has joined #foot
<dnkl> sosodank: 8 and 9 are typically the extra "side" buttons
<dnkl> But yeah... Not sure how common mouse with a tiltable scroll wheel is...
<dnkl> Anyway, foot now generates escapes for them (when the application enables mouse event reporting)
<dnkl> Just wanted to see it in action :D
novakane has quit [Ping timeout: 272 seconds]
novakane has joined #foot
<dnkl> Emacs (of course) supports it
<dnkl> (require 'mwheel)
<dnkl> (setq mouse-wheel-tilt-scroll t)
<dnkl> (setq mouse-wheel-flip-direction t)
<dnkl> (mouse-wheel-mode t)
<testuser[m]> has anyone faced this issue where foot shows tiny grey lines at random places that go away when you move your cursor near them ?
<dnkl> testuser[m]: you on GNOME?
<testuser[m]> no
<testuser[m]> sway
<dnkl> Fractional scaling?
<testuser[m]> 1.25
<dnkl> That's likely related
<testuser[m]> the "lines" are 1 column wide and appears at places the cursor has visited
<testuser[m]> like in a text editor
<dnkl> It *is* remnants of the cursor
<testuser[m]> yeah thats the issue
<dnkl> My guess is it's either a compositor or driver issue
<testuser[m]> the linked one
<dnkl> testuser[m]: try the workaround mentioned there: tweak.damage-whole-window=yes
<dnkl> That's known to work on at least the GNOME+fractional scaling combo
<testuser[m]> is it supposed to be exactly `tweak.damage-whole-window=yes` ? it says "invalid key"
<ericonr> section tweak
<testuser[m]> oh
<ericonr> the key will be "damage-..."
<testuser[m]> seems to be working fine now
<testuser[m]> i mean the cursor bug
<dnkl> testuser[m]: might be rounding error or similar when the compositor downscales the application provided surface.
<dnkl> But could also be a driver issue; I've never been able to reproduce on Sway
<dnkl> Or, perhaps the window and/or cell size matters
<dnkl> testuser[m]: in https://gitlab.gnome.org/GNOME/mutter/-/issues/1071, it looks like it _used_ to work in GNOME, then they changed something.
<dnkl> Even though that's GNOME, and not Sway, it shows that foot isn't the only client suffering from this
<dnkl> And that a change in the compositor triggered it
<testuser[m]> hmm
<dnkl> testuser[m]: is your monitor rotated? Trying to figure out why I can't reproduce...
<testuser[m]> no
<testuser[m]> my editor is vis so i guess you could try it with some populated file and see if it happens
<testuser[m]> it happens less often in regular terminal usag
<testuser[m]> e
<dnkl> testuser[m]: if you start foot from another terminal it will log stuff on stderr. What cell size does it say it's using?
<testuser[m]> width 21 height 47
<dnkl> testuser[m]: definitely a compositor issue; I was able to reproduce with a specific window- and cell size, but only at a specific row.
<dnkl> Move the window one pixel up or down and the issue went away
<dnkl> Move the window back and the issue reappears
<dnkl> I updated #350 with my observations
yyp has quit [Read error: Connection reset by peer]
yyp has joined #foot
<dnkl> ericonr: do you have a plan for how to deal with ncurses providing a foot terminfo?
<dnkl> I need to handle it in Arch AUR in some way... 😅
<dnkl> I'm considering doing what kitty does; install our terminfo to a non-standard location and setting the TERMINFO environment variable...
<ericonr> dnkl: does the new file conflict directly with your "official" ones? :/
<dnkl> Or maybe TERMINFO_DIRS. It would be nice if it would search the default location if our terminfo isn't present (for e.g SSH)
<ericonr> I think TERMINFO sounds bad, since you could end up screwing up launching *another* terminal from within foot
<dnkl> ericonr: the ncurses one lacks a couple of standard ones (some due to misunderstandings)
<ericonr> appending to TERMINFO_DIRS sounds more reasonable
<dnkl> But more importantly, it's missing the non-standard capabilities
<dnkl> Used by e.g tmux
<dnkl> Yeah, I didn't really like TERMINFO. That would break ssh too, if the remote system doesn't have our terminfo
<testuser[m]> did you try contaciting thomas dickey
<testuser[m]> contacting
<ericonr> dnkl: and I meant conflict as in default file location :p
<dnkl> testuser[m]: I'll do that for the things that are obviously wrong
<dnkl> But he's not going to add the non-default capabilities
<ifreund> huh? how'd foot terminfo end up in ncurses?
