[ODE] autoconf/automake

[Russ Smith]

> topic HUGE_VAL
> 
> perhaps the easiest is anyway 1.0/0.0 ?

will this be portable?

> but float f=HUGE_VAL; works too

really? i didn't realize. will this work across all platforms?
actually dInfinity does not have to be the IEEE infinity, it could
simply be a Really Big Number (10^whatever) that is close to the
representation's limit.

russ.

--
Russell Smith
http://www.q12.org