Possible fix for the Vista/ATI disaster

[Zebulon Pleasure Beast]

(This is all based on wild googling that resulted in 6 months old posts, I am unsure of the current status of the ATI OpenGL driver but one thing is sure, it doesn't work properly on Vista)

So, my amateur research has brought me to the conclusion that the problem might be caused by ATI's Vista drivers lacking OpenGL ICD (whatever this means) which means OpenGL on an ATI card is somehow layered ontop of Direct3d and thus works like **** and crashes all the time.

Now, on one forum I picked this up:

"If you really want opengl in vista and don’t care or don’t have support for AERO interface, you can just install the xp driver and go on as normal. You just won’t get the 3D effects or transparency (thats the best part of vista!)"

Apparently it might be somehow possible to force Vista to run on older drivers, but I'm unsure on how exactly to go about doing this and whether CMSF would run is a mystery.

Just thought I'd give a heads-up since apparently the Vista/ATI connection is being neglected and dualcore processors are being blamed. Ofcos I might be wrong in which case shoot me.

I have a fresh Vista install on a single processor with the newest Catalysts and the exact same problem that is being reported by Vista/ATI users all over the forum manifests itself. I haven't yet seen a single person with Vista&ATI card reporting that CMSF works properly on his machine, dual core or not.

Correct me if I am wrong, just trying to help.

[rune]

Beast,

I have been doing the same researching as you. If you have an older card, can you try this driver?

http://www.ngohq.com/home.php?page=Files&go=cat&dwn_cat_id=9&go=giveme&dwn_id=399

If you don't, don't worry, I asked in the other thread to see if things are better. Again atioglxx.dll is their driver. Anyone have any connections at ATI where I can talk to them?

Rune

[Zebulon Pleasure Beast]

I have an X800, which is old, but is it old enough?

Anyway, downloading the driver...it does say it's for XP/2000 though.

[rune]

Sorry may have gone to the wrong one, there is a Vista one on the web site. Hang in there a sec. Nope. OK let me find a modified Vista driver. I see on their forums they are saying Vista and ATI still have problems, but hang loose. Thanks for trying tho Beast.

Rune

[rune]

Oh and if someone is having the problem with XP and ATI, can you give those drivers a go.

Rune

[Zebulon Pleasure Beast]

Computer won't boot Vista with those drivers installed and you have to remove them in safe mode. The search continues...

[Zebulon Pleasure Beast]

Has anyone tried the Catalyst 7.8 leaked beta?

[MikoyanPT]

Originally posted by Zebulon Pleasure Beast:

Has anyone tried the Catalyst 7.8 leaked beta?

No, but i heard about. It it seems dangerous, i am not as good with software to risk it ! :eek:

[Wburn]

I've had the game for two nights, and I'm just starting to wonder if there is a possiblity of a refund? I think if you're going to do a release you would have tested the game under your own reccomended specs! Currently the game is completely unplayable and after paying 55 dollars I'm getting more than alittle frustrated at the incompetency of those who released this piece of crap and thei inability of releasing a patch to resolve the ATI/Vista issue.

Wayne

[ASHBERY76]

I'm curious to know if anybody with ATI cards and Vista have the game running.I have no dual core so that is not the problem.I'm glad to have waited for the demo.

[ July 29, 2007, 10:51 AM: Message edited by: ASHBERY76 ]

[The DesertFox]

Originally posted by ASHBERY76:

I'm curious to know if anybody with ATI cards and Vista have the game runing.

Highly unlikely from what can be read on the board. Maybe they should label it NVIDIA only.

I'm glad to have waited for the demo. [/QB]

Me too. The demo ist totally arsed on my Notebook with VISTA Turion and ATI card which runs CM:AK smooth as silk. Crash to desktop happens 100% of time. Very dissapointing. If I get similar problems on my desktop PC (C2D and NVIDIA) I´m done with this bugger.

[Steiner14]

Wburn,

i understand your frustration, but do you know what the problem is?

ATI's OpenGL implementation is crap! But ATI sells their GPU with OpenGL compatibility.

