considering that gnome just deprecated X11, i have a weird feeling that my life is about to get even harder and more painful, unless I choose to conform to one of the mainstream ideal linux GUIs

wayland will be the thing which finally kills all the small WMs that still remain. I consider that to be a cultural disaster.

@domi nah, I don't think it'll kill small compositors. The existing "meta" of X11 WMs will very likely die because it's a massive effort to detach a project like that from X11, but with libs like wlroots, swc or smithay the bar to clear is muuuch lower than reinventing the whole userspace display stack from scratch again.

If anything, X11 has to finally die for real instead of being kept on perpetual life support for those smaller WMs to kick off for real.

@dragoonaethis
@domi
The bar has been lowered by wlroots & co for almost a decade now. If we're not seeing an explosion of small compositors yet, that'd suggest the lowered bar is still too high.
@wolf480pl @dragoonaethis @domi we have sooo many compositors and many good/maintained compositors. mutter, kwin, cosmic-comp, niri, sway, swayfx, wayfire, labwc, wlmaker, phoc, cage, gamescope, pinnacle, jay, river-classic, river, etc. that's all without mentioning lots of abandoned ones and the fash one. If that's not an explosion I don't know what would be one??

@valpackett @wolf480pl @dragoonaethis i don’t think you’ve seen how many WMs X11 has? and real WMs, not glorified single-purpose hacks like cage and gamescope.

regardless of the quantity: let any of those go unmaintained for five years, and suddenly you have a codebase that doesn’t compile and requires a substantial amount of work to fix. meanwhile i’ve resurrected X11 WMs that were stale for 20+ years with a diff of 3 lines…

@domi @wolf480pl @valpackett it helps a little that X11 hasn't meaningfully changed for the past 20 years -_-
@valpackett @wolf480pl @dragoonaethis @domi and the new river alone has like 15 window managers built on top of its protocol, all of them created even before it was officially released a few days ago: https://codeberg.org/river/wiki/src/branch/main/pages/wm-list.md
wiki/pages/wm-list.md at main

wiki - Wiki for river

Codeberg.org
@IngaLovinde
which kinda proves the point that implementing a whole compositor is too much work, and there is a need for a way to implement a WM separately from a compositor
@valpackett @dragoonaethis @domi

@wolf480pl @IngaLovinde @valpackett @dragoonaethis to make my point clear:

  • it’s good that river is a thing now
  • it’s a shame that it took what, 16+ years until wayland got something like it?
  • it’s also a shame that I’m only hearing about it because of this thread

there are some conclusions that could be drawn from the 2nd point, but i’m not interested in exploring them in this thread

@domi @wolf480pl @IngaLovinde @dragoonaethis @valpackett
you might be interested in wayland.fyi (thought i sent my reply but i guess not)
Git Lab of experiments in varying quality.

@wolf480pl @dragoonaethis not looking forward to one of those libs getting deprecated X years from now :/