dnkl changed the topic of #foot to: Foot - fast, lightweight and minimalistic Wayland terminal emulator || 1.16.2 || https://codeberg.org/dnkl/foot || channel logs: https://libera.irclog.whitequark.org/foot
<gbrlsnchs> I love both Codeberg and SourceHut. Whichever is preferred by the project's creator works well for me. If I never had to open GitHub again and only had to switch between the former two, I'd be so happy :-D
erectus has quit [Ping timeout: 240 seconds]
<sewn> ifreund: that's understandable
xenrox has quit [Remote host closed the connection]
xenrox has joined #foot
uncomfy has joined #foot
sentriz has quit [Ping timeout: 276 seconds]
rrogalski has quit [Quit: Quit: Dog probably tripped over power cord again.]
rrogalski has joined #foot
sentriz has joined #foot
h-erectus has quit [Remote host closed the connection]
erectus has joined #foot
cbb has joined #foot
uncomfy has quit [Remote host closed the connection]
erectus has quit [Remote host closed the connection]
erectus has joined #foot
sentriz has quit [Ping timeout: 252 seconds]
sentriz has joined #foot
Arsen has quit [Quit: Quit.]
Arsen has joined #foot
Biolunar has quit [Ping timeout: 260 seconds]
Biolunar has joined #foot
rrogalski has quit [Ping timeout: 245 seconds]
cbb has quit [Quit: WeeChat 4.1.2]
cbb has joined #foot
anarcat has joined #foot
<anarcat> hello
<anarcat> i have just had a foot terminal closing a *bunch* of other ones because i had the misfortune of closing the parent of multiple terminals opened with control-shift-n
<anarcat> is that normal?
<cbb> anarcat: which version are you using?
<anarcat> 1.13.1
<cbb> anarcat: this is the first thing that comes to mind: https://codeberg.org/dnkl/foot/pulls/374#issuecomment-179430
<anarcat> uh!
<cbb> but that fix seems to have landed prior to 1.13.1
<anarcat> but that was fixed in 1.7.0
<anarcat> so i take it that's not expected :)
<cbb> I wouldn't expect it, no... but I can't reproduce it either
<anarcat> ack
<cbb> anarcat: can you reproduce it realiably?
<cbb> and do you get any errors in the logs?
<anarcat> i can't reproduce it if i start foot from an existing foot terminal
<anarcat> so there must be something with my launcher
<dnkl> I can't reproduce it, neither when starting foot from foot, nor when starting foot via fuzzel (i.e. a launcher)
<cbb> anarcat: if you run "foot" from a shell, the shell sets up the session/pgroup in a way that prevents that kind of thing happening
<cbb> even as it was previously, before that fix was in
<anarcat> right
<anarcat> so i start foot from fuzzel, but with a twist: the whole thing is launched from a systemd-run thing
<anarcat> it's quite convoluted
<anarcat> actually, no, i start it just with systemd-run --user --collect foot
<anarcat> so no fuzzel involved
<anarcat> that's from sway, bindsym $mod+Return exec systemd-run --user --collect foot
<anarcat> and i do that because otherwise podman takes over the Main PID of sway.service because it does service discovery signaling
<anarcat> and when a container exits, it crashes my entire session
<dnkl> guessing systemd-run just kills the entire process tree when its top-level process exits?
<anarcat> dnkl: yep, exactly
<anarcat> i think i would need some special config to keep that form happening, i'll look around
<anarcat> dnkl: does foot do sd notification?
<dnkl> I've been going back and forth, in my mind, whether we should double-fork ctrl+shift+n sub-terminals. We currently don't, so they all end up in the same process tree
<dnkl> explain sd notification?
<anarcat> probably irrelevant
<anarcat> i think i need ExitType=cgroup
<anarcat> the default is =main which kills children when the parent exits
<dnkl> we try hard to stay away from special casing e.g. systemd, or any other init system/service manager
<anarcat> sd notifications is a process by which a process signals to systemd (or whatever process manager) that its startup is complete
<anarcat> i don't think it's relevant here
<dnkl> foot --server has --print-pid, which can either write its PID to either a file, or to an already open FD. That's the only service startup notification mechanisms we support
<anarcat> yeah that's fine
<dnkl> ExitType sounds like it could work
<anarcat> and it does!
<anarcat> thanks for the help
<dnkl> \m/
<Arnavion> anarcat: I run all launcher-launched things through https://paste.rs/w0LKG which for foot will result in .../app.slice/app-foot.slice/app-foot-$unique.slice/
<dutchie> this reminds me that i suspect running foot.service/footclient in my user systemd instance is stopping the url opening from working be
<Arnavion> The --scope means the cgroup will wait for all processes in that cgroup
<dutchie> ...cause it can't connect to my firefox instance
<dutchie> (thanks dog for leaning on my keyboard)
<anarcat> Arnavion: why PartOf?
<Arnavion> Yes, you should do `systemctl --user import-environment DISPLAY WAYLAND_DISPLAY` or `dbus-update-activation-environment --systemd DISPLAY WAYLAND_DISPLAY` etc if your compositor doesn't do it already
<Arnavion> anarcat: So that they are all stopped automatically and "nicely" when I exit sway
<Arnavion> as opposed to having the socket closed on them and crashing
<anarcat> Arnavion: nice!
<anarcat> not sure i need the --slice stuff
<anarcat> nor systemd-cat
<anarcat> i guess it makes for cleaner logs
<Arnavion> Yes, like I said the --scope is the main one I was pointing out
<anarcat> what do you do --slice for?
<anarcat> or the command -v?
<Arnavion> Mainly because it looks nice in systemd-cgls, but also for resource-control stuff
<anarcat> avane: does that fancy script have a name?
<anarcat> i might just use it as such as well
<Arnavion> like steam and all games get put in app-steam.slice, so I have a drop-in for that slice to set its memory max
<Arnavion> `command -v` is to convert a command name to a full path, because the launcher I use (bemenu) only gives the command name
<anarcat> uh
<anarcat> and systemd-run isn't happy with th at?
sevz has quit [Quit: Client quit]
<Arnavion> I don't remember about systemd-run specifically, but systemd units themselves don't take $PATH into account so I just do it to be sure
<anarcat> systemd-run seems to do the right thing here anyway
<anarcat> Arnavion: i think i'll adopt a variation of your script! do you mind? how did you name the script?
chymera has joined #foot
cbb has quit [Quit: WeeChat 4.1.2]
cbb has joined #foot
<anarcat> Arnavion: ping?
hexa- has quit [Quit: WeeChat 4.1.1]
hexa- has joined #foot
Biolunar has quit [Ping timeout: 256 seconds]
Biolunar has joined #foot
cbb has quit [Quit: WeeChat 4.1.2]