Mikez2015

> * Where does the direction vector that is being visualized come from? My program takes a visualized direction vector from "dir": ``` int ccdMPRPenetration(const void *obj1, const void *obj2,...

Strange, I checked with another algorithm, and it turned out that **ccdVec3PointSegmentDist2** returns the correct value. Now I'm at a dead end.

Another video with the same problem, now a box-box for simplicity, with depth printing and normal penetration. It is amazing how such a issue was not noticed at the very...

I am very grateful to you that you continue to help me. Your help is very much needed. > Here's my suggestion for tracking down this bug: > > 1....

> The new visualization is wonderful! You can see the bad behavior is right on the very edge of the triangle. That suggests to me that it's related to the...

> The new visualization is wonderful! > Wow! That's an amazing visualization of MPR! Thank you, I'm glad you liked it. :)

> 4\. Now put print statements _everywhere_. We know the final calculation is producing a normal in a bad direction, but we don't know where that _started_, so, ideally, we...

Have same problem - fatal error LNK1120. I tried to do as advised **vekkuli**, but it did not help.

> > > I have the same problem. You can try to add extern int ccdGJKIntersect(const void *obj1, const void *obj2, const ccd_t *ccd) before main() function. warning C4273: ccdGJKIntersect:

> > Warning means you can run it but you still have some potential problem. In most situation, it is just a useless hint. It depends on your compiler. But...