[ec2m] Re: ec2m-list digest, Vol 1 #13 - 2 msgs

[Chris Leger]

> > From: ec2m-list-admin at alife.org [mailto:ec2m-list-admin at alife.org]On
> > Behalf Of Adam Rotaru

I'd add one other, which is satisfied by most packages I've seen (or
written :) ):

- The ability to simulate new mechanisms without recompilation.  At
least one dynamic simulation package generates source code based on a
mechanism description, and if you change the mechanism, you've got to
re-generate the source, compile, and link it.

I would also suggest the capability for interacting with more complex
fixed objects (e.g. terrain) than a simple ground plane.  The previous
comment about the infeasibility of absolute repeatability agrees with my
experience; compile with -g instead of -O, or on a different platform,
and you'll likely get different results.  If the ALife organisms being
generated are that sensitive to minor numerical deviations, then either
some sort of adaptive control or stochastic, multiple-run evaluation is
probably necessary.

And finally, I'd like to point out that, to the best of my knowledge
(which is slightly out of date), there are currently no methods for
dynamic simulation of general frictional contact between articulated
rigid bodies that are physically accuracte, numerically robust, and
fast.  
Pick any two  :)

Chris

-- 
"The worst habit you can possibly acquire is to become uncritical 
towards your own concepts and at the same time skeptical towards 
those of others. Once you arrive at that state you are in the grip 
of senility, regardless of your age."       -Karl von Terzaghi