Dear Larry,

On the long-term plan, some improvements in the Boissonat surfacing algorithm
would be very useful. John's current implementation works fine most of the time
for simple concave structures but may generate non-manifold mesh (ie: not
"water-tight" and with intersecting triangular faces). The appearance is usually
OK but problems starts to appear if you try to calculate the enclosed volume or
use smoothing functions.
With my regards and thanks for your support,
Thomas Launey

--
Thomas LAUNEY, PhD
Unit Leader
RIKEN, Brain Science Institute (BSI)
Launey Research Unit for Molecular Neurocybernetics
2-1 Hirosawa, Wako-shi 351-0198 Saitama, Japan
Web: www.brain.riken.jp/en/t_launey.html