From ... From: Erik Naggum Subject: Re: IS(O-)Lisp status? Date: 1996/07/23 Message-ID: <3047102938994282@arcana.naggum.no>#1/1 X-Deja-AN: 170445810 references: <4t0b6d$ge1@trumpet.uni-mannheim.de> <3047066329923457@arcana.naggum.no> <4t22fj$11q@goanna.cs.rmit.edu.au> organization: Dept. of Informatics, University of Oslo, Norway newsgroups: comp.lang.lisp [Richard A. O'Keefe] | I don't think the ease of calling C functions from Sun's ESH can be beaten. | If I read the (somewhat patchy) documentation, this even extends to C++. | Step 1: slurp the appropriate header using "ix" | Step 2: load the Scheme file written by "ix" | Step 3: you're away laughing. you realize, of course, that this can be true for any product and any language, regardless of how much work this "ix" thingy does. | What this establishes is that there is no reason for gratuitous ISLisp/ | Scheme incompatibility on the grounds of ease of foreign interfacing. non sequitur. unless you're in the business of selling "ix"'es or "esh"'es. | ESH provides this via a borrowing (without acknowledgement) of the old | Pop-2 idea of "updaters". Again, it's an *extension* to Scheme, with a | proven record of working extremely well in practice since the '60s, | that doesn't need to break existing code. the issue is that ISO Lisp has opted for a notation for dynamic variables that differs significantly from the traditional way to declare special properties on variables, and has not done so consistently or open-endedly. I actually fail to see what ESH is doing in here at all, unless it has been a major influence on the design of ISO Lisp, which I sort of doubt. #\Erik