Wayland's Weston 13.0 Released With Multi-Backend Support & Overlapping Outputs

[phoronix]

Phoronix: Wayland's Weston 13.0 Released With Multi-Backend Support & Overlapping Outputs

Weston 13.0 has been released as the latest major update to this reference Wayland compositor that attracts various experimental features and other innovations as developers experiment in the post-X11 world...

https://www.phoronix.com/news/Wayland-Weston-13.0

[ezst036]

I am pretty sure I have seen the answer to this before, but.

Why isn't Weston used in distros and DEs? Everybody keeps re-inventing the wheel? For a smaller desktop that may not have a lot of manpower, why not just build on top of/with Weston?

[primary]

I am pretty sure I have seen the answer to this before, but.

Why isn't Weston used in distros and DEs? Everybody keeps re-inventing the wheel? For a smaller desktop that may not have a lot of manpower, why not just build on top of/with Weston?

I've thinked this as well alot It feels like wlroots is most updated when it comes to new features.

[Sethox]

I am pretty sure I have seen the answer to this before, but.

Why isn't Weston used in distros and DEs? Everybody keeps re-inventing the wheel? For a smaller desktop that may not have a lot of manpower, why not just build on top of/with Weston?

You can use Weston as a test environment or as a small Desktop Environment (DE) for smaller systems like in cars, planes, trains, fridges, etc. The answer to your question was that when the other DE started to implement Wayland, the concept of having a test suite wasn't fulfilled (developed to this degree) but the Wayland concept was good enough to implement. Since the DE went ahead and implemented their own solutions (for better or worse), it was too late to implement Weston as a backend (it's still possible though, but you can imagine the time to refactor the current implementation of said DE).
There is much more to this than what I am saying but it's the gist of it as I understand it.

As it's "stupid" to reinvent the wheel for most, it's actually good experience and practice to do so, e.g. documentation gets better, code gets highlighted (good and bad) and perspectives clash that causes discussion. Which leads up to your question (what we have today).

[mrg666]

I am pretty sure I have seen the answer to this before, but.

Why isn't Weston used in distros and DEs? Everybody keeps re-inventing the wheel? For a smaller desktop that may not have a lot of manpower, why not just build on top of/with Weston?

So, you think all those different wheels in the market were "reinvented" all over just to be "manufactured"?

Or, do you "reinvent" FFT algorithm (or another) to merely "implement" it?

And, can you buy any wheel on the market and think it fits on your car? Plus, how do you know Weston is the best implementation possible?

[rmfx]

I am pretty sure I have seen the answer to this before, but.

Why isn't Weston used in distros and DEs? Everybody keeps re-inventing the wheel? For a smaller desktop that may not have a lot of manpower, why not just build on top of/with Weston?

To me the waste is more with the multiple distros rather than the multiple implementations competing to get the best result out of a given protocol.

If Linux Foundation was doing a kind of Arch-like unified distro with a (edge, stable, longterm branch to fit all the needs), Linux could really take off as a true unified effort.
Distros would still exist but they would be the same dev target with much much less redundancy, and be mostly a user experience/cosmetic preference for end users.

[Vistaus]

And, can you buy any wheel on the market and think it fits on your car? Plus, how do you know Weston is the best implementation possible?

Very good point. A lot of people keep rehashing that phrase, but wheels are not all alike either.

[M.Bahr]

Michael
FYI The third part of winewayland has been merged.: https://gitlab.winehq.org/wine/wine/..._requests/4522