I'm munging order for my own purposes.
>>>> "Lars" == Lars Magne Ingebrigtsen
<larsi(a)gnus.org> writes:
Lars> "Stephen J. Turnbull" <turnbull(a)sk.tsukuba.ac.jp> writes:
> 4. Where the semantics of common constructs differ among
> elisp, Common Lisp, and Scheme, how hard is it going to be for
> casual elisp programmers to make the switchover? How hard is
> it going to be for core developers and Lisp implementors to
> make the switchover?
Lars> At a guess, I would think that Emacs Lispy people would (in
Lars> general) feel more comfortable switching over to Common Lisp
Lars> than to Scheme.
I know about comfortable. How about correct?
Lars> There are many subtle (and not so subtle)
Lars> differences between Emacs Lisp and CL, but I'd think that
Lars> many elisp people have gotten used to having those handy
Lars> functions and macros that cl.el provides.
Lars> With Common Lisp one gets a language environment that is
Lars> reasonably standard -- you can expect books like CLtL2,
Lars> commercial environments like Allegro, free environments like
Lars> CLISP, and (oh, joy of joys) an Emacs that one would know
Lars> how to use if one was familiar with any of the others.
Well, which is it? "Many subtle, and not so subtle, differences," or
"reasonably standard"?
I'm not being obtuse; logically, there is no necessary contradiction.
Cognitively, it's not clear to me that everybody can have both.
People who live in Emacs's Lisp, whatever flavor, will learn Scheme if
they have to, although they may not like it (this is important and
deserves heavy weight, but my point is otherwise). My question is:
are people who occasionally program in Emacs Common Lisp going to
introduce more problems than it's worth because they think they know
what they're doing because it "looks like" good ol' Emacs Lisp?
Lars> Anyway. Few would argue that Scheme isn't more
Lars> ideologically pure, but I dislike cleanliness. Common Lisp
Lars> is a great, towering beast, but you don't have to be
Lars> familiar with even a fraction of it to get lots of mileage
Lars> out of it. And as you learn more, you find that it's
Lars> basically all there.
Buffers? Fonts? Faces? Info? ...
Really?
Lars> The problem with Scheme is that's there's generally nothing
Lars> there, so you have to write all the code yourself, which
We will, anyway: buffers, fonts, faces, codesets, info, etc, etc. And
there are extension packages to make Scheme look like Common Lisp.
Chipsy claims they're pretty close to the real thing. Where are they
likely to break down, so I can go look and see if maybe they're good
enough for me (without learning all of that "great, towering beast"
just to take a half-informed look at Scheme-based CL extensions)?
Lars> means that you get a gazillion little functions that work
Lars> their way into the code base, all functions with their own
Lars> little peculiarites, which (in the end) means that Scheme
Lars> turns out to be even bigger than Common Lisp. (How big is
Lars> Guile now?)
Definitely not as big as CMU CL, but I don't recall the comparison
with CLisp.
Lars> Besides, Scheme doesn't have `when'. One can't not have
Lars> `when'.
One can. I do. Why can'two?
--
University of Tsukuba Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
Institute of Policy and Planning Sciences Tel/fax: +1 (298) 53-5091