LXQt Now Has Full Qt5 Support

[curaga]

(short edit times)

That's because "knowledge" by definition includes what the video detector does. If an app has a GL-rendered part it truly cannot handle in a better way, it has the knowledge of which pixels changed.

[erendorn]

(short edit times)

That's because "knowledge" by definition includes what the video detector does. If an app has a GL-rendered part it truly cannot handle in a better way, it has the knowledge of which pixels changed.

Well, X applications, a real world example of "knowledge apps", suck for network use of anything animated or GPU rendered
I'd say that Citrix's ICA fares a bit better than X in this regard (for handling multimedia), but looks quite complex and not that incredibly better either.

While the video compression core in you GPU does a pretty good job of compressing anything from animations to multimedia, without much effort for anybody, including programmers. I guess it fares already pretty well for compressing translations (scrolling), but given the wide use of this kind of damage on large surfaces, you can just simply add a "translate subsurface" to the "update subsurface" of your protocol and build on that (yes, this is a bit on the knowledge side), without adding much complexity to your display server (possibly even smoothing things locally).


I fully agree that the theoretical quality upper bound of knowledge app is higher than a compressed output delta app (but not by much: the theoretical limit of compression is pretty high too).
But in the real world, a constant programming cost, video networking works better.

[GreatEmerald]

So the actual topic lasted for a whole four posts. That must be some sort of a record...

[tuubi]

So the actual topic lasted for a whole four posts. That must be some sort of a record...

Hey, this is Phoronix. We've seen worse.

[curaga]

Indeed, the record is 0 posts.

[Luke_Wolf]

So the actual topic lasted for a whole four posts. That must be some sort of a record...

Quick! Call Guinness!

[pouar]

I think I just thought of an even bigger reason why Wayland won't succeed X. It only works with the Linux kernel.

[Luke_Wolf]

I think I just thought of an even bigger reason why Wayland won't succeed X. It only works with the Linux kernel.

Wrong http://www.phoronix.com/scan.php?pag...tem&px=MTMwMzE
Weston is written to the Linux API however Wayland itself is AFAIK OS-agnostic, since all it cares about is buffers.

[ua=42]

So the actual topic lasted for a whole four posts. That must be some sort of a record...

It's like clockwork. You even breathe the word wayland in an article and someone will say "It will fail because it doesn't have network transparency". I'm pretty most of them have to be trolling at this point as it has been mentioned allot these past several years that X also hasn't been network transparent for a long time either.

[Luke_Wolf]

It's like clockwork. You even breathe the word wayland in an article and someone will say "It will fail because it doesn't have network transparency". I'm pretty most of them have to be trolling at this point as it has been mentioned allot these past several years that X also hasn't been network transparent for a long time either.

this +1.
It's basically the "If it ain't broke don't fix it" crowd who as always fail to take a vastly more important (read: Actually useful) principle into account: "Don't live with broken windows", and X is one very ugly very broken window.