[om-list] Mathesis -- was Re: Language

[Tom and other Packers]

Mark

    I'm just dying to make MPL.  It's like a physical hunger.  I'm driven to
make it by an uncontrollable appetite.  This has been on my mind constantly
for the past weeks and months.

    Can the two of us talk about our two designs some time, to see how
similar they are in detail, to see how easily it might be to collaborate and
satisfy both of our desires in the same language?  Any one else may also
join in the discussion, especially as it may concern the whole OM group and
goals.

    Okay, venting time -- but it has a sublimational justification as an
object lesson.

    At this particular moment, instead of being hungry, I'm actually
wallowing in self-pity because my fellow citizens of Farmington have denied
a "certain positive" because they feared an "uncertain negative".  In other
words, something bad might have happened to them, so they denied the
certainty of something good happening to somebody else.  Is this not the
essence of selfishness and ignorance?

    I'm taking it personally.  I know they probably didn't intend anything
personal by their decisions and arguments, but I'm taking it personally,
because it affects me personally.  They will have cost me hundreds, if not
thousands, of dollars.  And it all could have been solved, to the benefit of
all people involved, if people would simply rely on truth and logical
thinking, such as what I intend mathesis and its implementations to provide.

    Let me explain.

    My dad wanted to finish the upper story of his detached garage, to
convert it into a dwelling place, in which I could have lived after getting
married.  We had a public hearing in Farmington because this plan does not
quite fit the current zoning restrictions.  My dad was applying for a
"variance".  The variance was to simply extend an arbitrary four feet
definition into an hundred feet definition.  As the law now stands, the
detached garage could have been considered part of the same structure as the
house if it were less than 4' from the house; we wanted to loosen that
restriction a little.  That's it.  And yet the ignorance and emotionalistic
associations of mere mortals will change that purpose into something bigger.
At the public hearing, some of our neighbours offered completely irrelevant
objections to the motion -- which I'm confident you would agree were
irrelevant if I were to explain them in excruciating detail, so I won't --
and the quorum of three (and the one lady who came in late) voted on the
decision, and based on their own irrelevant objections, denied the
application.

    (There was one member of the quorum who did not support the denial, but
... the other three were enough to end the issue.)

    This is just a minor thing, in the grand scheme of things.  But it is an
immediate and personal illustration of the damage that ignorance can cause.

    It is my goal in life to fight ignorance wherever possible.

    I propose a tool, not unlike the "new tool", or "Novum Organum", that
Bacon proposed in his famous essay of that title.  The tool I propose is
called Mathesis, which loosely means two things in the language of
Descartes: "a mathematics-like structure or language" and "a method".
Mathesis will be a tool or language which will facilitate the human mind in
reaching any general or special, real-world decisions, based on common or
other arbitrary knowledge, in a perfectly logical manner.  Any logical
thought process already possible in the human mind can and will be
represented and aided by this proposed language.  (Written proof to come
later.)

    After the language is designed, its use can be automated.  Once its use
is automated, and other more user-friendly forms of representation are
devised for front- and back-end interface with users, (such as natural
spoken language), this method/structure called mathesis can and will be put
into common use, in ordinary, arbitrary situations, for the benefit of those
situations and all people involved.

    Think of this as the ultimate practical AI.

    Just as logical arguments were put forth by great men at the end of the
Age of Enlightenment, in the form of simple prose essays, for example Thomas
Pain in "Common Sense", and "Publius" (Hamilton, Madison, Jay) in "The
Federalist Papers", with the purpose of convincing the ignorant masses of
the value of a new Constitutional system of government, so will the writings
in the new language convince the ignorant masses in our own generation to
follow other valuable truths for the benefit of individuals and whole
societies.

    Am I delusional?  Feel free to tell why.

ciao,
tomp

----- Original Message -----
From: "Mark Butler" <butlerm at middle.net>
To: "Tom and other Packers" <TomP at burgoyne.com>
Sent: Tuesday, August 28, 2001 9:47 AM
Subject: Re: Language


Tom,

 It sounds like you have lots of interesting things going on.  My
programming
language (should it be implemented) would probably work pretty well for what
you describe, but I haven't convinced myself that it is worth my spare time
for a decade to write it.

I usually call it the Epic programming language, and it is a declarative /
procedural hybrid.  The key characteristics are:

1. Primarily declarative
2. Set oriented (by default all results have arbitrary cardinality)
3. Formal hierarchical namespace
4. C++ style procedural language
5. Distinguishes between logical classes and physical data types
6. Language level concepts of list, string, set, expression,
   predicate, etc.

Example:

x = 3;
y > 0;
z = 5;

x^2 + y^2 = z^2;

print y;

In a compiled environment, this would reduce at compile time to:

print 4;

In other words, the language uses an internal symbolic constraint solver
similar to the one in Mathematica that it uses to solve problems
non-procedurally.  If the y > 0 constraint were left out, it would print 4,
-4.

also

> print employees[salary > supervisor.salary].name;

John Smith


  The procedural language is used much like a high level version of C++ to
implement process features like user interfaces where necessary.

The general idea is to implement a language that can do operations as high
level as LISP or Prolog at run-time but can also compile operations like
vector arithmetic or data compression into machine code as efficient as that
generated by C/C++ compilers.

Of course, it is much more likely that a first version would be interpreted
only, because it is much easier that way.

 - Mark



>     (Mark, how similar might this programming language be to the one you
> want to write?  Probably not very.  This one would be rather high-level,
> like Lisp and Prolog.  But, ... I may ask you if you'd like to think about
> helping me write a language some time.  It would probably be called MPL:
the
> "Mathetical Programming Language".)


--
Mark Butler        ( butlerm at middle.net )
Software Engineer
Epic Systems
(801)-451-4583