2011 catalyst news:tearfree,weight loss,h264Flash acceleration

[Qaridarium]

2011 catalyst news:tearfree,weight loss,h264Flash acceleration

yes 2011 is funny time we got a light weight cytalyst because they put out the r300-r500 part and they put the xorg bin versions together.

and we got tear-free desktop and surprise surprise finally we got h264 codex gpu acceleration in flash.

and maybe we got an Adobe Creative Suite Linux version

http://www.golem.de/1012/80235.html

[Mr James]

I'll believe it when I see it.

[evolution]

2011 catalyst news:tearfree,weight loss,h264Flash acceleration

yes 2011 is funny time we got a light weight cytalyst because they put out the r300-r500 part and they put the xorg bin versions together.

and we got tear-free desktop and surprise surprise finally we got h264 codex gpu acceleration in flash.

and maybe we got an Adobe Creative Suite Linux version

http://www.golem.de/1012/80235.html

Well... the r300-500 part was already removed at Catalyst 9.3... Unless they didn't remove some old shared code in the driver...

About the xorg bin versions "together", that will mean they'll just support Xorg Server 1.9.x / 1.10 versions?

About the tear-free videos / H264 flash video acceleration: (btw, who needs flash with HTML5? ) Like Mr James said, I'll believe when I see it...

Cheers

p.s.: It would be nice they could fix the H264 video acceleration problem in UVD1 video cards, which is present since Catalyst 10.8 (at least)...

[bwat47]

2011 catalyst news:tearfree,weight loss,h264Flash acceleration

yes 2011 is funny time we got a light weight cytalyst because they put out the r300-r500 part and they put the xorg bin versions together.

and we got tear-free desktop and surprise surprise finally we got h264 codex gpu acceleration in flash.

and maybe we got an Adobe Creative Suite Linux version

http://www.golem.de/1012/80235.html

How are we getting h.264 acceleration in flash? Doesn't it only use vdpau?

p.s.: It would be nice they could fix the H264 video acceleration problem in UVD1 video cards, which is present since Catalyst 10.8 (at least)...

Yeah I would really like this to be fixed.

[Qaridarium]

Well... the r300-500 part was already removed at Catalyst 9.3.

you are wrong they hold the r300 parts because the windows version brings bugfixreleases for r300-r500 after that.

10.2 was the last windows r300 driver means they start to pull out r300 stuff after 10.2 but its a lot of work.

in 9.4 they only turn of the r300 part on linux.

[Qaridarium]

About the xorg bin versions "together", that will mean they'll just support Xorg Server 1.9.x / 1.10 versions?

all versions they just put duplicatet parts togeter.

in the past they use full version of every version number in the future they use 1 version and all other versions are generatet by difference part files.

[monraaf]

2011 catalyst news:tearfree,weight loss,h264Flash acceleration

yes 2011 is funny time we got a light weight cytalyst because they put out the r300-r500 part and they put the xorg bin versions together.

and we got tear-free desktop and surprise surprise finally we got h264 codex gpu acceleration in flash.

Have they fixed suspend/resume issue, Xv colorspace issue and high cpu load when using shader based yuv to rgb conversion with mplayer gl output issue? I think not. All those 'goodies' you are talking about are absolutely worthless to me if the driver can't even do a proper suspend/resume. Besides anything you'll get with 11.1 they'll take it away with 11.2. That's what catalyst has always been about: regressions.

[evolution]

Have they fixed suspend/resume issue, Xv colorspace issue and high cpu load when using shader based yuv to rgb conversion with mplayer gl output issue? I think not. All those 'goodies' you are talking about are absolutely worthless to me if the driver can't even do a proper suspend/resume. Besides anything you'll get with 11.1 they'll take it away with 11.2. That's what catalyst has always been about: regressions.

1. Suspend/Resume issues? That might be more a problem from your system configuration (hardware/software used) than something else. Personally I don't have any of that problems with my cards, and I believe most people there don't have problems with Suspend/Resume (Unless they're using ToI, I think).

2. Xv is old technology and uses CPU (the graphics card is only used for color conversions/scaling)... Long life to XVBA/VAAPI-based video acceleration!
About the high CPU usage with YUV-to-RGB conversion + GL rendering in mplayer, I don't have such issue (At least, for me, it's using about the same CPU percentage as Xv uses, (and as I said in the Xv topic, GL/GL2 rendering is only used for color conversions/scaling))... Btw, in Xine, I can confirm there's a real problem (but that might be more a developer's problem than a driver one).

Cheers

[Kirurgs]

PM has been always great with catalyst, my experience is just excellent...
I don't know whether to believe Mr. Q, it's not clear whether he speaks truth or not. If he is under NDA, he should already have been kicked in butt but he isn't, so it seems that more or less he's making things up...
Anyhow he's not bitching about everything is bad with catalyst, there is no offense from him, nice person I would like to say

[Qaridarium]

Have they fixed suspend/resume issue, Xv colorspace issue and high cpu load when using shader based yuv to rgb conversion with mplayer gl output issue? I think not. All those 'goodies' you are talking about are absolutely worthless to me if the driver can't even do a proper suspend/resume. Besides anything you'll get with 11.1 they'll take it away with 11.2. That's what catalyst has always been about: regressions.

yes they should care but i can not force them to care.

"Have they fixed suspend/resume issue"

i don't know.

in my point of view powersaving features are nice to have but not essential.

maybe essential for notebooks but i do not use this kind of hardware.

"Xv colorspace issue and high cpu load when using shader based yuv to rgb conversion with mplayer gl output issue?"

maybe there are a openCL output in the future.

but yes i do not know the status of your bugs.

" Besides anything you'll get with 11.1 they'll take it away with 11.2. That's what catalyst has always been about: regressions"

yes thats why i love the opensource driver radeon and the git-bisect function.

yes this is an real problem they just can not deliver an end-user friendly bisect option with an closed source driver.

if you care about regression testing the radeon driver is your favor.