Incidentally, tiling WM:s are even more retro, even if they're all the rage right now. GNU Screen and Oberon are both from 1987.
anyway, just a data point after happily seeing your comment.
I keep wanting the impossible. Something minimalist like i3, FVWM etc. but comprehensive, so I can go without what kept feeling like all the hassle of having to write custom utilities for so many operations and spending a lot of time having to configure the damn thing.
I've been running a whole band (16in, 8out) with Ardour, 3x Guitarix and "standard effects" (EQ, Compression, Barry's Satan Maximizer, etc) for some years now on a 10th gen i7.
That's with -rt patches and running live.
I considered switching to Xfwm, but even with Gnome3 I never had issues with performance.
With IceWM, you need to take care of the SSH agent that KDE or Cinnamon would start for you. You can do it with something like this in your `.xinitrc`:
# SSH agent
mkdir -p "$XDG_RUNTIME_DIR"/ssh
eval "$(ssh-agent -a "$XDG_RUNTIME_DIR"/ssh/agent)"
export SSH_AUTH_SOCK
# ...
# WM
exec icewmWell, for most of us it is.
X11 is not actively developed, a fork is.
That's a weird framing.
Original X was forked as XFree86 and included in all the major distributions.
Xorg was a fork of XFree86, and subsequently included in all the major distributions when some devs/sponsors wanted to take it in a different direction.
XLibre — if that's what you are referring to — is another fork in a long line of forks.
So when you say X11 is not being developed but a fork is, which one are you referring to?
X11 and Xorg, as we know it, are not in active development but maintenance mode. They have planned to wind down for a while, and this has been old news for over six years now... The code rollback and freeze inquiries should've signaled people off.
Within the Linux community, the primary X11 implementation has been the XFree86 server, which forked into XOrg over licensing disputes, which forked into XLibre over the development roadmap.
Other implementations include XQuartz (also based on XOrg) and Xlibe (not a traditional server but implements X11 by translating into native Haiku graphics API calls). There's also Xming and android-xserver (written in Java!), among many others.
XOrg may be abandoned. That doesn't mean the entire ecosystem is.
Xorg is not actively developed. X11 is fine.
Red Hat and Canonical made their calls, and basically every other desktop environment outside of their control as well. Some were late, but are now here. Outside of legacy systems, I just don't see why people are going out of their way to get back on X11/Xorg outside of perceived FOSS politics and posturing. See - XLibre code of conduct and the Arch fiasco.
Thankfully, that's not really how open source ecosystems work. There are still people interested in both using and developing X11.
> Outside of legacy systems, I just don't see why people are going out of their way to get back on X11/Xorg outside of perceived FOSS politics and posturing.
Well, when Wayland supports the features that people use on X11 I suppose we can discuss it. As of the present day, moving "forward" to Wayland would mean massive regressions in a11y tools that I'm not willing to give up just yet, and accepting a fragmented landscape where questions like "how do I change the keyboard layout" or even "is it possible to change the keyboard layout to anything but us+qwerty" are "It Depends™, what compositor do you use" (because Wayland deeply believes in reinventing the wheel over and over again, forever).
Being a protocol doesn't mean it can't be actively developed (look at HTTP at least), but if it's good it doesn't NEED to