So i'd suggest you write ATI, why their drivers do not support the OpenGL API like promised.

You are long enough registered here to know, that in CMx1 the ATI cards also had a huge problem: they couldn't display fog.

Don't blame Battelfront for the faults of another company, which is already famous for their crap drivers. And if you as customer accept that, things will become worse and worse: such companies will care less and less about correct implementations as long as their customers do not make them responsible, but the victims of their crap drivers: the game developers.

[rune]

Guys, at least the ones that expect miracles, I want you to call ATI right now, their tech support, and see who is working on a weekend. Tell them you have a problem with a game, and you will give them under 48 hours to recode their driver. Listen to their response, [well will have to wait till Monday, they don't work weekends, but you don't care, then think about it when you say it is CMSF. Here is a list of KNOWN bugs in the catalyst 7.7 drivers:

Known Issues Under the Windows Vista Operating System

This section provides information on known issues associated with the ATI Catalyst™. These include:

The default audio device may be set to ATI HD Audio after installation of the display driver for the ATI Radeon™ HD2900, HD2600, and HD2400. This issue may also be experienced under the Windows XP operating system. Further details can be found in topic number 737-28483

3DMark06: An error message may be displayed when running the application for an extended period of time (6-8 hours or more). Further details can be found in topic number 737-28494

City of Heroes: Attempting to play the game on a system with a dual link panel connected to it may result in the Windows Vista operating system failing to respond. Further details can be found in topic number 737-28495

Ghost Recon Advanced Warfighter: Pressing the Alt+Tab keys when playing the game may result in the Windows Vista operating system failing to respond. Further details can be found in topic number 737-28491

Lost Planet: Setting AA to either 4x or 8x may result in the game failing to respond. Further details can be found in topic number 737-28492

Resuming from an S3 state may result in a longer than expected time for the Windows Vista operating system to resume. Further details can be found in topic number 737-25798

Hot-plugging a CRT to a system running Windows Vista and containing an ATI Radeon™ X1100/1150 series of product may result in extended desktop mode failing to be applied. Further details can be found in topic number 737-25801. This issue is resolved in Catalyst™ 7.8

Enabling CrossFire™ followed by disconnecting and then reconnecting the CRT display device may result in duplicate video options being seen within the Catalyst™ Control Center. Further details can be found in topic number 737-27457. This issue is resolved in Catalyst™ 7.8

Catalyst™ Control Center->OverDrive™: The OverDrive™ setting are not retained for individual accounts. This issue may also be experienced under the Windows Vista operating system. Further details can be found in topic number 737-25805

Enabling CrossFire™ followed by setting AA to 14x and AF to 16x may result in the AA showing a setting of 0x level when disabling CrossFire™ and unchecking the AA option in the 3D aspect. Further details can be found in topic number 737-27458. This issue is resolved in Catalyst™ 7.8

Making changes to the Color aspect found in the Catalyst™ Control Center without applying the changes may result in the Reactivate ATI color controls button remaining active after the Catalyst™ Control Center is closed and then re-launched. Further details can be found in topic number 737-27459. This issue is resolved in Catalyst™ 7.8

The Record function currently fails to work when using an ATI All-In-Wonder X1300 and time shifting is enabled. Further information can be found in topic number 737-22014

A blank screen may be displayed after unchecking all the Component Video formats except when the 480i30 custom mode is applied. Further details can be found in topic number 737-27620

Enabling CrossFire™ followed by setting the OverDrive™ clocks to their maximum settings may result in the Catalyst™ Control Center failing to accurately report the requested OverDrive™ values. Further details can be found in topic number 737-27621. This issue is resolved in Catalyst™ 7.8

A green or black screen may be displayed when configuring MCE to ATSC mode. Further details can be found in topic number 737-27622

Setting the Avivo Video De-interlacing setting to auto, motion adaptive or vector adaptive may result in certain DVDs, menu videos and some special features failing to benefit from the options. Further details can be found in topic number 737-27196

