Software Development for Intranet Applications
by Dan C. Marinescu
Computer Sciences Department
Purdue University
West Lafayette, IN 47907

Full Paper in Postscript File


Tutorial Proposal for the IFIP Workshop on Dependable Computing


There is very little doubt that network computing will be the
driving force in the software development for the coming decade.
Several "killer applications" including electronic commerce,
telemedicine, and distance learning will define the actual requirements
for network computing and will justify the enormous investment in  this
area.

Fist we will address some of the issues in electronic commerce and
discuss
Java Commerce and the Open Network Environment of Netscape. The novel
approach to software distribution and maintenance supported by the
"cassette mechanisms" will be presented.To conclude the first part of the
tutorial we will outline some of the challenges of Intranet Applications,
namely fault tolerance, quality of service and security.

Next we will cover the enabling technologies for the development of
Intranet applications. Fist we overview the use of Java as a network
programming language, then we'll survey the new features of Java 1.1
and discuss introspection, reflection and object serialization. The next
topic is the Java Server Architecture and servlets. Last we will discuss
CORBA and CORBA/Java environments and introduce basic concepts like
Object Request Broker, ORB, the Internet InterOrb Protocol, IIOP, the
Interface Description Language, IDL.

The third part of the talk will cover our experience in developing
a metacomputing environment. We will start by describing a problem in
computational structural biology and some of the design requirements
for the Bond system and compare this software project with other
software development projects we have been involved with, the
development
of a real time data acquisition and analysis system for nuclear and high
energy physics and the development of parallel programs for phase
refinement and extension in structural biology.


Dan C. Marinescu is a Professor of Computer Sciences at Purdue
University.
He was the chief architect of data acquisition analysis system used
for the study of super heavy elements. He is currently the Principal
Investigator of a Grand Challenge project sponsored by the National
Science Foundation for the application of parallel and distributed
computing in computational structural biology. He has published more 
than 90 papers in the areas of scientific computing, parallel and 
distributed computing, and performance evaluation.
--
Dan Cristian Marinescu  Voice/FAX: (765) 494 6018/0739
dcm@cs.purdue.edu  http://www.cs.purdue.edu/people/dcm
CS Dept. Purdue University, West Lafayette, IN, 47907, USA;