From: curtis@stsci.edu (Gary Curtis)
Sender: owner-trans@stsci.edu
To: rpg@steam.stanford.edu
Cc: trans@stsci.edu
Subject: LUGM99 Paper Overview
Date: Fri, 30 Jul 1999 15:32:13 -0400 (EDT)

Hi,

This message contains a short overview of a paper we intend to
submit for LUGM '99. The paper discusses an innovative approach
to application development that we are using in the re-engineering
of one of the key software systems used to prepare observations
for the Hubble Space Telescope at the Space Telescope Science
Institute.

Thanks,
Gary.
-----

The Constraint Sequencing Infrastructure (COSI) is an extension to
the Common Lisp Object System (CLOS) which supports a constraint
based object-oriented programming model. The CLOS Meta-Object
Protocol (MOP) is used to specialize the behaviour of CLOS to
support this model.

COSI has three core capabilities that support the constraint
oriented paradigm: relationships, object management and constraint
sequencing.

Relationships simplify the task of working with a network of objects.
A relationship is a virtual bi-directional pointer between objects
with arbitrary cardinality. By using a relationship, all objects in
that relationship are aware of related objects, and the relationship
is automatically maintained by the relationship manager as objects
are added and removed.

The object manager is responsible for managing the instantiation of
objects as they are created by constraints. Constraints specify the
complete set of objects that should exist given some current state.
If a constraint re-runs and creates an object that was instantiated
with the prior invocation of the constraint, then the object manager
returns the same object, rather than a new one. If a constraint re-
runs and does not create an object that was created with the prior
invocation then the object manager will delete the object, removing
it from relationships and resetting its state.

A constraint is a specialized method which will be automatically 
re-run by the COSI infrastructure whenever any of its input values
change. Input values are any of the object attributes that are 
accessed by the constraint, and which are therefore assumed to
alter the processing within the constraint.

Whenever a state change occurs those constraints which depend upon
that state are added to a propagation queue. When the system is
queried a propagation cycle runs ensuring that the state of the
system is consistent with all constraints prior to returning a value.

Our paper will provide a detailed description of this infrastrucuture
and its implementation and use.

-----