Subject: Re: Avoiding unintentional variable capture
From: Erik Naggum <erik@naggum.no>
Date: 1999/09/11
Newsgroups: comp.lang.lisp
Message-ID: <3146052864905406@naggum.no>

* Iver Odin Kvello <iverk@hfstud.uio.no>
| Gambit Scheme ...
| 
| Also, in the R5RS under 6.3.3, Symbols: "Some implementations have
| 'uninterned symbols' ...".[2]

  we're talking about the _language_ Scheme, here.  you just can't base a
  _language_ feature on something that only exists in some implementations,
  and you certainly can't dismiss it as if the existence of implementations
  that go way beyond the standard is a feature of the language.  quite the
  contrary in my view.  so I'm being anal and discuss the language as per
  specification.

| The program-as-text thing of the RnRS is pretty strange though; I've
| never seen any real argument of why it's supposed to be a good idea[4].

  well, it does make it much easier to construct tools for it in non-Lisp
  languages, using lex/yacc, and more "usual" compiler techniques.

| DEFMACRO obviously isn't a safe-syntactic-abstraction-Right-Thing all by
| itself, but it seems pretty Right-Thingish anyway, given  what it can do.

  I agree with this, but I also think it is sometimes impossible to get it
  all at once, and when this is the case, it's more important to settle on
  _something_ and agree that the issue has been settled than to oscillate
  between things you never settle on.

#:Erik
-- 
  it's election time in Norway.  explains everything, doesn't it?