Trotterwatch
Active member
Thanks a lot, this works great 
and Happy Easter
and Happy Easter Needs Tester For An Experimental Glide Wrapper
- Thread starter Hacktarux
- Start date
olivieryuyu
New member
hmm can someone test Mario golf with a radeon 9600 with 5.3 ATI drivers ?
i have some white textures but i 'd like to know if it's only me that have this issue
Thanks
i have some white textures but i 'd like to know if it's only me that have this issue
Thanks
The Siskoo
Member
I have some white trees !!! And it's slow...
olivieryuyu
New member
here a ss of mario golf
many games have this problem
many games have this problem
The Siskoo
Member
Same here !!!
olivieryuyu
New member
another one
The Siskoo
Member
Same problem with Pokemon snap just before the start screen.
I have the official catalyst 5.3...
I have the official catalyst 5.3...
Trotterwatch
Active member
Hacktarux said:
I'm using a Radeon 8500 with 5.1 ATI drivers and am experiencing none of the graphics issues as shown above.
A quick question:
I don't think the Radeon 8500 supports the shading language does it? I have it selected, but I've a feeling it auto defaults to off if it can't be used.
olivieryuyu
New member
old drivers (4.10) give me same results
OP
- Thread Starter
- #274
Well, the code is very simple, i can't really run it through GLSL validator as the shaders are dynamically written by the wrapper but i know nvidia and ATI compilers doesn't report any error. Nvidia's compiler has an option that enables some portability warnings, it doesn't report anything about the code produced by the latest version of the wrapper.
I've seen errors in both compilers but at least errors on nvidia's compiler gave me compilation errors, not glitches...
I've seen errors in both compilers but at least errors on nvidia's compiler gave me compilation errors, not glitches...
Reznor007
New member
Hmm...maybe you could capture one of the shaders that is producing the bug on ATI cards and send it to ATI's developer relations. Or maybe post it on Beyond3d.com on the forums as a few ATI and nvidia employee's post on there.
One of the ATI employee's that posts there(Humus) makes small graphic demos using GLSL from time to time so he might be able to help find out where the problem is.
One of the ATI employee's that posts there(Humus) makes small graphic demos using GLSL from time to time so he might be able to help find out where the problem is.
OP
- Thread Starter
- #276
I've just checked the 21 shaders used in mario gold with the glsl validator. They are all valids. Now the question is... which one is causing problems...
Maybe it can be the shaders that are using fogs... let's check if google knows any problem that involves fog, glsl and ATI
Maybe it can be the shaders that are using fogs... let's check if google knows any problem that involves fog, glsl and ATI
Reznor007
New member
Found something while looking up some GLSL stuff that reminded me of the previous bug on ATI cards.
Your comment "it seems like ATI compiler don't know how to convert 0 to 0.0" is actually a good thing.
From NeHe's OpenGL GLSL introduction:
GLSL is 100% type safe. You are not allowed to assign an integer to a float without casting (by constructor):
float my_float = 1; // Won’t Work! “1” Is An Integer!
float my_new_float = 1.0; // Will Work!
From here: http://nehe.gamedev.net/data/articles/article.asp?article=21
Haven't found anything relating to fog bugs though.
Your comment "it seems like ATI compiler don't know how to convert 0 to 0.0" is actually a good thing.
From NeHe's OpenGL GLSL introduction:
GLSL is 100% type safe. You are not allowed to assign an integer to a float without casting (by constructor):
float my_float = 1; // Won’t Work! “1” Is An Integer!
float my_new_float = 1.0; // Will Work!
From here: http://nehe.gamedev.net/data/articles/article.asp?article=21
Haven't found anything relating to fog bugs though.
Noxious Ninja
New member
Reznor007 said:
Heh. The ATi drivers get a 98% on the 3Dlabs parser test, you say? The latest nVidia drivers, 76.41, score a 49%.
OP
- Thread Starter
- #280
Well, it's not that important, these tests are really weird. Anyway, the problem i'm having isn't about the parser, the shaders are compiled with both compilers, it means they are parsed without any errors. From the tests i did, it seems like the issue is on the fog, maybe a wrong fixed functionnality for the fog on ATI's hardware ?