3D Nature and Partners Declare ATI drivers Unsuitable for Professional Visualization

[Kaotik]

XenonofArcticus, just wondering, what's up with using Quality texture settings rather than High Quality, as seen here:

 

[CATALYST MAKER]

hotfix has been released

http://support.ati.com/ics/support/default.asp?deptID=894&task=knowledge&questionID=22263

 

[rwolf]

XenonofArcticus, do you have any other bugs you need fixed?

 

[mjtdevries]

Does this mean that it actually helps to create such a press release?

 

[Geo]

http://support.ati.com/ics/support/default.asp?deptID=894&task=knowledge&questionID=22263

Knowledge Base
No Results Found - 22263

Yumpin yimminy --a hotfix even? Not even just rolling them into the next monthly Cat? I foresee a plethora of News Releases in an attempt to get your attention in the future.

"Minneapolis, MN FOR IMMEDIATE RELEASE --Geo announced today. . . "

Seriously, nice turnaround there when you did jump on it.

But since you're here anyway. . .and wouldn't feel needed if the sound of griping wasn't in your ears . . . why is it when I navigate thru customer care to the WinXP Radeon page that Cat 5.5 is what greets my eyes until I hit "more" and go to a second page? Wouldn't it be better for that stuff to be sorted most recent first?

Edit: Btw, when signed in I got the "results not found" from the included link. As soon as I signed out from ATI's site I was able to see the hotfix description.

 

[micron]

XenonofArcticus, just wondering, what's up with using Quality texture settings rather than High Quality, as seen here:

On a mobility chip running a modern app, high quality just might not be feasible

 

[wishiknew]

While I like to see drivers of the gamer line be up to stuff running professional applications, I can understand Atis point of view of allocating resources toward opengl.

At least its fixed and ID engine games work.

Does the latest catalyst drivers actually incorporate code from the so called opengl rewrite yet? Could the world run out of petro this happens?

 

[Geo]

The last time I saw Terry address it in public, he said:

We are not re-writing our OpenGL driver from scratch. What we are doing is re-writing certain portions of it to improve stability and performance. There will by no means be massive performance gains after this work is complete. This effort will attempt to improve the quality and speed of our OGL driver as much as possible, but until it is done (which still is a ways away) we cannot speculate on how much of a performance improvement users will see.

http://www.driverheaven.net/dhinterviews/terryandrew/

Otoh, it seems to me I've heard suggestions/hints that they might be slipstreaming bits in, rather than a "grand release". Tho it would certainly be nice to know if they have a target date for "work is complete" and "until it is done".

 

[wishiknew]

I remember reading an interview with a firegl guy a couple months back and something along the lines of they weren't going to switch over to the 'rewrite' branch until it was stable.

One day.

 

[Geo]

I remember reading an interview with a firegl guy a couple months back and something along the lines of they weren't going to switch over to the 'rewrite' branch until it was stable.

One day.

Yeah? Hunt it down and link it please?

 

[karlotta]

Yeah? Hunt it down and link it please?

well as a gamer ATi's OGL is fine, i only miss the Open Source versions of NVDA proprietary extensions... Oxymoron drivers..

 

[wishiknew]

Yeah? Hunt it down and link it please?

I tried but I just couldn't find it. I remember it wasn't in one of the big name sites though.

 

[Rys]

I wrote something along those lines, after talking with Dinesh Sharma, here. Was it that?

 

[Geo]

I wrote something along those lines, after talking with Dinesh Sharma, here. Was it that?

Ah, thank you. That would seem to leave open the possibility that they may be slip-streaming it into Catalyst. . .but not necessarily. It could just as easily be read to mean there will be a grand release to Catalyst. . . followed at a later date by a grand release to FireGL after consumers have. . .err, thoroughly "beta tested". . .it. :smile:

 

[wishiknew]

I wrote something along those lines, after talking with Dinesh Sharma, here. Was it that?

Hi Rys. I count Hexus a big name site.

I remember it was more of a workstation/cad site. But close enough.

EDIT. ... maybe this is it. God old age already caught up to me.

 

[XenonofArcticus]

More from Xen...

@Kaotik: Yeah, at the time I made that screengrab I was using lower quality textures. Sometimes it's a trade off, skimping on quality can get you more resources available on-card, and sometimes that's worth it. This laptop machine uses the mobility variant and only has 128Mb of memory, so you do what you can to stretch it. I think "Quality" makes it use 16-bit textures instead of 24-bit ("High Quality") and that can stretch your texture memory budget further on a limited machine.

I am happy to see a hotfix, and no, all the bugs we found were documented on that OpenGL Compliance page last year. We haven't found anything further to report. If we do, we'll report them to ATI developer relations just like we (initially) did this time. Only if it's a nasty unworkable problem that gets ignored would we go to these lengths.

We will be confirming this fix as soon as we can actually test it. I can't install this driver on my own (only) ATI test hardware, but some other guys in the OSG developer group are going to test it and let us know how it fares. Unfortunately, the hotfix doesn't do most of our customers any good -- in most big organizations (US Gov't especially) you aren't allowed to install anything other than a fully released, supported, signed, official driver. So those folks are stuck until it becomes part of a real release. And so am I, because Omega won't release a new update until ATI makes it official. Such is life.

As far as the issues of Linux drivers from ATI, laptop drivers from ATI and NVidia -- someone else can champion those crusades. I've got a day job to do, and the issues that were holding us up are resolved.

Am I going to crow about how ATI fixed this so quickly? No. They can do that if they care enough. In my book, it still took 6 months and a lot of hair-loss , so that's not "quickly" in my mind. It was only quick after they started feeling the publicity pain.

Do I recommend this approach for people with other driver problems? No. It's a last resort, and it's pretty bitter for all parties. I'd have been a lot happier if it had been resolved months ago during any of the many communications I had with ATI devrel.

 

[Geo]

Heh. Except the damage was already done by the press release. At that point they could have safely ignored it for another six months!

I dunno who your devrel contacts are, but in the future you might try to hunt Terry Makedon down yourself and beard him. This side of digitalwanderer, he's about the easiest guy to find on the 'net. . .

I'm not saying as step 1 (for which Terry would pummel me with his dayplanner if I said anything of the kind). But if step 1 is "Pardon me, ATI . . .but 'ow!'" and step 10 is "I shall now smite you from one end of the 'net to the other with a list of OGL living legends!", then maybe around step 5 or 6 you should try hunting CATALYSTMAKER down directly. Just my $.02.

 

[IgnorancePersonified]

I'm just glad there's a happy ending... I like happy endings.

 

[wishiknew]

! http://www.sharkyextreme.com/features/idtrInterviews/article.php/3606746

I'm probably going to live long enough to see it in production!!!

 

[Geo]

Cool, thanks. I remember when SE was a must-visit site (geo fondly remembers).

We are planning to release some new OpenGL code later this year that will be the result of this effort.