Re: [Chicken-hackers] Numbers egg interaction with other compiled code.

[Tony Sidaway]

On 10/22/09, John Cowan <address@hidden> wrote:
>
> I've been talking privately with felix about a strategy for changing
> this, using hooks that the numbers egg installs procedures onto rather
> than just rebinding identifiers, but he's worried about performance.
>

An alternative might be to provide chicken-install with the capability
to produce parallel egg binaries--one with numbers and one without--on
an egg-by-egg basis.  Fork the local library repository
(/usr/local/lib/chicken/4 or whatever) and make bignums infectious by
causing bignum-using modules transparently load the bignum version of
other eggs (from /usr/local/lib/chicken/4/numbers/ or something).  An
egg compiled with "-require-extension numbers" (a "Bignum egg") would
have code (perhaps as part of the numbers egg) to change the dynamic
load behavior to always load the bignum version of a dependency if it
exists.

For those who aren't interested and want to retain Chicken's speed,
the local library repository would remain unforked because they'd
never switch the dual-compilation facility on.  Bignum eggs would fall
back to loading the vanilla version of a module on which they depend,
so their behavior would be as it is at present.