Fast JAVA2D/JOGL bridge >= JAVA 1.6 not working with NVIDIA drivers

.

Which version of JOGL do you use?

.

That’s not accelerated OpenGL.

Cas :slight_smile:

.

princec is right, “Renderer is GDI Generic from Microsoft” means that Windows is using a generic driver rather than the real OpenGL driver of your graphics chip. GLJPanel requires PBuffer or PBO support whereas “GDI Generic” supports only a few features. This problem has nothing to do with JOGL, it is a problem of OpenGL configuration on your machine.

.

The current drivers for your 9600 GT, win 64, are 260.99. Here.

:edit:

Ah, just read your first post - you’ve already tried them =]

I just gave the example a spin on my Dad’s GTX 260 (win7 64bit), and all I get is a white screen =/ No crash, but nothing is rendered either.

It failed in the same fashion on my Radeon 5870 (win7 64bit) the first time I launched it, but it worked fine on subsequent launches.

.

The Netbeans OpenGL Pack uses an obsolete version of JOGL and you should use the option noddraw=true.

.

Does it mean that your problem has been solved?

The properties sun.java2d.opengl and sun.java2d.noddraw have to be set very early in the source code or as VM options. Someone on the official JogAmp forum changed the JOGL version used by Netbeans OpenGL Pack:
http://jogamp.762907.n3.nabble.com/JOGL2-need-some-advice-tp2181185p2181185.html

.

I already told you that the JOGL version of Netbeans OpenGL pack is obsolete and that your renderer should NOT be “GDI Generic”. Fix those problems and then, if the problems persist, fill a bug report. I use JOGL in complicated Eclipse RCP applications and in games, it works very good.

.

It is a known bug, Wade Walker is going to fix it. Schabby does this:

[quote]if (height <= 0) // avoid a divide by zero error!
{
height = 1;
}
[/quote]
whereas it should not be necessary, it is a crappy workaround. The animator should be launched only when the canvas and its container have a non null size. Rather use my example on Wikipedia. I will provide another one as soon as possible for Eclipse RCP.

.

.

.