A Summary of Project so far [Re: Kernel LISP - how low down can it go?]

Patrick Logan patrickl@servio.gemstone.com
Thu, 22 May 97 08:51 PDT


>>>>> "Henry" == Henry G Baker <hbaker@netcom.com> writes:

    Henry> In the case of OODB's, you have multiple users trying to
    Henry> access the same objects.  If the objects are functional,
    Henry> then they can be shared without any restrictions, while if
    Henry> they are mutable, then each user is building up a different
    Henry> view of what the OODB looks like.  At commit time, the
    Henry> multiple users lay their cards on the table, and the winner
    Henry> gets to commit, while the losers have to roll back.

Which is one reason I am finally getting around to learning Haskell, a
lazy, pure functional language. I'd like to know more about the
potential of programming with non-mutable data, and then try to
understand the implications in a distributed multi-user world.

Interestingly, a lot of applications avoid mutation at the conceptual
level, but not at the implementation level. Think of business
transactions and records. Conceptually, these items are *never*
mutated.

Another area where Gemstone is used a lot is in manufacturing. One
part of the system controls the operations... creating things. Then
facts and artifacts of these operations are stored off to be reported
to management and used to further control operations. A lot of
creating and no destroying, conceptually.

I am leaning with some of Henry's suggestions that a Lisp machine for
the 21st century can and should push against the conventions of the
20th century Lisp machine.

-- 
Patrick Logan                 mailto:patrickl@gemstone.com
Voice 503-533-3365            Fax   503-629-8556
Gemstone Systems, Inc         http://www.gemstone.com