I think that’s fixed now. I hope. Maybe.
My ATI computer has some weird linux on it atm, so I can’t test it myself.
Yes, I think so, but no reflections?
it works on nvidia, and ati gives no errors… Not sure… what’s…
… wrong.
Ah, well. ATI sucks. 
PBuffer inited
Link warnings for /programs/water.vertex.shader + /programs/water.fragment.shader
Link successful. The GLSL vertex shader will run in hardware. The GLSL fragment shader will run in hardware.
edit: no reflections btw
That “warning” thing is outputted by wurm (I assumed that any output would be a warning, but I was wrong).
That “Link successful” is outputted by the ATI drivers. It’s either a bit fat lie, or there’s some weeeird bug in my shader.
[quote=“Markus_Persson,post:383,topic:15864”]
Yep, ATI’s GLSL implementation is… wrong. 
It has been a year since we moved Marathon to GLSL and not only it has never worked 100% bug-free on ATI cards, it actually seems to be getting worse with each Catalyst release. They’re supposed to have been working on it all this time, fixing bugs (I have personally reported ~6), but no apparent progress has been made (at least on what we need). The most critical bug (and our major showstopper) was fixed just last week, which means we’ll get it in an official driver in two months (minimum).
Patience, still.
Edit: Cool water btw
Ok, so there’s some confirmation. Thanks. =)
I’ll add an option for turning off GLSL in the meantime then.
It’s a shame that ati owners won’t get to see this:
http://www.mojang.com/screenshots/ultrawater.jpg
muoahahah!!!
WOW! :o I’m glad I dont use ATI ;D
sob That is just so, so, so beautiful
Notch, have you tried any of these two programs, http://developer.nvidia.com/object/nvemulate.html and http://developer.3dlabs.com/downloads/glslvalidate/index.htm ?
That water is just to good to be ruined by bad ati drivers sob
I get a 404 on the GLSLvalidate thing. =/
It seems like the most useful one of the two, as the other one just emulates nvidia cards.
I heard some rumors of the water looking “special” in the area where the sun would normally be reflected on ati cards, so that indicates something’s happening with the faked HDR rendering for the sun at least, but I haven’t gotten a screenshot of it yet to confirm.
I’ll keep trying to fix it. =)
If the problem was bad code, it would fail at compile time.
When writing a shader, both NVEmulate’s “Strict Shader Portability Warnings” and 3DLabs’ GLSL Validate are very helpful (if you’re developing on NV hardware). Unfortunately, ATI has a perfect GLSL compiler (IIRC, they used the open source one from 3DLabs…), but (what seems like) a very bad linker/optimizer. So, no matter how well-formed your shader is, it still has to pass through there.
That doesn’t mean I can’t try to find workarounds for it. =D
I was replying to AndersDahlberg. ;D
As for workarounds, if you make sure it’s a shader problem, try reordering the operations (without breaking the result ;)). It sometimes helps with ATI cards. NV cards are not affected, they always produce the same order of low-level ops, depending on the card profile.
The link works now.
Ah great, thanks. =)
And because I’m incapable of not posting a screenshot, I won’t not post one:
Really cool game man, i especially like the GUI
Launches ok, runs fine untill i click on play, then there is a black error window titled “wrong jogl version”, but I have the “july 12” release installed.
I have ran the game before, a month ago or so, whithout jogl localy installed (webstart downloaded the jogl components on the fly back then) and it ran great, but i’ve updated my video card and i would like to see how it does
How did you “install” jogl? NEVER copy files into the jre/jdk folders as it will cause conflicts. Delete any files you’ve copied into the jre folders and try again.
I did that to run art of illusion with opengl and to code with jogl without having to add the library and classpath properties on every launch, but i’ll look for some other solution, thanks.
btw, shouldn’t webstart override the installed jogl and load his own?