🚀 go-pugleaf

RetroBBS NetNews Server

Inspired by RockSolid Light RIP Retro Guy

Article View: comp.lang.prolog
Article #4070

Re: Prolog interpreter/compiler on IBM (4MB memory) machines

#4070
From: Joaquim Baptista (pxQuim)
Date: Fri, 26 May 1989 17:19
55 lines
3100 bytes
Please consider using Prolog-2. I know Prolog-2 and Arity/Prolog, so I
will compare both. Prolog-2 is fully compatible with Edimburgh
format, with the following constraints and features:

- record, recorded, etc. are _simulated_ using assert/retract and thus
  _not_ recomended;

- It has tail recursion optimization, something you won't find in
  Arity/Prolog and that I find most usefull;

- It features modules (just as Arity/Prolog does). Both can simulate
  virtual memory;

- Prolog-2 can index a predicate using a hash function based in the
  functor and the arity of the first argument;

- Arity/Prolog has many unique features, like explicit B-trees and
  Hash tables, which are absolutely not portable;

- Both can be compiled, must you should be prepared to handle the bugs
  of the compiled code - perfectly good code will behave strangely
  after compilation (usually unification-related problems);
  Arity/Prolog beats Prolog-2 in compilation time (we never compared
  the execution times);

- Some Arity/Prolog predicates were not designed to be backtrackable.
  This means that when you backtrack over some system predicates which
  should succeed only once, you sometimes get strange behavior.

- I'm not sure of speeds, but they seem to have approximate speeds;
  perhaps you shoud choose one or the other based on the other
  features.

- I'm also not sure if any of this products uses any extended memory,
  but I'll bet they don't.

By the way, I used Prolog-2 and saw Arity/Prolog. I tend to like
Prolog-2 because of tail recursion optimization, and rejected
Arity/Prolog because of its many non-standard features (which, at some
time, you are compelled to use). However, Arity/Prolog version 5 seems
to have a better and more integrated environment than Prolog-2 version
2 (not that it's not an integrated environment with a very good help
function, which doesn't exist in arity/Prolog; I just don't like the
text editor, and ended up using 'vi').

I wanted to include the address to quest about Prolog-2, but I'm
having some trouble into getting to it... It will follow later.
--
--------
Joaquim Manuel Soares Baptista	| BITNET/Internet: px@host.fctunl.rccn.pt
Dept. Informatica		| UUCP:	px@unl.uucp
Universidade Nova de Lisboa	| ARPA: px%host.fctunl.rccn.pt@mitvma.mit.edu
2825 Monte Caparica		| PSI/VMS: PSI%(+2680)05010310::PX
PORTUGAL			| Sound: (+351) (1) 295 4464 ext. 1360
Technical Details
Message-ID: <PX.89May26171919@uniq.unl.fctunl.rccn.pt>
Path: arc1-ca.pugleaf.net!utzoo!attcan!uunet!mcvax!inesc!unl!uniq!px
References: <7535@charlie.OZ> <6963@bunny.GTE.COM>

Show Full Headers
Complete Article Headers
Path: arc1-ca.pugleaf.net!utzoo!attcan!uunet!mcvax!inesc!unl!uniq!px
From: px@unl.fctunl.rccn.pt (Joaquim Baptista (pxQuim))
Newsgroups: comp.lang.prolog
Subject: Re: Prolog interpreter/compiler on IBM (4MB memory) machines
Message-ID: <PX.89May26171919@uniq.unl.fctunl.rccn.pt>
Date: 26 May 89 17:19:19 GMT
References: <7535@charlie.OZ> <6963@bunny.GTE.COM>
Sender: px@unl.fctunl.rccn.pt
Followup-To: comp.lang.prolog
Distribution: comp
Organization: Universidade Nova de Lisboa, Portugal
Lines: 55
In-reply-to: slzr@GTE.COM's message of 22 May 89 14:03:26 GMT
Posting-Front-End: GNU Emacs 18.44.13 of Wed Jun 24 1987 on uniq (berkeley-unix)