scsh-users
[Top] [All Lists]

Re: Scheme disutility

To: scsh@martigny.ai.mit.edu
Subject: Re: Scheme disutility
From: sperber@informatik.uni-tuebingen.de (Michael Sperber [Mr. Preprocessor])
Date: 20 Apr 1997 15:38:07 +0200
Sender: sperber@informatik.uni-tuebingen.de
>>>>> "Martin" == Martin Cracauer <cracauer@wavehh.hanse.de> writes:

Martin> But what's wrong with SLIB's approach? Provide the interfaces you want
Martin> on almost every Scheme implementation, either as Scheme code on top of
Martin> existing functionality - maybe slow - or you could go and implement a
Martin> feature on a given implementation  in an efficient manner.

The major problem with SLIB is that it imposes a packaging approach
which does not port well to, say, batch-compiled systems such as
Bigloo and Gambit-C.

Here is generally a big problem with the usability of Scheme: How do I
ship a piece of Scheme code and have it install under multiple Scheme
implementations?  At least, all major C compilers are batch compilers
and have pretty much isomorphic (read: "make"-customizable)
interfaces.

Admittedly, this problem is a hard one to solve.

--
Cheers =8-} Mike
Friede, Völkerverständigung und überhaupt blabla

<Prev in Thread] Current Thread [Next in Thread>