Thought I should add a brief comment to Shlomi's last on this subject.
I concur essentially with his description of the state of affairs. To
update, I've yet to tackle the specific problem to which Shlomi referred,
because along the way a couple of other problems cropped up which took some
time. I'm now finally clear to look at that remaining "minor" problem.
The multiple soft thread system is now functioning properly, and I'm
very encouraged by the performance I'm seeing. For FcPro users' principal
use of a solver, i. e., for a spot check whether a given position is
winnable or not, my preliminary observation is that FCS might turn out to be
the superior solver as against the modified Don Woods and Patsolve. Except
for the "flaw" that it can announce a false impossible. That means, for a
user, that while it might be faster on the average than the other solvers,
users cannot trust an "impossible" verdict. This isn't a fatal flaw, as hot
key usage can be made of either other solver when a user sees FCS say
"impossible". But it will be an annoyance. Shlomi has indicated he has an
idea on how to eliminate that problem, and I'll look forward to integrating
that when it comes.
Once I feel it's "ready for prime time", I'll make it available for
downloading..
-------------------------Adrian Ettlinger
Received on Sat Jun 01 2002 - 20:24:52 IDT