[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: is that useful?
>>>>> On Tue, 26 Feb 2002 15:53:12 +0100, sperber@xxxxxxxxxxxxxxxxxxxxxxxxxxx (Michael Sperber [Mr. Preprocessor]) said:
> Sure, but we're not writing a language standard, we're writing a
> library proposal. The problem with Common Lisp is that the
> redundancy is not sufficiently modularized. There's no problem with
> redundancy in the SRFI process --- indeed we're expecting and
> soliciting a lot more of it. Will Clinger has some excellent
> thoughts on the issue at
As one of the original editors of SRFIs, obviously I am in favour of
the process, and the structure. However, I am not very keen on the
idea of having my program starting with a pile of SRFI-0 code to
determine what SRFIs are available. Obviously in some cases this is
necessary and appropriate (I don't necessarily agree with Will's
argument, but in the case of SRFI-6 & friends it makes fair sense).
But in the case of SRFI-26 or other very small bits of code, I am
*far* more likely to simply define my own macro that does what I need
than to have a bunch of SRFI-0 code to find out if what I want is
available and if not define my own. On the other hand, if it was a
more general SRFI that solved most of the problems, I'd be more likely
to use it.
All that, of course, is independent of the choice of a confusing name.
While the Scheme community and the Haskell community may be fairly
disjoint, I don't think it is true of Scheme and ML (who also talk
about curry all the time). It would also be really nice at ICFP or
other places where Schemers and Haskell types get together if the
Schemers don't look like complete idiots when they misuse the term
`curry' because of this SRFI's poor choice of name. Finally, honoring
Curry with this SRFI is like honoring a great artist by naming a
water-treatment plant after them - a nice thought, but not
(BTW, is it possible for the email address to show as
srfi-26@xxxxxxxxxxxxxxxxx rather than an address at uni-tuebingen.de?)