Subject: Re: source access vs dynamism
From: Erik Naggum <erik@naggum.no>
Date: 1999/09/04
Newsgroups: comp.lang.lisp
Message-ID: <3145425004322182@naggum.no>

* William Tanksley
| Fortunately, the purpose of this document wasn't to understand Lisp, but
| rather to explain and compare it.

  huh?  if one end tries to explain and compare, what other than understand
  should the other end try to do?  I must have missed something.

| I can't bring myself to understand why anyone objects to snapping off the
| word "library" when referring to a collection of functions.  Would you
| prefer "magazine" or "clip"?

  "library" has very specific connotations and annotations in programming
  language, none of which are applicable to Common Lisp.  the whole concept
  is very, very different from what we do in Lisp.  I'd expect "library" to
  be the equivalent of "package" in Common Lisp.  since there is but one
  standard package, which holds the entire language, there _is_ no library.

| Read "On Lisp"; it's a magnificent book.

  I know.  I read it when it came out.

| The core problem is unintentional variable capture.

  well, the problem with hygienic macros is that you can't have intentional
  capture.  however, the problem is so much more severe in Scheme that many
  are actually afraid of macros because of it.  this is not something I say
  because I don't like Scheme.  I don't like Scheme because it has this and
  many other problems that Common Lisp doesn't have, mainly because it did
  retain the two namespaces that Scheme discarded.  I've been bothered by
  the same problem that Scheme has in other languages, too, like C.  it's
  just stupid to have one namespace: a function can be called and variables
  can't and that's a sufficiently fundamental difference that I think two
  namespaces fall out naturally, and I have created the distinction when I
  needed to design my own languages.  nobody seems to think it's a problem.
  except Scheme people.  so I'd say the _severity_ of the macro problem is
  an artifact of the one namespace decision, and that the problem is so
  easy to curtail otherwise that the problem is not perceived as a problem
  that needs solving.

#:Erik
-- 
  save the children: just say NO to sex with pro-lifers