Subject: Re: Guide to Lisp, v1.20
From: Erik Naggum <erik@naggum.no>
Date: 03 Sep 2002 17:04:16 +0000
Newsgroups: comp.lang.lisp
Message-ID: <3240061456940076@naggum.no>

* Wolfhard Buß
| Allen calls "these representation-dependent coding tricks" dangerous.

  Does he have any alternatives in mind?  Common Lisp has made a trade-off in
  this regard where the general mechanisms depend on run-time type information
  to make their decisions and compilers generally do not rewrite calls based
  on available type information.  This is in sharp distinction to statically
  typed languages, where syntactically identical access mechanisms may be used
  independent of the types, but with only abstractly similar semantics.

  Some Scheme courses I have seen construct an abstraction barrier between the
  lower-level access functions and the abstract names for their accessors.
  This may have its merits, but I found it needlessly verbose and error-prone.
  Common Lisp offers defstruct with a `:type´ argument that may be used to
  construct lists with accessors.  Is that something Allen would prefer?

-- 
Erik Naggum, Oslo, Norway

Act from reason, and failure makes you rethink and study harder.
Act from faith, and failure makes you blame someone and push harder.