Ireozar has quit [Read error: Connection reset by peer]
Ireozar has joined #river
flower_ has joined #river
eoli3n has quit [Remote host closed the connection]
eoli3n has joined #river
eoli3n has quit [Remote host closed the connection]
eoli3n has joined #river
Guest2 has joined #river
<Guest2>
im running fedora, and i installed river via dnf install, but i always get a blank cyan screen when running it
<Guest2>
there is no config file afaik, and im pretty new to using linux
<ifreund>
Guest2: that behavior is expected if you don't configure river at all
<ifreund>
configuration is documented in the river and riverctl man pages
<ifreund>
river will run an init script on startup, which you can use to invoke riverctl and setup keybinds, start programs, etc.
<Guest2>
fair point, with other DEs there was a bit more documentation to help beginners. sometimes even fedora specific
<Guest2>
can i just create a init script myself, or is that not advised?
<Guest2>
as in manually
<ifreund>
creating an init script yourself is exactly what is advised :)
<Guest2>
alright, ill try to look into it, thanks ~~
<ifreund>
and yes, I'm quite aware that river isn't terribly beginner friendly, it's also completely free to use :P
<Guest2>
i know i know, not complainin
<Guest2>
when it comes down to it i just want to choose the best DE for me, and hopefully that is river
<Guest2>
or, create it myself so to speak
<Nickli>
reminds me of first time starting OpenBox
<Guest2>
too bad there are not many beginner friendly step by step guides, was hoping that river was popular enough for those to occur naturally similarly to something like hyprland.
<Guest2>
i tried using sway with fedora, but i disliked the i3-style, and wanted something closer to hyprland. the last hyprland fork i used (by kool) didnt run well on my system
<Guest2>
so in other words, something more efficient and less needlessly flashy, while still being dynamic
<leon-p>
zuki: do you have experience writing wayland clients?
<leon-p>
if yes, then your first starting point is looking at the protocol xml
<leon-p>
Maybe I'll eventually get around to writing a simple example WM in C as I've done for the layout protocol, but not until rwm is merged to master
<zuki>
I haven't rewritten a wayland client before (unless hacking onto the layout.c included with the river repo counts.) and I've been reading a bit of the wayland book to try and understand things better.
<zuki>
I've managed to setup a basic makefile for using river-window-management-v1 with c.
<zuki>
Thanks for the tip to read the protocall, I assume you mean "river-window-management-v1.xml"
<gbrlsnchs>
https://wayland-book.com/ helped me a lot with getting started, I had no prior experience with Wayland client development
<sewn>
personally wayland-book explained nothing to me, reading the protocols manually using wayland.app and example clients explained much more
<zuki>
I'm currently struggling to get river-rwm to show anything other then a black screen, although that's probably because I'm trying to use a pull request version river-rwm that adds basic zig-14 support.
<zuki>
still black even with hyprpaper in init it shows nothing but a blackscreen and a cursor.
<zuki>
I'm using the normal rwm branch now with zig-13
notchoc has quit [Remote host closed the connection]
aryak has quit [Remote host closed the connection]
aktina has quit [Remote host closed the connection]
notchoc has joined #river
aryak has joined #river
aktina has joined #river
<leon-p>
zuki: rwm does not support layer-shell right now, so that's why you are not seeing a wallpaper
<zuki>
ah, I'm still not able to get any window to show up visually, ether via putting 'kitty &' into my river init or manually setting WAYLAND_DISPLAY. The proccess doesn't crash and acording to rwm's log its connecting just fine but doesn't show anything on screen.
<zuki>
I've also tried using rwm from both the tty and from within river-3.10
<leon-p>
in your WM try setting position and size of windows, that should make them appear
<zuki>
Do you need a wm to get rwm to show a window? If so that explains my issue since I wanted to check that rwm worked, before I got too deep into making my own wm.
<leon-p>
yes
<leon-p>
I can confirm it works, I've been hacking on a WM for some time now
<zuki>
thanks I feel a little silly now
<leon-p>
no worries :)
<ifreund>
zuki: yeah, currently the only real docs are the protocol xml
<ifreund>
it would be very helpful to know what parts of the protocol XML you find most confusing though as a relative newcomer to Wayland protocols
<ifreund>
(its completely useless currently, I'm still figuring out some event loop stuff and getting my repl workflow sorted)
<ifreund>
rivercompat on the rwm is probably the most complete low level example right now, though I suspect my janet stuff will pass it quickly when I find more time
d34db33f has joined #river
<zuki>
I'm having a issue were using wl_registry_bind to bind to river_window_manager_v1_interface causes a compositor crash of rwm
<ifreund>
zuki: messing around in rivercompat with foot windows works for me, what client are you using?
<ifreund>
kitty?
<ifreund>
hrm, kitty works for me too
<zuki>
wdym what client am i using?
<ifreund>
are you not running some program inside the compositor?
<zuki>
river-rwm with nothing connected crashes for me when running rivercompat or when my w.i.p wm tries to bind to iver_window_manager_v1_interface
<ifreund>
aside from rivercompat I mean
<zuki>
nothing aside from rivercompat
<ifreund>
oh, I bet your river init file is starting something
<ifreund>
can you reproduce with `river -c true`?
<ifreund>
there's definitely some xdg shell client connected to the compositor based on that stack trace
<ifreund>
(it's definitely a compositor bug too, if that wasn't clear. Some race condition I haven't hit yet myself)
<zuki>
ok so after trying it again it seems to work fine with rivercompat (screen turns white no crash) and with my w.i.p wm it doesn't crash and keeps the connection
<zuki>
AH if i run river-rwm and then connect kitty and then kill kitty and then try to run rivercompat the compositor crashes
<zuki>
that's really weird because running kitty->rivercompat works fine if kitty isn't killed before rivercompat is started
<ifreund>
interesting
<ifreund>
I think I have a decent idea of what's going on from reading the code, just a sec
<ifreund>
nevermind, I need to go make food before I dig in deeper
schneid3306 has joined #river
Cornelius-Figgle has joined #river
<ifreund>
zuki: I pushed a commit that I think should fix that issue
<ifreund>
I didn't manage to reproduce it myself yet though, so if you could confirm that it's fixed that would be great
<zuki>
ifreuned: seems fixed! unable to reproduce now.
notzmv has joined #river
<ifreund>
zuki: awesome, thanks for testing!
<ifreund>
this is why I love assertions
<ifreund>
they make sure that river's code actually works the way I think it does
<ifreund>
in other news, I figured out my janet event loop problem, I can now send wayland requests from the repl :)
<ifreund>
all single threaded cooperative multitasking
zuki has quit [Quit: Client closed]
schneid3306 has quit [Ping timeout: 252 seconds]
notzmv has quit [Ping timeout: 248 seconds]
flower_ has quit [Quit: Lost terminal]
zuki has joined #river
autisticshark has quit [Ping timeout: 260 seconds]
autisticshark has joined #river
schneid3306 has joined #river
donio has quit [Remote host closed the connection]