[World] Yet another step on the way to constructive modeling

[Daniel Phillips]

Another slight adjustment of the two tetrahedra gives an intersection 
consisting entirely of interlocking triangles as handled by the algorithm 
described in the previous post, except for one case where two edges of one 
triangle pierce another. This intersection therefore must be solved by  a 
combined algorithm.

The solution in the image is correct in the sense that if we cut loose the 
corner of the green tetrahedron inside the red tetrahedron along the white 
cutting arc, turn it inside out, and glue it to the red tetrahedron along the 
cutting arc, the result is a solid that is exactly the green tetrahedron 
subtracted from the red tetrahedron, an honest to goodness constructive 
boolean operation.

Even though the result is correct, the algorithm is not yet correct. In two 
places where it made the wrong decisions I just plugged in the right decisions 
as magic numbers. There is still some unknown amount of work ahead of me to 
make it come up with the correct decisions on its own. Still, it looks pretty 
cool, no? This is one of those algorithms where shortly after becoming 
functional on a few simple test cases it will most likely just work on meshes 
of arbitrary size and complexity.

Regards,

Daniel
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://phunq.net/pipermail/world/attachments/20120610/7c38b0b7/attachment-0001.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: csg4.png
Type: image/png
Size: 123489 bytes
Desc: not available
URL: <http://phunq.net/pipermail/world/attachments/20120610/7c38b0b7/attachment-0001.png>