Solve remaining capsule collision issues about ioq3ztm HOT 3 CLOSED

The "groundEntityNum" bug probably could be fixed than. However, cylinders 
might have their own drawbacks. In common it seems that the math behind 
cylinder collision is not that fast as for boxes/capsules. There are many 
articles about this, e.g.:

http://forum.unity3d.com/threads/63967-why-cylinder-collider-doesn-t-exist-!

..."Mathematically (with respect to collision detection at least), cylinders 
are much harder to work with than triangles, rectangles, spheres, capsules, or 
boxes. Algorithms involving cylinders are generally less optimal (considerably, 
in some cases) than similar algorithms involving other simple primitives, so it 
makes sense that cylinders would not be supported as a basic primitive."...

It would be very helpful to optimize the math/algorithms behind Q3's q_math.c 
first, I think. Unfortunatly, at Ioid3 we don't have the skills to do that 
(atm).
So, yes, I think cylinders would fix some of the issues, though I wonder how D3 
did the collision.

Doom3 uses a box for player collision. (I don't know if it rotates or if it's 
axial like in Quake3.) There is a cvar for enabling 8-sided 
cylinder-approximation for players, pm_usecylinder.

I can see how a 8-sided cylinder-approximation would be simpler to handle, 
makes it possible to treat cylinder more like boxes and map brushes.

Moved to github issue tracker https://github.com/zturtleman/spearmint/issues/6