On Sep 22, 2011, at 9:17 PM, Stéphane Ducasse wrote:

...

Now if some platforms don't adhere to IEEE754, then I think that we must first try to find a compatibility layer at VM level

What about just using http://www.mpfr.org ? The main goal of MPFR is to provide a library for multiple-precision floating-point computation which is both efficient and has a well-defined semantics. It copies the good ideas from the ANSI/IEEE-754 standard for double-precision floating-point arithmetic (53-bit significand). (I think I asked that questions once alrady...) Marcus -- Marcus Denker -- http://marcusdenker.de