(Attn -- AMG) revision of the Abstract Models for Module Processing and Attachments

I revised my earlier submission on Abstract Models for Module Processing
and Attachments in order to facilitate comparisons with SOAP's model for
Module Processing.  It now includes a a discussion with each concept to
illustrate how it compares with SOAP.  I also made some minor changes in
response to earlier feedback, and I included an "optionality" operator
to model behavior similar to the mustUnderstand feature in SOAP.

From an Abstract Model perspective, my intention is to introduce a
clean, more expressive way of describing module processing than the
model in SOAP -- one module (actor) per block which consumes the block
and an optionality indicator.  I was inspired partially by things I had
seen elsewhere in the AM document (Figures 2.1 and 5.2).  Composition,
sequences of module applications, and the ability to target the same
block at multiple modules are the most important concepts. Parallelism
is probably less important (but implied in Figure 5.2?).   Most of thee
cocepts are orthogonal to some extent and we can pick and choose.  That
is why I wrote the concept list as I did.  A different set of concepts
would yield a somewhat different model.

--
Mark A. Jones
AT&T Labs - Research
Shannon Laboratory
Room A201
180 Park Ave.
Florham Park, NJ  07932-0971

email: jones@research.att.com
phone: (973) 360-8326
  fax: (973) 360-8970

Received on Thursday, 8 March 2001 16:50:36 UTC