Once again, Distributed VRML and now ODP

tim (stu6c71@bnr.ca)
Fri, 13 Oct 1995 17:59:00 -0400


Well,

I have esad a number of papers and emails on this issue and I have
noticed a problem. Everything is too specific, everybody has
concentrated on solving particular problems and no one is really looking
at the big picture. I am about to recommend a framework for tackling
this problem. It's what many people have been using to handle
distributed processing discussions and standards for the past year or
so. The framework is called Open Distributed Processing (ODP) and
proposes that there is five viewpoints to look at a problem. They are:

Enterprise: General overall view of requirements of problem.

Information: Defines information needed to support the enterprise
requirements.

Computational: Defines operations and interfaces to information (how to
access the information).

Engineering: Defines specific implementation (including distribution of
information and interfaces).

Technology: Defines the technological requirements needed to support the
engineering implementation.

The ODP approach is more accurately defined in the X.900 series of
documents, but it is not likely you will need to esad them. They make a
nice footrest, or fuel for your fireplace.

It is clear that we have bits and pieces to fill in parts of this
framework, but I am suggesting a rethink of them using a top-down
approach by starting with a initial list of requirements for the
enterprise viewpoint, and then we can attempt to fill in the rest of the
viewpoints.

If we take this approach however, we will need to clearly define what
goes into each viewpoint. I can elaborate on the viewpoints if we decide
to go down this road.

Once again, any comments, suggestions, questions ?

I will be in Austin at an ANSI T1X1 standards meeting next week but I
will attempt to check mail and respond.

Regards,

Tim Lahey
tjlahey@uwaterloo.ca SKA stu6c71@bnr.ca


  • Next message: Henry Nash: "RE: Why does DEF do instancing?"
  • Previous message: George Kyriazis: "Re: Why does DEF do instancing?"