Maybe someone should make a minimal by default rootless network transparent #X12 able to stream gfx and audio?

Take as a given that on the server/display side something else will do the window management, composing, displaying, passing data from window to window, etc…

Hell, perhaps use something like display postscript to do it more efficient? :D

@puppygirlhornypost @stefano As far as I know (from what I read, from other contributors, from lots of users e.g. on the forums), #wayland works fine on #FreeBSD (as in, "production quality" as far as this is possible talking about wayland at all). DRM/KMS isn't really an issue, #FreeBSD uses the original drivers from #Linux (which are liberally licensed, not GPL ...). Sure, they require Linux-specific kernel interfaces, but there's #LinuxKPI in FreeBSD bridging that.

Still, I don't like the idea of wayland. It tosses out more or less everything, shifting it to either the compositor or the client application. I never tried it though, mostly because I'm perfectly fine with #X11. I even started programming for "raw" X11 very recently, which gives much more insight than just using it, and what I found basically boils down to: Yes, X11 core drawing is useless nowadays, but #XRender is pretty sane and it's a shame its development stalled ... IMHO, the way forward shouldn't be wayland, but something like #X12, tossing out the stuff that's *really* useless nowadays and instead declaring some extensions (XRender, XKB and a few others) mandatory ...

@winterschon Thanks a lot for the offer, I'll most likely get back to that, but it might take a while (it feels like I'm 80% there, but OTOH, a lot of building blocks are still missing, like e.g. menus and tooltips, have to see how to handle them correctly, as they must be "transient" windows without decorations and with specific behavior ...)

Regarding #wayland, my feeling is quite similar: Sure there are issues with #X11, but wayland didn't really convince me it's "the solution". I always prefer non-disruptive, evolutionary improvements, something that's in my experience prevalent in #BSD systems (while the #Linux ecosystem loves to reinvent the wheel and break everything every other day, to exaggerate a bit 🙈)

Maybe an #X12 could indeed be the better solution. Take it with a grain of salt, as I really have no "inside" experience with X, but "drop all stuff nobody uses any more and at the same time declare *some* extensions mandatory" would look pretty great to me ...

[1] Please don't use 123456789012 as a receiver ID. Just why?! What kind of megacorp can't afford to designate an [arbitrary] test ID that doesn't suck?

[2] If you're manually editing an ISA segment, [a] why?! & [b] mind the heccin' field width, ya dingus . . .

<w<

#edi #lame #x12