The no future of X.org, FreeBSD becomes headless?

Considering the amount of work being done, I would not assume this.
I was referring to Marcan/Hector being behind the vtuber. The Linux port itself is obviously a much larger team.

It wasn't about libinput, it was about an X protocol limitation after libinput.
Does that not sound odd to you? A protocol that could handle vastly more complex keyboards (i.e the Sun Type 5 keyboard) can't handle a fairly basic consumer keyboard. The same one used perfectly fine for the Intel mac models. The guy is absolutely blagging it to push his agenda.
 
From what I read (and linked to earlier) …

So that people don't have to click more than a hundred times to load comments:

<https://mega.nz/file/Yd8G3CbK#2SY_CfMWYiO6KjQkAESlQlSbQmHV68tn6UqTESfmo3c>
  • Think twice about Wayland. It breaks everything! also known as Think twice before abandoning Xorg. Wayland breaks everything!
  • 9feb7c20257af5dd915e3a9f2d1f2277.pdf
  • A4, 1701 pages, 371.7 MiB
  • includes an edition (version unknown) of <https://gist.github.com/probonopd/9...malink_comment_id=4811191#gistcomment-4811191> (2023-12-29) from Stephan Sokolow
  • no fonts (sorry), maybe a limitation of printing from qutebrowser, at a later date I might attempt a more accessible PDF.
 
So that people don't have to click more than a hundred times to load comments:

I read the piece. I found it quite convincing despite my ignorance of the underlying, but you can call me a luddite.

I read a few of the early, accessible comments. Why would anybody want to read any more, I wonder.

A4, 1701 pages, 371.7 MiB

Surely you jest? What value could that much commentary on a relatively tiny article possibly contribute to, well, anything? Who could have the spare time to read it?

Can you outline, without more links, why or how all that could matter to this thread?
 
I was referring to Marcan/Hector being behind the vtuber. The Linux port itself is obviously a much larger team.
I'm referring to the vtuber being part of that team. Lina does GPU stuff. Hector doesn't seem to touch that. Don't really see any evidence that Lina is Hector's sockpuppet.

Does that not sound odd to you? A protocol that could handle vastly more complex keyboards (i.e the Sun Type 5 keyboard) can't handle a fairly basic consumer keyboard. The same one used perfectly fine for the Intel mac models. The guy is absolutely blagging it to push his agenda.
No, 8 bits for keys from the dawn of X11 doesn't sound odd at all.

Here's some of the posts after having to manually search:

Tangentially related:
 
No, 8 bits for keys from the dawn of X11 doesn't sound odd at all.

Here's some of the posts after having to manually search:
Hmm, I have come across these in the past. It still isn't convincing; for example this part of it:

Right now that driver is a mess of hacks and tables to quirk around keyboard layout differences and implement the Fn+F-key combinations for either F-keys themselves or machine-specific media keys.

That is patently the wrong way to do all this. All that is keyboard layout stuff, and should be handled in xkb. The Fn key should be exposed as a normal key. It already is modulo the magic mappings, except it doesn't work on Xorg because *insert decades-old bad decisions*. But we don't care about Xorg.

... everyone else manages to work around it quite well and there are much more "exotic" platforms than Apple's consumer offerings. This guy is just looking for excuses because Xorg is "old" and this guy sees himself as a bit trendy.

The OpenBSD guys seem to manage just fine for example.
 
The point is why should there be a workaround. Why can't it just handle larger scancodes as they come off the keyboard?
 
I would like to say that this whole thread was invaluable to a guy like me. I do multimedia, audio, animation, CAD, and remote desktopping work on freebsd, mac, linux, android, ios, arduino, pi, and m1cr050ft. There are some good articles in the thread that expedited my understanding of the future for wayland, pipewire, oss, gnome, kde and other mutlimedia desktopy things.

I saw someone suggest to close this thread because of politics talk. I'm glad they did not close this thread. I've been trying to decide what to do about making applications on freebsd and linux since its 2024. I always wonder what the current devs of the GPU, audio, and other input peripherals were going to be doing for 2024-2025. I have the mailing lists, but they're bloated with emails from all areas. But this thread alone, pointed me in the right direction to so many things.

I have alot on my plate and this thread had a very condensed amount of info. Don't close it if you don't have to. Keep these talks and links open. I was literally trying to figure out how to draw opengl/vulkan apps to wayland and xorg. Good thread.

Thanks.

P.S. I use freebsd in some very edge case situations in multimedia for rendering video and audio while using zfs^^ufs as the filesystems.
 
The point is why should there be a workaround. Why can't it just handle larger scancodes as they come off the keyboard?
Because that would break backwards compatibility of a small part of Xorg.

... instead they want to break not only backwards compatibility with the whole display system, but the entire software ecosystem. 👍
 
Because that would break backwards compatibility of a small part of Xorg.

... instead they want to break not only backwards compatibility with the whole display system, but the entire software ecosystem. 👍
Conversely, Xorg's refusal to create X12 or whatever's necessary to natively support larger scancodes is probably an example as to why momentum coalesced around something completely different.

Look, I would prefer X just get updated so I don't have to learn a new thing, but we should not fool ourselves about what's going on.
 
