Re: IEEE floating point support for guile?

[Mikael Djurfeldt]

Jim Blandy <address@hidden> writes:

> - By default, floating-point operations in Guile should raise
>   exceptions, rather than generating Inf or NaN values.  On the IA-32,
>   this implies enabling software exception handling for underflow,
>   overflow, zero divide, denormalized operands, and invalid operation
>   exceptions.
> - Guile should eventually provide some way to support users who want
>   to work with extended doubles, or prefer non-signalling operations,
>   but implementing the earlier suggestions should not be delayed until
>   we agree on the interface to these more sophisticated facilities.

Since this is very basic functionality which has had one behaviour
under the decade in which the SCM/Guile code has existed, I think
changing it will break many programs.

I therefore think we should delay generating exceptions for
floating-point operations until we have designed and implemented the
interface for disabling these exceptions.

Besides, no one of the major Scheme implementations (including Chez,
RScheme, Bigloo and MzScheme) generate floating point exceptions.  I'd
like to understand this issue better before making the change.