🚀 go-pugleaf

An Open Letter to Richard O'Keefe, 

   I`m sharing this with the net because I think
we need a strategy for dealing with the problem
Richard brought up.

REASONING:

It is tempting to argue with people who have
not thought things through,  or are driven by
myths,  etc.  However,  mere talk doesn't work because 
things like "no one uses Prolog" has been repeated
so many times it is considered to be gospel.

DEMONSTRATION:

I have had good experiences *demonstrating* to
people that they may be mistaken (by "good experience"
I mean that out of every ten demonstrations I do, 
one company gives serious consideration to using Prolog :-)

Would it be possible for you (Richard) to write
(or adapt) a small program which demonstrates some
of the diagnostic capability and then *show* them
what maintenance of such a program would entail.
With the proper design,  I imagine a great deal
of maintenance might involve editing sets of facts.

Then challenge them to write a prototype in C with
exactly the same capability. Once they see the 
minimum 10X increase in source code,  they will
have to rethink their "maintenance" decision.
(for instance,  in a C program,  the driving facts would  
more than likely require some kind of parser to be built
-- Eg.they would be seen as a kind of "data" which is
removed from the "program" - a distinction we can
usefully merge in Prolog).

If Yuval Lirov is listening,  perhaps he can help you
out with the diagnostic systems he worked on at AT&T?

INTIMIDATION:

Or maybe we *should* take the argumentative approach!
But do it right. Tell the Prolog Vendors Group who
this mysterious company is and let them be inundated
with information about worldwide Prolog applications.

FEAR:

Once,  in the depths of despair,  I came up with what is
rapidly becoming my favorite strategy of all:  If you
really think that company C is missing a strategic advantage
by not using Prolog for their application,  then find
other companies in the same industry and provide
the demonstration to them!  Now you have a 
situation in which all three companies can
potentially use this technology and they all understand
that whoever does take this step might have a
substantial advantage -- e.g. that the conservative
approach might not be the best for them to take.

Actually,  if one of the other companies starts using
Prolog,  this strategy might be thought of as REVENGE.

and
LOATHING:

Perhaps these are not the best people to be using
a Logic Programming language.



One last thing. Get permission from the company to post
their name to this newsgroup. We promise not to jeer at
them,  its just that they may need to see resumes from
people with Prolog experience to understand that the manpower
for maintenance is there.  I can think of at least one Prolog
programmer that would like to spend some time in Austrailia.

Richard,  the best thing that *you* can do is to produce
students who can give good demonstrations of Prolog
technology and are not afraid to influence the companies
they will be working for after they leave the University.



            - Peter Reintjes        pbr@quine.sofla.ingr.com

_____________________________________________________________
| As far as I know,  these *are* the opinions of the company |
| I work for. That's why they hired me ;-)                   |
|------------------------------------------------------------|