Experimental (found amongst <https://www.freshports.org/ports-new.php?interval=month>):
… fork of xscreensaver which makes it possible to run some of the hacks as animated wallpapers on Wayland compositors that support wlr-layer-shell. With swaylock-plugin, it is possible to use these as backgrounds for a lock screen on some Wayland compositors as well. Note: this is a very rough work in progress, …
 
We all have individual use cases. For me, I use openbox or dwm, depending on various things. I gave labwc and dwl a thorough test run, in case xorg disappears before I die. (not that likely, I'm pretty old). I could do everything that *I* need to do. I don't care about screensavers, I don't remote to other machines with X, I do, on occasion, input Japanese. At work, it's all ssh into terminal, so doesn't matter at all. So, for me, I see no real difference if X is completely replaced by Wayland. Again, this is just *my* use case, no one else's. I also suspect X will be around, regardless of what RH does, for a long time. Fedora, which is a bit of a testbed for RH among other things, is not dropping X yet, and I doubt it will be a worry for anyone for a long time.
Somewhat ironically to me, FreeBSD's dwl worked better than Fedora's. Fedora wouldn't run urxvt on dwl, though it does on labwc. Also, the only terrminal out of a few that I tried in Fedora, that would input Japanese was xfce4-terminal--foot and alacritty wouldn't, and as mentioned, it wouldn't run urvt.
There are probably reasons for that, but at present. Wayland pretty much works as a drop in replacement for me. I don't plan to switch to it, because I see no reason to, but if I had to, my life would be pretty much the same.
 
Last edited:
X.Org on NetBSD - the state of things (May 4, 2024)

NetBSD has its own downstream ecosystem of X.org in the xsrc repository which functions as a downstream fork. Customizations are kept separate from regularly being upstreamed to X.org.

This X.org implementation is built exclusively using NetBSD's toolchain. It uses
BSD makefiles, not X.Org's based on GNU autotools.

NetBSD has developed and improved a handful of video drivers, which many are still only available on BSD.

xf-86-input-ws is exclusive to development on NetBSD and OpenBSD. It's a generic input driver which:
doesn't assume the device is a mouse, and can support advanced touchpad and touchscreen features.
Other NetBSD developed drivers include:
xf86-video-pnozz, xf86-video-mgx, and xf86-video-crime.

Also, NetBSD updated drivers to work with the newer EXA driver, which were abandoned in Xorg and elsewhere. XAA was the older 2D driver for Xorg.

A lot of these drivers were developed or otherwise contributed by

NetBSD also focuses on clients or programs which use Xorg.


This and Xenocara need to be a collaborative project of a common fork shared between FreeBSD, NetBSD OpenBSD and other BSD's.
 
Yes, but:

NetBSD don't push upstream based on their fork
OpenBSD might/do

At this point, fork or not, they should be considered as separate entities.
 
This and Xenocara need to be...
A lot of people say that, in both Linux and BSD camps. Who's got the skill and motivation to make it a reality, though? My thinking goes, it will start with a 'proof of concept' - someone will set up 3 machines (or VMs), a FreeBSD, a NetBSD, and an OpenBSD - and set up a CI infrastructure to develop and test that fork, and maintain it... If you think it through, this fantasy of a project is a LOT of work to set up, let alone get right. 😩
 
NetBSD don't push upstream based on their fork
OpenBSD might/do

At this point, fork or not, they should be considered as separate entities.
It's kind of the point, that they are split from Linux. Upstream is so different than what suits BSD. It's the effort to upstream improvements, which suit more purposes including for Linux, which one BSD operating system does and another doesn't. Let them be in maintenance or slow mode for Linux to use like that. Then have a BSD independent fork, where each BSD can differ from that.

A lot of people say that, in both Linux and BSD camps. Who's got the skill and motivation to make it a reality, though? My thinking goes, it will start with a 'proof of concept' - someone will set up 3 machines (or VMs), a FreeBSD, a NetBSD, and an OpenBSD - and set up a CI infrastructure to develop and test that fork, and maintain it... If you think it through, this fantasy of a project is a LOT of work to set up, let alone get right. 😩
It really is a lot. It would take collaboration between the foundations to different extents. At minimum, they would have to agree on a fork, which can simply be what Xorg used to be. Alternatively, a fork could be independently run opting not to install the traditional Xorg. OpenBSD and NetBSD are already collaborating on some parts. Then, each one would have to be tested on the machines, and there's a lot of ports. So both Xorg versions would have to be available side by side on FreeBSD, until one is depreciated. The newer Python still hasn't replaced the older one, and the newest xaw3dxft hasn't replaced the older xaw programs for example,

But, if it were to happen, collaboration would occur with less effort, and with any effort from any of these operating systems. Driver news and collaboration would spread faster. That's for another venue though.
 
It's kind of the point, that they are split from Linux. Upstream is so different than what suits BSD. It's the effort to upstream improvements, which suit more purposes including for Linux, which one BSD operating system does and another doesn't. Let them be in maintenance or slow
mode for Linux to use like that. Then have a BSD independent fork, where each BSD can differ from that.
I don't think it has anything to do with Linux. So far, the forks of xorg in NetBSD and OpenBSD that I've noted are more to do with integrating them to the buildsystem ecosystem. There's few, impactful deviations to standard x.org
 
That's the hard part, it needs to be automatic and reliable, to the same extent as what Xorg provides.
Has anyone written a Wayland script? I'm sure this isn't an unreasonable expectation that could easily be accomplished with a script. I am by the way not interested in writing the script.
 
That's the hard part, it needs to be automatic and reliable, to the same extent as what Xorg provides.
I have default configuration in wayland as display server, and some changes in window management things.
But when most of applications is X11 applications, why I should use Wayland?
 
Back
Top