wagner@rocky.STANFORD.EDU (Juergen Wagner) (10/24/87)
Following the request I sent a personal email message to Steve because
I dodn't want to overload this newsgroup with this kind of stuff. Reading
this response, however, I have to make a comment. I thought, Steve was
supposed to give an AI course. Turbo Prolog is (as you already mentioned in
your article) not a real Prolog (this is not only *MY* opinion, there
has been a long discussion in this newsgroup some time ago).
So, how can people interested in AI (and typical AI languages, whatever
this might be) learn something about Prolog when working with
TurboP{rolog,ascal}. For some applications, Turbo Prolog might be a good
system. It has a very nice environment as opposed to many other Prologs
around. And last, but not least, it is fast. So far with the positive
aspects of Turbo Prolog.
The other side of the coin presents a language which is full of
declarations (sigh!), and therefore does definitively *NOT* allow the
typical rapid prototyping style which makes AI programming so
particular. Developing programs is something I would not like to do
under TurboProlog because changing one clause could require changes
throughout a whole bunch of declarations in other parts of my program.
The method of exploring a problem and its solutions is exactly what
beginners with AI languages have to learn who have already experiences
with classical programming languages.
So, my personal opinion is that using a REAL Prolog (C-Prolog,
BIMProlog, whatever) or a REAL Lisp (Scheme, CommonLisp, ...) not only
helps to grasp the philosophy of the respective programming language,
but also introduces into and affects the way of thinking and reasoning
about particular AI problems. It is of great importance that the first
contact with such a symbol manipulation language gives the right
impression rather than just presents yet another language some people
out there are using.
Juergen Wagner, (USENET) gandalf@portia.stanford.edu
Center for the Study of Language and Information (CSLI), Stanford CA
The usual disclaimer: No, it wasn't me, somebody must have used my account.