This is the mail archive of the
guile@sources.redhat.com
mailing list for the Guile project.
Re: pssyntax.ss and module system
- To: "Matthias Clasen" <Matthias dot Clasen at poet dot de>
- Subject: Re: pssyntax.ss and module system
- From: Jost Boekemeier <jostobfe at linux dot zrz dot TU-Berlin dot DE>
- Date: 25 Aug 2000 17:02:15 +0200
- Cc: <guile at sources dot redhat dot com>, <djurfeld at nada dot kth dot se>, <kxn30 at po dot cwru dot edu>
- References: <012401c00e9c$f6a4dae0$e30b0ac0@poet.de>
["clean" modules]
"Matthias Clasen" <Matthias.Clasen@poet.de> writes:
> (http://www.cs.indiana.edu/~dyb/papers/popl99.ps.gz) in which Waddell and
Thanks for the pointer! There is one thing that concerns me however:
They seem to extend the Scheme language itself with the special forms
"module" and "import" and build the module system on top of the R5RS
language + the macro facility + the two special forms, don't they?
The problem with this is: How should the module system scope the macro
facility, the R5RS language or the two special forms istelf?
In Rees' system the module system is *not* part of the language
itself and although the module configuration language looks similar
to scheme, it can only be used to configure the module system.
I think it pays to compare their solution with what Tung proposes in
his dissertation. -- For example Tung's module system "grabs"
auxiliary bindings from a syntax module, while their solution does not
require special support from the module system. Which solution is
better?
Jost