Announcement
Collapse
No announcement yet.
The Performance Penalty Of Xfce/Xubuntu On XMir
Collapse
X
-
Originally posted by MrTheSoulz View Postim not in love with either mir or wayland, im a neutral in here trying to learn.
you dont see me flaming either wayland or mir and i wont at least untill they are both finished and can actuly be compared to each other, comparing mir or btw wayland to x at this point is a waste of time since non of them are finished products.
if not more is testing mir and giving a change for cpu vendors to get theyr drivers ready and get feedback from people who can actuly say something good and arent just writing for war fun.
On testing, there are only a few Mir things you can test through XMir, and they're noticeable enough to get polished in an alpha release of whatever version ends up using a native Mir desktop. To name the obvious ones, you can test modesetting, XMir itself (I think that's the only non trivial to test), page flipping, eventually, when it's ready, compositing bypassing, and a subset of what is input handling (specifically, identifying the input, but not which surface is it going to, since you only see a fullscreen X server which in turn will handle the input Mir hands to it).
Also, I think it's wrong to make forced testing. The people who tests alphas and betas do it willingly, knowing it's likely to not be for production use, and they do it because of own reasons (I do as a way to collaborate, some people just like to be more bleeding edge while using Ubuntu); since this will go to 13.10 and 14.04, it's a different story.
Comment
-
Originally posted by LinuxGamer View Postyes it is
looks like ill be unstalling Xubuntu on my wife's ComputerLast edited by Nuc!eoN; 06 August 2013, 04:51 PM.
Comment
-
Originally posted by LinuxGamer View Postyes it is
looks like ill be unstalling Xubuntu on my wife's Computer
Keep your shirt on, sparky!
Comment
-
Originally posted by Andrecorreia View Postwho cares about finnish? lol
No country is better than any other, it's just a matter of chance where you happen to be born.
Comment
-
Originally posted by mrugiero View PostFair enough, we are the same kind.
AFAIK, Wayland is a finished product (as finished as software can be), just there are mesa and kernel changes not already there, and some things to polish on compositors. But Wayland itself is already complete, AFAIK.
On the time, they could have the time they need just by Ubuntu using X.org until Mir and the drivers are ready, so that's not a good reason either.
On testing, there are only a few Mir things you can test through XMir, and they're noticeable enough to get polished in an alpha release of whatever version ends up using a native Mir desktop. To name the obvious ones, you can test modesetting, XMir itself (I think that's the only non trivial to test), page flipping, eventually, when it's ready, compositing bypassing, and a subset of what is input handling (specifically, identifying the input, but not which surface is it going to, since you only see a fullscreen X server which in turn will handle the input Mir hands to it).
Also, I think it's wrong to make forced testing. The people who tests alphas and betas do it willingly, knowing it's likely to not be for production use, and they do it because of own reasons (I do as a way to collaborate, some people just like to be more bleeding edge while using Ubuntu); since this will go to 13.10 and 14.04, it's a different story.
in the comsumer eyes its not finished untill its running some kind of DE (my opinion).
the kind of benchmark i would want to see is weston running a full DE vs xmir.
"On the time, they could have the time they need just by Ubuntu using X.org until Mir and the drivers are ready, so that's not a good reason either."
making a DE that only runs with EGL (wayland/mir) is forcing them, what canonical is doing is better, they offer both so they can ease into it (not forcing) wich means wayland is the one forcing it :P
"On testing, there are only a few Mir things you can test through XMir, and they're noticeable enough to get polished in an alpha release of whatever version ends up using a native Mir desktop. To name the obvious ones, you can test modesetting, XMir itself (I think that's the only non trivial to test), page flipping, eventually, when it's ready, compositing bypassing, and a subset of what is input handling (specifically, identifying the input, but not which surface is it going to, since you only see a fullscreen X server which in turn will handle the input Mir hands to it)."
testing the drivers is the moust importante thing i belive, if the drivers are ready they can then easly make it run with unity 8 without messing with the drivers at all.
"Also, I think it's wrong to make forced testing. The people who tests alphas and betas do it willingly, knowing it's likely to not be for production use, and they do it because of own reasons (I do as a way to collaborate, some people just like to be more bleeding edge while using Ubuntu); since this will go to 13.10 and 14.04, it's a different story."
i dont consider anything but the LTS stable and in my opinion no one should any release outside lts is bleesing edge wich means its for testing to be then put into the next lts.
they could do like windows did have a BIG waiting time untill everything is ready and then on launch day find out whats wrong in the real world.
Comment
Comment