XWayland 2D Performance Appears Better Than XMir

[spacetoilet]

GNOME uses GPL+ and LGPL_ while XFCE uses tri-licences of GPL+, LPGL+ and BSD.
LGPL3+ would be fine for display manager. The main issue is Canonical behaviour towards upstream considering that Mir display manager is a clone of Wayland.
It is not the first Canonical acted that way, Unity was originally a copy of early Gnome Shell.

eazy now your going to make the Usheep mad telling them how Mir was Developed so fast and we do have Pic's of the older Gnome's

[Delgarde]

It is not the first Canonical acted that way, Unity was originally a copy of early Gnome Shell.

Not really... the other way around, if anything. In terms of design, Unity evolved from Ubuntu's netbook edition - a bunch of Gnome 2 hacks to optimise things for a small screen, etc - which was then subsumed into the regular desktop, with stuff like the modal view for application launch and search and stuff. And if you look at early Shell designs, there's actually less similarity than later on - e.g Shell had a more complex sidebar, which was then replaced with a dock similar to that of Unity. Personally, I think Shell did it better (which is why I now use Fedora rather than Ubuntu), but they *did* copy a lot of ideas from what Unity had already done.

[finalzone]

I don't see your point. The post you quote explicitly states Canonical are *the only ones* who can sublicense. Comparing with dual or even triple licensing is mixing apples and oranges: in those cases, you still have the same rights than everyone else, and in Canonical's they have *more* rights than everyone else. And this extra rights come with extra control.

I deliberately omitted the Clause Licensing Agreement part to show that LGPL3+ is a better license for display manager rather than GPL3+ and specify the used licenses from respective Desktop environments i.e. Gnome and XFCE.

[mrugiero]

Not really... the other way around, if anything. In terms of design, Unity evolved from Ubuntu's netbook edition - a bunch of Gnome 2 hacks to optimise things for a small screen, etc - which was then subsumed into the regular desktop, with stuff like the modal view for application launch and search and stuff. And if you look at early Shell designs, there's actually less similarity than later on - e.g Shell had a more complex sidebar, which was then replaced with a dock similar to that of Unity. Personally, I think Shell did it better (which is why I now use Fedora rather than Ubuntu), but they *did* copy a lot of ideas from what Unity had already done.

http://en.wikipedia.org/wiki/Unity_%...environment%29
Canonical starts their Unity idea in November 2010.

http://en.wikipedia.org/wiki/GNOME_Shell
GNOME Shell became available for the first time in August 2009.

[Sidicas]

Mir's superiority and awesomeness is secured at the protocol level:

So no matter what any benchmarks show - Mir is better than Wayland, period.

Assuming that it doesn't expose bugs in the Android drivers they plan to be using, in *ANY* of the devices they plan to target, you're right.. If it does expose bugs, then they're SOL because the hardware manufs. aren't on the Mir boat, probably won't be anytime soon, and won't care to fix driver bugs related to Mir since they're focused on Android.

If Canonical runs into bugs that aren't fixable because they're hooking into binary blob drivers from Android, things could go downhill for them very fast. Whereas Wayland would be in a much better position to get similar problems fixed in Mesa, although as has been noted before, fixing bugs in Mesa related to Wayland isn't a cakewalk either.

[mrugiero]

Assuming that it doesn't expose bugs in the Android drivers they plan to be using, in *ANY* of the devices they plan to target, you're right.. If it does expose bugs, then they're SOL because the hardware manufs. aren't on the Mir boat, probably won't be anytime soon, and won't care to fix driver bugs related to Mir since they're focused on Android.

If Canonical runs into bugs that aren't fixable because they're hooking into binary blob drivers from Android, things could go downhill for them very fast. Whereas Wayland would be in a much better position to get similar problems fixed in Mesa, although as has been noted before, fixing bugs in Mesa related to Wayland isn't a cakewalk either.

Wayland uses the same. libhybris was designed for Wayland to take advantage of Android drivers.

[Sidicas]

http://en.wikipedia.org/wiki/Unity_%...environment%29
Canonical starts their Unity idea in November 2010.

http://en.wikipedia.org/wiki/GNOME_Shell
GNOME Shell became available for the first time in August 2009.

Ugh, Not this again..

....and in 2010 Canonical's CEO gave a long speech about how Unity was built from everything they learned from Ubuntu Netbook Remix which dates back before 2008.. Going on about how it all culminates into the initial design of Unity and how netbook manufacturers like Asus have a great interest in using Ubuntu's Netbook Remix for the future netbook EeePC sales.

[mrugiero]

Ugh, Not this again..

....and in 2010 Canonical's CEO gave a long speech about how Unity was built from everything they learned from Ubuntu Netbook Remix which dates back before 2008.. Going on about how it all culminates into the initial design of Unity and how netbook manufacturers like Asus have a great interest in using Ubuntu's Netbook Remix for the future netbook EeePC sales.

OK. I don't care about GNOME Shell or Unity anyway. The only thing I remember from those times is when the menu got integrated into the taskbar, but I think MacOS X was the one to come with that idea.

[kgunn]

how can standalone Xorg be better for any case than Xorg+Wayland

The claims of the article that this is an apples2apples comparison of what the Xmir configuration is would appear to be inaccurate. The test results linked to in this article seem to indicate that XWayland (Xorg+Wayland) are actually better in 5 out of 9 cases vs standalone Xorg. It does not seem reasonable that Xorg would be slower than Xorg+"anything".

Xmir is Xorg+mir as a system compositor. I am tempted to believe as one other poster has pointed out that in these instance Xorg is not involved, and rather what is being measured is a native Wayland rendering.

[mrugiero]

The claims of the article that this is an apples2apples comparison of what the Xmir configuration is would appear to be inaccurate. The test results linked to in this article seem to indicate that XWayland (Xorg+Wayland) are actually better in 5 out of 9 cases vs standalone Xorg. It does not seem reasonable that Xorg would be slower than Xorg+"anything".

Xmir is Xorg+mir as a system compositor. I am tempted to believe as one other poster has pointed out that in these instance Xorg is not involved, and rather what is being measured is a native Wayland rendering.

This was discussed one thousand times. When you run single apps on XWayland or XMir, you get to ignore several parts of what X alone would do. For instance, compositing will be handled by Mir/Wayland, which are supposedly more efficient than X on that area, and X will work as if there is no compositing. A full X desktop + anything, yes, will probably be the same or slower than pure X, but neither XMir nor XWayland are designed for that, but rather for running single apps (the fact Canonical will run desktops on top of XMir is just a way for them to do early in real world testing of Mir, nothing else). For single apps you will get, in terms of cost for the processing, pure X.org costs minus whatever you achieve to ignore of X (input grabbing/releasing and compositing are the most obvious ones) plus whatever it costs to do the ignored operations in the Wayland/Mir side. If Mir or Wayland is more efficient on this areas than X, then you will get a better performance than running the same set of applications on a pure X.org session.