[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: A short name.

On Thu, Feb 21, 2002 at 12:41:38PM -0600, Mario Latendresse wrote:
> I make this little proposition to Sebastian for a short syntax to his
> curry srfi.
> I think the most practical use of it is to reduce clutter in Scheme
> code.  
> Instead of `curry', I propose `_j'. This name comes from the
> projection notation in math. Also, I would use `_' instead of `<>'and
> `._.'instead of `<...>'.
> So it would look like this
> Instead of:                       You see:
> (curry cons a <>)                 (_j cons a _)
> (curry list 1 <> 3 <> 5)          (_j list 1 _ 3 _ 5)
> (curry list)                      (_j list)
> (curry list 1 <> 3 <...>)         (_j list 1 _ 3 ._.)
> And you get rid of the name problem of curry, partial, etc.
Mmm. In Mathematica, this operation is denoted with the & symbol.
(And the slots are denoted with #.)
So let me one-up your proposal and suggest we use &.

 (& cons a _)
 (& list 1 _ 3 _ 5)

As an added bonus, it makes Scheme look like Perl...

OK, so this isn't a serious proposal, but at least it is better
than using curry . I really don't understand how people can seriously
suggest to choose curry and go for a head-on clash with the whole FP
lterature. Especially since "real" curry *is* really possible to write
in Scheme, moreover, it is even occasionally useful. If you want a
short nice name, then in ${DEITY}'s name, please use _j, or &, or
tandoori, rather than curry.

If it really must be `curry', then I'm afraid that I am not going to
use this SRFI. OK, so that didn't scare anybody ;-) ,  but I want
to leave no doubt that I feel very strongly about this point, and
I haven't seen a single convincing reason to stick with curry.
There are 26^5 names with the same "shortness properties" as curry
and most of them do not clash with any established FP jargon.