The Future Direction & Purpose Of Wayland's Weston Is Being Revisited

[phoronix]

Phoronix: The Future Direction & Purpose Of Wayland's Weston Is Being Revisited

Pekka Paalanen, the Collabora developer who has effectively been serving as the Wayland/Weston release manager with Kristian H?gsberg still being away from Wayland development, is seeking some discussion from Wayland developers about their intended view of Wayland's Weston compositor and what direction it should take moving forward...

http://www.phoronix.com/vr.php?view=MTg1NzI

[89c51]

Haven't all the DEs around linux implemented their own compositors effectively making Weston just a test thing?


BTW the worrying part in Pekkas email is that the maintenance/review/etc of Wayland ends up going through one person only and with no official corporate support. Quite weird thinking how important wayland is. (but it probably already does what companies need (TV, mobile) i suppose)

[yakman2020]

Haven't all the DEs around linux implemented their own compositors effectively making Weston just a test thing?


BTW the worrying part in Pekkas email is that the maintenance/review/etc of Wayland ends up going through one person only and with no official corporate support. Quite weird thinking how important wayland is. (but it probably already does what companies need (TV, mobile) i suppose)

I can't help asking these days, do software companies write software anymore?

Seriously. It seems like the actual development thing is expected to be occuring elsewhere, perhaps in peoples spare time.

[89c51]

I can't help asking these days, do software companies write software anymore?

Seriously. It seems like the actual development thing is expected to be occuring elsewhere, perhaps in peoples spare time.

Even if you use something already made it is a hell of a job into turning it into something production ready for the general public. Wayland is just the protocol after all. I don't know if anyone is using weston as the compositor.

[mark45]

Wayland is just the protocol after all.

6+ years for "just the protocol" is way too much time spent on. Even Vista/Longhorn wasn't such a failure. I hope the corporations get their shit together and start investing time into Wayland to fully finish it and release production ready desktops (don't care about vehicles on wayland). It seems after they did 95% of the work they're slowly giving up.

[yakman2020]

But it seems to me there is a pattern of companies expecting the product to fall from the sky.

In terms of wayland, even weston, it seems to me there is considerable momentum/mindshare for doing it, and the time is pretty much ripe to do so. The product itself is or should be pretty straightforward for a small team (not a single individual) to manage. But everybody seems to be waiting for someone else to do it, so they can package and sell it.

[johnc]

Wayland, like BTRFS, is just around the corner.

[dh04000]

Wayland, like BTRFS, is just around the corner.

Oh god, I forgot that was even a thing...

Just like Wayland, we've been waiting since 2008 for it to be "about a year or two away".

Still waiting....

[Ericg]

6+ years for "just the protocol" is way too much time spent on. Even Vista/Longhorn wasn't such a failure. I hope the corporations get their shit together and start investing time into Wayland to fully finish it and release production ready desktops (don't care about vehicles on wayland). It seems after they did 95% of the work they're slowly giving up.

Core Wayland Protocol has been set in stone and pretty much 'done' for a long time. Everything past that, everything being worked on now are things that are 'nice to have' and even very usable but not absolutely required to get graphics on the screen. Such as the security extension, the relative pointer patches and other stuff. Things had to get set in stone, things had to be decided on. Gnome-- With Red Hat / Fedora backing-- is getting closer and closer every release to be able to seamlessly run under Wayland and KDE is working closer everyday. I think the current goal is KDE 5.4 but I could be wrong on that. Why would you think that they are giving up after 95% of the work? X -has to be- replaced, its age is showing in many places and they cant be fixed without breaking the X Spec or X applications. So the ONLY choice is a square one rewrite of the display server: Wayland. Its a lot of hard work and to get it takes a long time to get it done right.

[valeriodean]

It seems you are forgetting some facts:

1. The desktop related part of the protocol is ready to be merged in the Wayland protocol, should be part of the next version.
2. With the xdg-shell in place the Wayland part of the work is done. From that point on the work is all in the DE side.
3. With the libinput completion (the next version should merge the tablet support branch) everything is more than ready for the widely adoption
4. The last missing piece is the blob driver compatibility

So, except for the first point, there isn't so much work on the wayland side, then what are you speaking about?