Newsgroups: comp.lang.dylan
Path: cantaloupe.srv.cs.cmu.edu!das-news2.harvard.edu!news2.near.net!news.mathworks.com!uhog.mit.edu!bloom-beacon.mit.edu!spool.mu.edu!howland.reston.ans.net!ix.netcom.com!netcom.com!netcom2!jrd
From: jrd@netcom2.netcom.com (Jon Degenhardt)
Subject: Re: Floats and (Re: The C/Dylan interface)
In-Reply-To: ram@cs.cmu.edu's message of 17 Nov 1994 17:28:27 GMT
Message-ID: <JRD.94Nov18210031@netcom2.netcom.com>
Sender: jrd@netcom.com (Jon Degenhardt)
Organization: NETCOM On-line services
References: <aberno-1611941121310001@lyapunov.stanford.edu>
	<3ag3rr$hma@cantaloupe.srv.cs.cmu.edu>
Date: Sat, 19 Nov 1994 05:00:31 GMT
Lines: 24

In article <3ag3rr$hma@cantaloupe.srv.cs.cmu.edu> ram@cs.cmu.edu (Rob MacLachlan) writes:

>  Some context on interfacing to other languages: Dylan implementors are not
>  currently trying to standardize this interface, but at CMU we believe that
>  standardization is important.  Currently we are working independently, but are
>  trying to avoid gratuitously diverging from Apple's current plan (as we
>  understand it.)
>
>  The general idea, which is clearly "the right thing", is to parse C header
>  files and automatically generate:
>   -- Dylan operations which can create and modify external-format data
>      structures, and 
>   -- Dylan functions which can be called to call foreign functions.

A related question: Do current plans provide ways for Dylan functions to be
called from a C/C++ main program? That is, will it be possible to insert
Dylan functions into what is otherwise a C/C++ executable control by a
C/C++ main() function?

I've wondered about this for quite a while, but have seen little
discussion.

--Jon Degenhardt
  DocuMagix Inc.
