Fare Rideau
Sat, 9 Aug 1997 16:07:48 +0200 (MET DST)

Dear Tunes people,
   some of you (Alaric and Sean, not to name them),
have asked me in private e-mail what platform did MACISTE run on,
and whether anything (code, docs, examples) about it was net-available.
   Well, the only machine MACISTE runs on is an old SparcStation 1+ under
SunOS 4.x, though it should be easily portable to any 32-bit architecture
under a similar OS (e.g. Linux/i386). However, this agains seem to be
no priority to Jacques Pitrat (though I hope I can have him change his
mind if I work with/near him), so don't expect anything soon.
MACISTE's deep internals are being reworked so as to remove
many builtin limits that were inherited from its ancient history,
and allow for automatic generation of numerous meta-heuristics;
hence it's unstable, and you wouldn't like to inspect its present state
(46MB worth of self-written source, bytecode, goto-ish C, binaries);
month old stable tapes could be made available, but it'd be difficult
to get much from it. Again, closed development is harming the success
of it quite a lot.
   The language includes features from Lisp, OO, Prolog, and more,
with an infix syntax that's the result of 40 years of evolution
loosely interacting with any particular other language;
historical reasons made it all upper-cases (it seems),
though this could have been easily fixed, if it wasn't in the author's taste.
Just everything is first-class, since the system is reflective;
of course, it's got lots of personal tuning by Jacques Pitrat.

   As for published Papers, Jacques Pitrat wrote quite a lot of them,
most in French I believe, about topics related to AI
(his first meta-program was a theorem prover in the 50's),
in all kinds of journals.
His vocabulary is particularly affected by his old-times AI background.
Subprograms/modules are called (fr)"expertise", etc.
Similarly, the syntax (that he demonstrated was easy to globally modify;
though again, no modularity in doing so was foreseen)
is not patterned after any particular existing language.
   I'll meet Jacques Pitrat again in September,
and will try to make things advance in the right direction,
while trying to find something of a bibliography of his publications.
Lots of it should be in French, though.

   Anyway, let all that not prevent you from having a nice summer!

PS: to Jecel -- yes, it's admittedly hard to promote open development,
all the more for things that are not of immediate use for lots of people.
So the challenge is precisely to write software that, without sacrificing
long-term utility, DOES enable people to do things useful things NOW in
a way compatible with their other activities... easier said than done,
but still, a priority. 'nough said.

== Faré -=- (FR) François-René Rideau -=- (VN) Уng-Vû Bân -=- ==
Join a project for a free reflective computing system! | 6 rue Augustin Thierry
TUNES is a Useful, Not Expedient System.               | 75019 PARIS     FRANCE -=- Reflection&Cybernethics ==