Re: GOOPS: New syntax for `define-method'?

[Ole Myren Rohne <ole dot rohne at cern dot ch>]

Michael Livshin <mlivshin@bigfoot.com> writes:

> 1. this is not standard.  AFAIK only SCM and Guile can do this.

Michael Livshin <mlivshin@bigfoot.com> writes:

> Ole Myren Rohne <ole.rohne@cern.ch> writes:
> > guile> (define ((test x) y) (* x y))
> > guile> ((test 3) 4)
> > 12
> > But why would anyone want to do that?
> 
> 1. this is not standard.  AFAIK only SCM and Guile can do this.

Just for the record:
mit-scheme yes
rscheme no
stk no

> 2. it can be handy:
> 
> (define ((curry+ a) b) (+ a b))
> (define 2+ (curry+ 2))
> (2+ 3)
>   ==> 5

Well, (define (curry+ a) (lambda (b) (+ a b))) is 9 full characters
longer:-)

> 3. I can't see why the new `define-method' syntax can't be made to
>    conform to this extension, except for the need to accomodate
>    defining setter methods.  you just can't have everything, I
>    guess...

And would it be worthwile anyway? This funny syntax started as an
unintentional side-effect of the define implementation, and should
probably be left as that.

	Regards,
        Ole