Artificial Intelligence work has fa

Artificial Intelligence work has faced issues that have only
recently been of serious concern to the database and
programming language worlds, and there is some quantity of
technology that might be transferred here. Some examples of
mechanisms in this category are automatic procedure invocation
(views, triggers, demons), complex type systems (multiple
inheritance of properties, etc.), complex integrity constraints, and
data inference. The database work has in the past been
concerned with issues less important to AI, e.g., large,
concurrently-accessed, secure, distributed, long-term, efficient,
highly reliable database management systems. The idea of
building AI systems on top of database systems was suggested
more than once. This would allow AI systems to take advantage
of the lower-level work in databases, for applications for which
this is desirable; however, often the higher-level type mechanisms
influence the lower-level implementation. The programming
language work has been involved with some of the database
concerns as well as others, e.g., more heterogenous data and
operations and ease of program development.
A point called to mind by comparison of areas but not always
made clear at the workshop is that combining many of the ideas
presented is much more complex than a simple juxtaposition of
the implementations or syntax. The high-level system models
and low-level system implementations just mentioned are just one
example of this. Much of our work on the Cedar database is
concerned with integration issues of this kind.
Acknowledgements
Mark Brown and Nori Suzuki are co-authors of the database
system that is now operational. Peter Deutsch and Jerry Popek
assisted us in the .design of the system. Mark provided
comments on an earlier version of this paper. I would also like
to thank John Smith, Dennis McLeod, Steve Zilles, Richard
Brodie, Joachim Schmidt, and the other spirited hikers for their
absorbing conversations enroute to Stormy Peaks Pass and
Cirque Meadows. John Smith deserves particular commendation
for remembering a compass.