[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: Uniform vectors: was Questions about floating numbers
From: |
Dirk Herrmann |
Subject: |
Re: Uniform vectors: was Questions about floating numbers |
Date: |
Wed, 21 Nov 2001 23:30:33 +0100 (MET) |
On Wed, 10 Oct 2001, Martin Grabmueller wrote:
> > From: Dirk Herrmann <address@hidden>
> > Date: Wed, 10 Oct 2001 20:18:28 +0200 (MEST)
> >
> > On Wed, 10 Oct 2001, Keith Wright wrote:
> >
> > > That would be srfi-4, but note also its anti-srfi, srfi-10, which
> > > proposes a more general syntax. I find srfi-4 to be unpleasantly
> > > full of special cases while missing e.g. bit vectors. Why should
> > > there be a special TAGvector-ref for each type of vector, instead
> > > of just letting vector-ref, or at worst uniform-vector-ref, check
> > > the type of its argument? Anti-virtualization!
> >
> > True. But, instead of providing our own set of uniform vector syntaxes
> > again - and running into compatibility problems later - it is wise to see
> > which #<letter> combinations are already taken.
> >
> > And, given srfi-10, there is no need for a new syntax for uniform vectors
> > at all.
>
> I don't know if you have noticed, but CVS Guile has srfi-4 as well as
> srfi-10 for quite a while now. Try it out, comment on it and maybe we
> can simply throw out the old uniform vector syntax...
Sorry for not answering for so long...
However, since I am not using the old uniform vector syntax, I'd say yes,
throw it out :-) Or, rather, first deprecate the old interface pointing
to your implementation of srfi-4. Then, we can throw it away in guile
1.10.x.
Best regards
Dirk Herrmann
[Prev in Thread] |
Current Thread |
[Next in Thread] |
- Re: Uniform vectors: was Questions about floating numbers,
Dirk Herrmann <=