I never sent out the "introduction" about myself that was suggested,
so perhaps I should say where I am coming from. I am an astrophysicist
with a history of computer language use, and scientific software
development, that started with the dark ages of FORTRAN but quickly evolved
through Algol, PL/I, and variants of Pascal and Modula/2 until I switched
to Oberon a few years ago for my purely personal programming. At that time
I was project scientist for a big C++-based effort that has consumed
several million dollars in personnel costs to produce object-oriented
software of such complexity that only the (third) project leader can use it
well. I left that project in disgust, and started up my own research project
to design and prototype a scientist-friendly system for programming and
data analysis. The heart of the problem is the management of complicated
objects with many state changes in a way that is both technically
correct and easily manageable by scientists who must continually adapt
software to the new problems of the day. Initially Oberon V4, and later
Oberon/F, were pretty good tools for rapid prototyping. Right now we
have a lot of miscellaneous tools, with the most interesting perhaps being
a system where both crisp and fuzzy numbers can be used as basic data types
in a linear algebra system, providing a new way for studying error
propagation in complex, non-linear computation.
However, it has become glaringly likely that Oberon is going to go nowhere.
The contrast to the explosion of good work (if you ignore the hype from
certain commercial sources) based upon Java is overwhelming. As a language
Java is sufficiently Oberon-2-like, but with a few important additions
and an even cleaner class extension system, that it is an example of
what could have been done with Oberon if people with the right vision and
energy tried to do similar things. It is amusing to me that indeed
the fast on-the-fly "link/compile/run" possibilities were utilized by
Kistler and Franz at U.C. Irvine to do faster "applet technology",
based on Oberon, years before the Java bandwagon began. But as usual
with Oberon, there is a largely academic flavor, and inertia, whereby
the hare (Oberon with Juice technology) seems to be rapidly overun by
the turtle (Java with simple byte code technology and the Java Virtual
Machine).
I suspect that most of you still "listening" to this forum also "listen"
to comp.lang.oberon, so you noted the recent postings where
"innocents" asked for a list of the "shrink-wrapped" applications
done in Oberon. We all know that there are none as yet. The attempts
at Oberon systems, like Oberon/F, do not count as "shrink-wrapped"
applications. Neither do add-on tools for data base developers.
The proof of the pudding is in the applications.
So, with great reluctance I conclude that the house of Oberon is
largely empty, and wonder if there is anyone will remember to turn out
the lights after I am gone. As has been noted, Oberon microsystems
has not lived up to the hopes we had for it. Almost two years ago
they were hoping for a Unix platform "real soon now", and we have
not heard about it. As noted, it has been almost a year since
there was an issue of the Oberon Tribune. There documentation
still is terrible unless you can memorize everything you have ever
seen driving down hyperlink road. They could have fastened on
Juice technology, and I heard a rumor that one of the implemetations of
it would be in the Oberon/F platform, with indications that it should
have been announced already, but again we hear only silence. Or
the announcement that it is available for the Mac platform of Oberon V4.
If I/we are wrong, I put in a strong plea for others, particularly
Oberon microsystems, to more clearly state their hopes and plans.
We could use some hype, but that takes a boldness that seems to being
missing from the Oberon world in general, and the Oberon/F world
in particular.
Is there any hope for the house of Oberon/F? Or are we "listening" in
the wrong places.
Bob Hjellming
Robert M. Hjellming | Internet: rhjellmi@nrao.edu | __
National Radio Astronomy Observatory | http://www.nrao.edu/~rhjellmi| \/ o\
P.O. Box O (US Mail only) | | /\__/
1003 Lopezville Road (UPS,FedEx. only)| Telephone: 505-835-7273 |
Socorro, NM 87801-0387 USA | FAX: 505-835-7027 | AMDG