Re: interface reductions

[Mikael Djurfeldt <mdj at mdj dot nada dot kth dot se>]

Dirk Herrmann <dirk@ida.ing.tu-bs.de> writes:

> I think we mostly agree.  Maybe the only thing we disagree about is what
> we consider as valid alternatives.

(I actually haven't considered that very carefully.  :)

> SCM_FAST_FLUID_REF    --> scm_fluid_ref
> SCM_FAST_FLUID_SET_X  --> scm_fluid_set_x

Those look like they have some special reason for their existence.
Always before removing things, we should find the ChangeLog entry
where they were added---that may give a clue.

Let me check if I can find the reason.

> SCM_BIGSIGN           --> scm_positive_p / scm_negative_p
> SCM_XXXP              --> scm_xxx_p  (for most predicates)

I'm not fond of the name SCM_BIGSIGN, but it is surely important to
be able to check the sign of a bignum quickly, since this can be very
time critical when implementing a new numerical operation for
bignums.  (And note that such an operation might be provided by a C
module.)

Also, in my view, all SCM_XXXP macros should be kept and exported,
because they can be time critical.


BTW, thinking about it, SCM_KEYWORDSYM falls in the category of not
being useful enough to warrant restricting the implementation the way
it does.  Let's deprecate it!  :)