03:15 <tleydxdy[m]> I mean the protocol is "unstable" so

10:02 <upsala> I'm having a weird issue with a clipboard tool called copyq. I have that autostart with 'riverctl spawn copyq' in my river init file and have Super+F4 bound to 'copyq toggle' to show/hide the main window. After starting river, the shortcut is only able to show the window but not hide it. When I close copyq and start it again, either via riverctl

10:02 <upsala> spawn or with rofi or via console everything is working as expected

10:02 <upsala> Any ideas what this could be?

10:07 <ifreund> upsala: that smells like a client bug tbh, the best way to debug would be to set WAYLAND_DEBUG=1 in the environment where copyq is started and redirect the stderr output to a log file

10:08 <upsala> Ok I'll try that thanks, I think I encountered the same thing when using Hyprland a couple of months ago, so you might be right that it's not river specific

11:17 <p00f1> <ifreund> "river doesn't have any similar..." <- i guess it is necessary in this case because i use kmonad which changes keymap (unless keymap does not mean what i think it means)

11:17 <p00f1> because the whole thing is inside a `if (keymap_changed || repeat_info_changed || config->reloading) {`

11:50 <Gozenka> https://github.com/riverwm/river/issues/24

11:50 <Gozenka> Is this forsaken? :)

11:50 <Gozenka> Is there another way to prevent header bars from some gtk apps they appear on?

11:52 <ifreund> Gozenka: I personally use this trivial gtk patch https://0x0.st/oTaC.diff

11:53 <ifreund> the gtk MR is forsaken as I didn't have the time or energy to learn enough about gtk internals to properly apply the state at runtime

11:53 <ifreund> the gdk part of the MR should be fine if anyone ever wants to pick it up

11:54 <Gozenka> (y)

14:26 <bw> is there somewhere i can see what's been added to river in the last year or so?

14:26 <bw> just the major stuff, if anything

14:27 <ifreund> bw: these release notes cover most of the features: https://github.com/riverwm/river/releases/tag/v0.2.0

14:28 <ifreund> there are some unreleased changes on master branch that are tracked here: https://github.com/riverwm/river/issues/796

14:35 <bw> noice

15:45 <bw> ifreund: that mentions "Interactive resize of windows with the cursor from all edges/corners." but i don't see how to do that

15:46 <ifreund> bw: it's the same resize pointer mapping, just grab the window close to the edge/corner you want

15:47 <ifreund> if the client sends river a resize request with an edge/corner set that works too

15:47 <bw> i guess let me make sure i'm actually building the correct river version first

15:47 <ifreund> yeah, that's pretty recent on river master, not in 0.2

16:31 <Guest7460> I apologize for my lack of wayland knowledge in advance: why is it that the 'focused_view' and 'view_tags' events are received in separate dispatches?

16:35 <angry_vincent> focused is an active "view" if more than one views are active

16:38 <ifreund> Guest7460: just to explain what's happening, river sends focused view immediately when keyboard focus changes and keyboard events are sent to a different client

16:38 <ifreund> this is all asynchronous though and that doesn't mean that that client is already being rendered

16:39 <ifreund> view_tags is only sent when the state river is rendering has been updated

16:39 <ifreund> to make every frame perfect, river needs to wait before updating the rendered state in many cases for multiple clients to ack configures and submit new buffers of the required size

16:44 <ifreund> really if we wanted frame perfection river should also wait for the river-status client to ack the new state and submit a buffer before rendering a new frame

16:44 <ifreund> probably do that next time I break things in that area

19:20 <Guest7417> ifreund: thank you for the explanation

19:21 <Guest7417> Would you consider adding an enter-previous-mode command to riverctl?

23:41 <ifreund> I guess I'm on mastodon now if any of you use that: https://hachyderm.io/@ifreund