From: David Dumas ([email protected],
September 17, 2003)
Hi William,
As per your request in the Neron login message, I'm writing
to let you know that I'm using this machine for some
computations.
Specifically, I'm running some big calculations of holonomy
representations of complex projective structures on
punctured tori. The set of comples projective structures
examined is limited to those with a fixed underlying
conformal structure, and it is this implicit constraint that
requires most of the calculations.
In addition to finding generators of the holonomy groups, my
program also guesses whether or not the group is discrete
(as a subset of PSL_2(C)) using an algorithm inspired by
Brian Bowditch's work on Markov triples and quasifuchsian
groups.
A typical application of the resulting data would be to
draw a picture of the Bers embedding of Teichmuller space.
I have recently made this software package available under
the GPL, though it is still in rough shape. The
semi-permanent home page is
http://www.math.harvard.edu/~ddumas/bear/
BTW, if the people at Sun who read your quarterly reports
have any suggestions regarding optimization and
parallelization of ODE solvers on the sun ultrasparc III
architecture, I'd love to hear about it!
thanks,
David |