The Enable ATI OverDrive™ clocks for 3d applications may be available on non supported products such as the ATI Radeon™ HD 2600 series of product. This issue may also be experienced under the Windows XP operating system. Further details can be found in topic number 737-28477

Performing an un-install of the display drivers on a CrossFire™ configured system may result in extended desktop mode becoming active resulting in all of the desktop icons failing to appear. Further details can be found in topic number 737-28478

Catalyst™ Control Center->3D->Anti-Aliasing: Unchecking the option to Use Application Setting found in the 3D aspect page when CrossFire™ is enabled may result in the maximum AA setting being 6x. This issue may also be experienced under the Windows XP operating system. Further details can be found in topic number 737-28479

Performing either a custom or express install of the display driver may result in the system failing to respond and the installation of the driver failing. This issue is known to occur when using an ATI Radeon™ X1950 CrossFire™ Edition product installed on in a system containing an Intel975 and nForce590. Further details can be found in topic number 737-28481

The 3D aspect may appear blank when launching a second instance of the Catalyst™ Control Center™. This issue may also be experienced under the Windows XP operating system. Further details can be found in topic number 737-28482

HDMI audio stutter may be heard when playing MPEG clip at certain resolutions on specific HDMI device. Further details can be found in topic number 737-28484

The color depth may change to 8bpp when enabling or disabling CrossFire™ for the first time. Further details can be found in topic number 737-28485

Connecting a TV followed by removing the display resolution of 640x480 from the mode table may result in the both the TV and LCD failing to display an image. Further details can be found in topic number 737-28490

The Catalyst™ Control Center version number may fail to be updated after the installation of a newer Catalyst™ Control Center. Further details can be found in topic number 737-28493

Dragging player window to secondary display while playing back HD content and moving it back to primary may result in blank player window. Further details can be found in topic number 737-28496

Windows Media Player: Playing a DVD in full screen mode may result in vertical bar corruption being noticed during playback. Further details can be found in topic number 737-28497

I am trying to narrow down what it is to see if a work around is possible, over MY weekend. I'll let you know if I find anything.

Rune

[The DesertFox]

Rune,

do you have a link for this 7.8 beta ? I can give it a spin and let you know how it turns out on my Notebook.

[ASHBERY76]

Originally posted by Steiner14:

Wburn,

Don't blame Battelfront for the faults of another company

Funny, ATI have never been the reason for any other game I have ever got not working.The small dog should make sure they are compatable with the big dog,not the other way round.

[Steiner14]

Originally posted by ASHBERY76:

</font><blockquote>quote:</font><hr />Originally posted by Steiner14:

Wburn,

Don't blame Battelfront for the faults of another company

Funny, ATI have never been the reason for any other game I have ever got not working.The small dog should make sure they are compatable with the big dog,not the other way round. </font>

[Wburn]

Let me give you a clue as to what the recommended specs stated.. I'm not making this up it's in plain view for all to see when they are deciding to purchase the game.

Operating System: Windows XP (Vista is supported)

Processor: Pentium IV 3.2 GHz or equivalent speed AMD processor or better

Video Card: GeForce 6800 or Radeon x850

(256 Megabyte VRAM or better and must support 1024x768 or higher resolution)

Sound Card: DirectX 9 compatible Sound Card

System Memory: 1 Gigabyte or more RAM

Hard Drive Space: 1 Gigabyte

Other Requirements: CD Drive (not needed for download version)

It plainly states ATI 850x as a standardized video card with Vista support. Where I come from that means the game should work with that hardware configuration. Why then is it ATI's fault or microsoft's vista OS?

Wayne

[MikoyanPT]

Originally posted by The DesertFox:

Rune,

do you have a link for this 7.8 beta ? I can give it a spin and let you know how it turns out on my Notebook.

Hi desrt fox, a couple of people have tried this. By rune sugestion but it doesn´t change the left click crash.

Have faith guys, Battlefront always delivers, and the game was released for 2 days only.

[tarball]

Originally posted by ASHBERY76:

I'm curious to know if anybody with ATI cards and Vista have the game running.I have no dual core so that is not the problem.I'm glad to have waited for the demo.

