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

Re: C# is not Dylan (was: Re: C# : The new language from M$)

Jeff Dalton <jeff@aiai.ed.ac.uk> wrote:
> Neel Krishnaswami wrote:
> 
> > Dylan plus its special forms is substantially less complex
> > than Common Lisp and its special forms.
> 
> Do you really mean special forms, i.e. omitting all the macros?

Yes, except for reader macros, which I regard as (partially) defining
the lexical structure of Lisp. There's a whole host of surprisingly
complex special forms (eg, TAGBODY/GO) and reader macros (eg the #n#
stuff for cyclic lists). This will all bulk up the size of a grammar
defining syntactically valid programs quite a bit -- more than many
people realize.

I mostly intended this as a mild challenge to a piece of common
wisdom. I'll agree it's not a terribly principled position, as I'm
drawing an arbitrary boundary between different things in the
COMMON-LISP package.

To compare the languages as they are actually used, something like
"All the macros and special forms in the Dylan-User module are less
complex than the macros and special forms available from the
COMMON-LISP package" would be a much fairer statement. It's still
accurate, too.

But don't read too much into that, please -- quasiquote is a much
bigger difference than any infix/prefix difference, and one that's not
been touched on at all in this discussion.


Neel