<ifreund> guess https://codeberg.org/dnkl/foot/issues/335 can be closed though...
<dnkl> Dickey added it all by himself. It's not released yet though. I just happened to see the changelog for the upcoming release
<dnkl> ifreund: ^^
<ifreund> kinda neat that it just happend without prompting as it appears alacritty/kitty needed to do
<dnkl> I have no idea how this came to be...
<ifreund> having it out of sync with the official foot terminfo is super awkward though :/
<dnkl> ifreund: supposedly he tested the terminal, and changed the capabilities he felt was wrong
<dnkl> For example, he removed the status-line capabilities because "foot doesn't set the window title"
<dnkl> Which is wrong
<dnkl> Unless you test on Weston or GNOME, with CSDs
<dnkl> (foot-1.8.3 will have support for title in CSDs too, btw)
<ifreund> nice
<dnkl> Apparently, bell also doesn't work
* ericonr knows people who like the bell with sound
<testuser[m]> what bell
<testuser[m]> does it refer to the boink sound when you press backspace on an empty line
<dnkl> testuser[m]: "echo -e '\a'"
<dnkl> testuser[m]: there's a couple of different things foot can do with it.
<dnkl> See the "bell" section in foot.ini(5)
<dnkl> ericonr: funny thing is, "tack" (test application) says exactly that foot bell: "you should hear a bell, otherwise it's not working"
<dnkl> S/foot/for
<ericonr> hm, so if you do a graphical bell, it's non conforming?
<dnkl> ericonr: I guess :)
<dnkl> Not that it matters; it doesn't affect any capabilities in the terminfo - it's just a comment in the terminfo source file
Hrmmm has quit [Quit: Hrmmm]
<dnkl> Fwiw: setting TERMINFO_DIRS seems to work best; it'll still search the default location (but only *after* searching TERMINFO_DIRS)
<ericonr> nice
<ericonr> supposedly you'd want to append to it
<ericonr> because users might have things already
<dnkl> ericonr: of course :)
phin has joined #foot
<phin> hi
<phin> is there a way to get gnome titlebars with foot?
<dnkl> phin: I'm guessing you actually mean GTK titlebars, in which case the answer is no
<dnkl> In GNOME, all applications draw their own titlebar
<phin> darn :(
<dnkl> Since foot isn't based on gtk, it won't, and can't, draw gtk-styled bars
<phin> ya that makes sense
st3r4g has joined #foot
<ifreund> if you want to go complain to someone, complain to the gnome and gtk people, they're the ones who pushed so hard for CSD everywhere on wayland :P
st3r4g has quit [Quit: Lost terminal]
themainman has quit [Quit: WeeChat 3.1]
themainman has joined #foot
themainman has quit [Quit: WeeChat 3.1]
themainman has joined #foot
<phin> naw no complaints.. just a bummer.
novakane has quit [Quit: WeeChat 3.2]
phin has left #foot [#foot]
<emersion> can always +1 that issue about xdg-decoration on their issue tracker
<emersion> sigh
<emersion> something something we don't care about non-gnome apps looking bad something
<lh> emersion: this is the only one I could find, that what you're thinking of? https://gitlab.gnome.org/GNOME/mutter/-/issues/217
tdeo has quit [Remote host closed the connection]
veiviser has joined #foot
veiviser has quit [Quit: veiviser]
veiviser has joined #foot
veiviser has quit [Quit: veiviser]
veiviser has joined #foot
veiviser has quit [Quit: zzz]
veiviser has joined #foot
veiviser has quit [Client Quit]
veiviser has joined #foot
ynakao has joined #foot
veiviser has quit [Quit: veiviser]
veiviser has joined #foot