bug in OpenGL ES glRotatex ?

[rsalmin]

It seems that rotations around vectors
(1, 0, 0) and (-1, 0, 0) at the some angle are equal (at least visualy) instead of supplemnt each other to 360 degrees. does someone experinced the same problem?

I am new in Symbian and OpenGL so may
be it's just my misunderstanding of what glRoate does?

Can anyoune help?

I am using SDK S60 2nd fp2

[Flame]

The reason why they are equivalent is that the rotation vector is the same vector.

Vectors (1,0,0) ; (-1000,0,0) ; (0.001,0,0) all translate to same normalized vector but with different scale (negative or positive).

Jani Vaarala
Graphics Architect
Nokia

[rsalmin]

but the rotation around vector (-1, -1, 0) not the same as rotation around (1, 1, 0)
the sign is significant! when you inverse vector the direction of rotation is also changed. It seems this isn't happen only for vectors with only one non zero coordiante.

[Flame]

Actually you are right (I just had a look at the GL spec). I forward this to our OpenGL ES implementation team to check.

Jani Vaarala
Graphics Architect
Nokia

[Flame]

OK. This is actually a bug in the implementation. Good thing about it is that
It only affects to axis-aligned rotation vectors (also the reason why it wasn't picked up earlier).

Thanks for the info.

Jani Vaarala
Graphics Architect
Nokia