Jiba wrote:
I think it is a trouble with the culling system ; i'm going to investigate it.
While you're at it, would it be possible to also arrange for front/back faces to be swapped when rendering in a mirrored coordinate system? That would be more efficient than having to use double-sided faces.
BTW, in the future, the new model deforming API may be another way for mirroring models.
Model deforming sounds interesting, but I'd rather not *have* to use it just to get a mirrored object. Currently I'm building models using SketchUp and loading them using a custom export/import script arrangement. It would be easiest if I could just use whatever transformation matrix was used in the SketchUp model, instead of having to figure out whether it's mirrored and do something different. On another topic, you can disregard what I said about Geoms earlier, since it seems it already works the way I wanted. Not sure why I thought otherwise -- there must have been something else wrong with my code. However, I've discovered another small limitation -- you don't seem to have provided individual access to the three collision functions (dCollide, dCollideSpace and dCollide2), instead providing one prepackaged recursive collision algorithm. That's handy, but it shouldn't be at the expense of access to the underlying functions, because the user might want to do things differently. As an example, for one application I have in mind I'd like to be able to test whether a particular object intersects anything else in the model. The natural way to do that would be to do dCollide2 between the geom of that object and the space for the model, but I can't, because dCollide2 is not exposed. Thanks, Greg _______________________________________________ Soya-user mailing list [email protected] https://mail.gna.org/listinfo/soya-user
