home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!munnari.oz.au!goanna!ok
- From: ok@goanna.cs.rmit.oz.au (Richard A. O'Keefe)
- Newsgroups: comp.lang.ada
- Subject: Re: AI and Ada (Was: Multiple Inheritance -- the last post?)
- Message-ID: <14425@goanna.cs.rmit.oz.au>
- Date: 9 Sep 92 06:23:22 GMT
- References: <1992Aug22.221840.2982@seas.gwu.edu> <59831@mimsy.umd.edu> <1992Sep8.195650.5728@sei.cmu.edu>
- Organization: Comp Sci, RMIT, Melbourne, Australia
- Lines: 31
-
- In article <1992Sep8.195650.5728@sei.cmu.edu>, jldh@sei.cmu.edu (Jorge Luis Diaz-Herrera) writes:
- > Briefly, Ada83 satisfies most "special" AI programming requirements,
-
- I like Ada, but _this_ I have trouble with.
-
- > although a complete solution to the problems is still lacking, Ada9X promises
- > to be an excellent software engineering (concurrent) language strongly
- > supporting many typical AI programming paradigms (such as, functional
- > programming, logic programming, frames, and, of course, "programs as data",
- > not to mention object-oriented technology).
-
- I have read several of the Ada9X documents, and really appreciate all the
- work that has gone into it. There's some wonderful stuff coming. Now, I
- haven't read _all_ the Ada9X documents (I don't think the department has
- enough paper), so I missed the bit where they were going to add support
- for logic programming. It really does come as a surprise that Ada9X will
- "STRONGLY SUPPORT" logic programming. I would definitely like to hear more
- about this, and I'm sure many other comp.lang.ada readers would too.
-
- > p.s.: by the way, studies have shown that Ada is several times more efficient
- > than traditional interpreted AI languages (just in case you are worried about
- > efficiency considerations).
-
- This is a very curious and rather irrelevant assurance. The AI languages
- _I_ use have compilers. Reasonably good ones. How does Ada-for-AI compare
- against traditional (dynamically) *compiled* AI languages? (Take as
- reference points SICStus Prolog generating native code, MIT Scheme generating
- native code, and CMU Common Lisp. While you're at it, you might try SML/NJ.)
-
- --
- You can lie with statistics ... but not to a statistician.
-