Re: interface reductions

[Dirk Herrmann <dirk at ida dot ing dot tu-bs dot de>]

On 1 Jun 2000, Marius Vollmer wrote:

> Dirk Herrmann <dirk@ida.ing.tu-bs.de> writes:
> 
> > The proposed change is shown as a patch below.  It basically
> > deprecates SCM_FLUID_NUM, SCM_FAST_FLUID_REF and
> > SCM_FAST_FLUID_SET_X, and provides as a replacement for these the
> > macros SCM_FLUID_DATA and SCM_SET_FLUID_DATA.  If there are
> > suggestions for better names, I'd be happy to hear them.  Otherwise
> > I am going to apply that patch.
> 
> Why not just name the macros SCM_FLUID_REF and SCM_FLUID_SET?

It was agreed upon in a discussion to have the SET immediately follow the
SCM_ prefix.  Thus, following your suggestion we would have:

SCM_FLUID_REF and SCM_SET_FLUID or, maybe
SCM_REF_FLUID and SCM_SET_FLUID

That would also be OK, I think.  However, it seems that the _REF postfix
is primarily used in guile for types where there is not just one single
datum to be extracted:  
SCM_FRAME_REF(frame, slot)
SCM_BITVEC_REF(a, i)
I. e. the _REF macro takes an additional argument.  (I think those are all
other _REF macros found in .h files - except SCM_FAST_FLUID_REF.)  In
contrast, the _DATA variant is used in cases where there is exactly one
datum to be extracted.

I don't want to claim that this is a policy that we need to follow - I
just had to decide for a name and thought it would make more sense that
way.

> But more importantly, there should remain documentation about how to
> properly use SCM_FLUID_DATA, etc.  I think the comment should not be
> removed, just the last sentence of it.

Good point.

Thanks for your comments.
Dirk