Yes, this Vista/ATI OpenGL issue is being conflated with the Multi-core issue and they are not likely related.

While I realize that nVidia is all the rage now, a significant number of computer users have ATI hardware as well. That same ratio is likely present in the userbase for CMSF - this entire demographic is SCREWED right now.

I am with the guy requesting a refund... it is unfortunate that this problem exists and I believe the developer(s) try - but I find it incredible that this wasn't picked up in testing. I know Vista is 6 months old, but there are plenty of folks out there who have this Vista/ATI combo - none of these folks are able to play the game at the moment.

Just for giggles, I tried this OpenGL benchmarking program - http://www.fl-tw.com/opengl/GeomBench/

I do not crash using it.

[ July 30, 2007, 06:01 AM: Message edited by: tarball ]

[The DesertFox]

Originally posted by tarball:

Just for giggles, I tried this OpenGL benchmarking program - http://www.fl-tw.com/opengl/GeomBench/

I do not crash using it.

Works flawless on my machine also. CMAK also works flawless btw. Seems like CMSF 1.01 has some serious graphic bugs hidden in its code which are in desperate need to be fixed ASAP unless they want to screw the ATI/VISTA users completely.

[Wburn]

"You are long enough registered here to know, that in CMx1 the ATI cards also had a huge problem: they couldn't display fog.

Don't blame Battelfront for the faults of another company, which is already famous for their crap drivers. And if you as customer accept that, things will become worse and worse: such companies will care less and less about correct implementations as long as their customers do not make them responsible, but the victims of their crap drivers: the game developers."

I don't dissagree and I remember the CM-O release and the outrage of no fog. But. The game mechanics worked without any freezeups and CTDS. The only problem was you didn't have smoke. Trying different methods to try to get CM-SF to run I've had no success.

1. Reverse mouse click (i'm left handed anyhow) still froze.

2. Reduced various commands in the catylst control panel, still no go.

3. Changed aperature and speed in the bios. No good.

4. kicked computer, worked great, computer blinked off and frustration of getting CM-SF to start becomes a moot point.

Wayne

[J Ruddy]

So... if the game works on Windows XP with my ATI x200 integrated POS, and the DEMO runs on Windows 2000 with my ATI x700 (haven't tried the game yet), but crashes on Vista / ATI combination, would this logically point to a problem with the game, with ATI hardware, with ATI Vista drivers, or with Windows Vista?

And for those that don't understand OpenGL - it is a programming interface, so that you can use the same code to speak to any number of video cards. In the old days (very old days) programmers had to write drivers for each video card - but with the introduction of standards for video interfaces (Directx and OpenGL) programmers, in theory, could write one piece of code that would work with all compatible video cards.

If the game code works for ATI on XP & 2000, and works for nVidia on XP, 2000 & Vista, is the problem with the game code? I imagine it *may* be possible to debug the problem with ATI/vista to the extent of "we make this call to OpenGL and it crashes" but I don't know if a work around could be implemented by BFC to get around the problem in the OpenGL layer.

[Battlefront.com]

All I can say is it is impossible for us to have tested all couple thousands of different possible variations of OSes, drivers, cards, processors, subsystems, user settings, patches, etc. Obviously it would have been helpful if one of our testsers had the Vista/newer ATI combo, but none happened to have that combo and therefore this problem went undetected. Sucks for everybody, but the bottomline is that if ATI's support for OpenGL was what it is supposed to be, we wouldn't be having this discussion. The bug itself is ATI's fault. The fact that we didn't catch it is our responsibility. But if ATI hadn't fudged it, we wouldn't have had anything to have missed.

A fix is coming out soon. There is a way to hack around it. It's just a matter of finding the right way to do it without causing other problems, especially for those that currently aren't experiencing any problems with this issue at all. That's why hacking around driver bugs is NOT the best way to go.

Steve

[Steiner14]

Steve,

may i ask out of interest, what the problem in more detail is? Is it a normal OpenGL function-call Charles uses, but this function-call leads to the crash?