From ...
From: Erik Naggum <erik@naggum.no>
Subject: Re: Avoiding unintentional variable capture
Date: 1999/09/29
Message-ID: <3147590311986519@naggum.no>#1/1
X-Deja-AN: 530659897
References: <37d6b784.46829757@judy> <3146036566024257@naggum.no> <7smo4l$ac898@fido.engr.sgi.com> <3147423680626348@naggum.no> <7ss1vq$b7oc7@fido.engr.sgi.com>
mail-copies-to: never
X-Complaints-To: newsmaster@eunet.no
X-Trace: oslo-nntp.eunet.no 938601515 17147 195.0.192.66 (29 Sep 1999 10:38:35 GMT)
Organization: Naggum Software; +47 8800 8879; +1 510 435 8604; http://www.naggum.no
NNTP-Posting-Date: 29 Sep 1999 10:38:35 GMT
Newsgroups: comp.lang.lisp

* Rob Warnock
| But I'm not ducking your main point (as I understand it), which I take to
| be that RxRS Scheme does not IN THE SPEC provide a low-level macro
| facility such as "defmacro" which includes the needed re-evaluation by
| "eval" that is needed to make any *use* of such generated sysmbols. It
| doesn't.

  Scheme people seem to take much pride in that specification.  I think
  it's fair to take their word for it and not sully the issue with what
  some implementations had to do to make it work in real life.  if we do
  that, we can't also complain that Common Lisp doesn't have features X and
  Y and Z when many implementations do support them.  there is a disturbing
  inconsistency in how Scheme people want to regard specifications, and I
  don't want to go along with that.  so of course it's IN THE SPEC that is
  the issue.

#:Erik