ifreund changed the topic of #river to: river - a dynamic tiling wayland compositor || https://github.com/riverwm/river || channel logs: https://libera.irclog.whitequark.org/river/
<Shinyzenith[m]> ifreund: https://pastebin.com/GUPd5DNJ I have logs this time
<Shinyzenith[m]> Closed a twitch tab and the entire compositor just crashed
<Shinyzenith[m]> segfault
<Shinyzenith[m]> as you can see, it unmapped the view, rivertile comitted, and it crashed
<Shinyzenith[m]> possible bug in the layout protocol?
<tleydxdy[m]> cool someone was able to catch it. I also have similar crash with emacs I think
<leon-p> Shinyzenith[m]: would be even more useful if the executable had debug info
<Shinyzenith[m]> <leon-p> "Shinyzenith: would be even..." <- oh do you mean the -log level flag?
<Shinyzenith[m]> looks like I already launched it with -log level flag
<Shinyzenith[m]> version: 0.2.0-dev.74+7c440b8
<Shinyzenith[m]> <tleydxdy[m]> "cool someone was able to catch..." <- I've consistently reproduced it with chromium only
notzmv has quit [Ping timeout: 260 seconds]
waleee has quit [Ping timeout: 260 seconds]
snakedye has quit [Read error: Connection reset by peer]
snakedye has joined #river
notzmv has joined #river
snakedye has quit [Ping timeout: 240 seconds]
<mizzunet> I'd like to waylock to be ran after resumption. How do I do that?
<mizzunet> I tried riverctl map-switch normal lid open spawn waylock
<mizzunet> But didn't help
snakedye has joined #river
kitty1 has joined #river
kitty1 has quit [Ping timeout: 244 seconds]
kitty1 has joined #river
<NickH> mizzunet: I normally achieve this by having swaylock run before suspend. What method are you using to suspend?
lxsameer has joined #river
<NickH> Or if using swayidle you can add the "before-sleep" option (also requires systmed)
NickH has quit [Ping timeout: 260 seconds]
NickH has joined #river
<leon-p> Shinyzenith[m]: no, I mean a river debug build. Your logs aren't super helpful, because the debug info has been stripped, so we can't actually tell /where/ the bug is.
waleee has joined #river
ayushnix has joined #river
waleee has quit [Quit: WeeChat 3.5]
waleee has joined #river
<Shinyzenith[m]> <leon-p> "Shinyzenith: no, I mean a..." <- Ok how do I make a debug build?
<Shinyzenith[m]> small side note, experienced the crash again right now
<ifreund> Shinyzenith[m]: don't pass -Drelease-safe when building, and don't strip the binary
<ifreund> if you're using an arch PKGBUILD to build river, you may have to take extra care to make sure arch's tooling doesn't strip the binary, I don't really remember
snakedye has quit [Read error: Connection reset by peer]
snakedye has joined #river
<Shinyzenith[m]> I will look into it and get back to you
snakedye has quit [Read error: Connection reset by peer]
snakedye has joined #river
nullndvoid[m] has joined #river
<nullndvoid[m]> hey, I'm having issues with binding the Print Screen key to take screenshots
<nullndvoid[m]> command used:
<nullndvoid[m]> riverctl map normal None Print "grim -g $(slurp) -c - | wl-copy"
<nullndvoid[m]> I need a spawn
<nullndvoid[m]> riverctl map normal None Print spawn "grim -g $(slurp) -c - | wl-copy"
<nullndvoid[m]> also does not work :/
<andrea> ifreund: do you think it is a good idea to remove stripping from the aur package?
<ifreund> nullndvoid[m]: you'll want to use single quotes, not double quotes, otherwise the $(slurp) gets run before the string gets sent to river
<nullndvoid[m]> thanks :)
<ifreund> andrea: not sure, I'm by no means an expert on arch packaging. Was I remembering correctly that it strips by default?
<andrea> Yes, it strips by default
<andrea> River is very small and this would relieve you from asking to do it every time someone has a problem
<leon-p> If there are separate release and -git packages, I think the latter one should not be stripped
<andrea> I could also build in debug mode
<andrea> leon-p: agree
<andrea> I'll do it right now
<ifreund> I personally just run river in debug mode all the time, I don't notice any perforance difference on my relatively beefy desktop. I can't speak for older laptops though
<ifreund> The bottleneck should be the GPU not cpu though, and the stuff running on the GPU won't be affected at all by build river in debug mode
<ifreund> I think the actual issue here is that zig stack traces should be working in release safe builds with debug info but they aren't
<ifreund> I haven't looked into trying to fix that yet though
<novakane> I haven't noticed any perf difference too on my way less beefy laptop
<leon-p> I mean, the most intensive thing river should be doing on CPU is iterating over all buffers about 60ish times a second depending on screen refresh rate when you have a video playing or something like that. That isn't exactly much. (assuming the video is not a shm buffer, because then you'd have a few massive memcpy's)
<andrea> done
<nullndvoid[m]> How can I get information for a window by clicking on it? Is there a riverctl command to help me with this
<nullndvoid[m]> or some sort of API to ask the compositor for this info?
<nullndvoid[m]> * this info? e.g. wayland protocol
<leon-p> no, that is not possible right now. There is no way for a client to now if some window was clicked on
<nullndvoid[m]> leon-p: how about through the compositor/wayland?
<leon-p> sure, but that would require river to implement such a thing, which it doesn't
<leon-p> what window information are you interested in, anyway?
<leon-p> you can get some window metadata, you just can't select windows by pointer
<nullndvoid[m]> I think I am asking the wrong questions here apologies
<nullndvoid[m]> I just need to be able to hide the screen share window that Chromium produces
<leon-p> ah, I see
<nullndvoid[m]> and clicking the hide button does nothing for some reason
<leon-p> I can see your train of thought, on X you'd hack something together to get the winid and than hide it, but that won't fly on wayland
<leon-p> What will work is just focusing the window and sending it to an unused tag
<leon-p> if it is a focusable window
<leon-p> if not, then you can't really do anything right now, unfortunately
<nullndvoid[m]> it doesn't look like I can focus it atm
<nullndvoid[m]> nevermind, I think thats all good, I'll try looking for any Chromium bugs to do with this if that is the issue here
lxsameer has quit [Ping timeout: 246 seconds]
lxsameer has joined #river
<nullndvoid[m]> how do I get the app-id for a window?
<nullndvoid[m]> also why is piping grim to wl-copy for screenshots not working for some reason, it works on my shell is all I know
<leon-p> nullndvoid[m]: for app-id use this: https://git.sr.ht/~leon_plickat/lswt
<leon-p> and when putting shell-commands into your init, keep in mind that the init is a shell script that is executed when river starts. If you don't escape your commands correctly, they won't work
<nullndvoid[m]> leon-p: I quoted them, I'll just write a script to do screenshots
<nullndvoid[m]> leon-p: thanks :)
<nullndvoid[m]> does the pattern field of the filter commands take regex or wildcards?
<ifreund> no, it's planned but not yet implemented
lxsameer has quit [Ping timeout: 272 seconds]
talismanick has joined #river
snakedye has quit [Ping timeout: 244 seconds]
snakedye has joined #river
snakedye has quit [Ping timeout: 246 seconds]
snakedye has joined #river
snakedye has quit [Read error: Connection reset by peer]
snakedye has joined #river
snakedye has quit [Ping timeout: 244 seconds]
snakedye has joined #river
snakedye has quit [Ping timeout: 246 seconds]
snakedye has joined #river