WICSA1 -- Working Group Plans

You are viewing a version of this page that has been archived via the Internet Archive Wayback Machine for the ICSA conference series history.
← return to the ICSA homepage.

22 February - Monday

Interoperability and Integration
Morven Gentleman and Mary Shaw

Note: In each case, the author is asked not to present the paper as such, but to use the paper and other relevant experience (including reading of the other papers in the session) to comment on these questions:

What are typical examples of serious interoperability/integration problems?
What are the most significant general classes of interoperability/integration problems the software development community faces today?
Which of these require research breakthroughs?
What would you propose as good candidate research topics?

Plenary session:

(9 min) Introduction to session, including the questions above
(12 min) Butler, based on experience report
(12 min) Gruhn, based on experience report
(12 min) Shaw, based on position paper
(15 min) Discussion

Group session 1 (afternoon):

(12 min) Gentleman, based on experience report
(12 min) Jaktman, based on experience report
(12 min) Kuusela, based on experience report
(12 min) Weil, based on experience report
(12 min) Bourgois, based on position paper
(30 min) Discussion in preparation for evening session

Group session 2 (evening):

Refine description of major problems
Identify promising research opportunities, describe them in enough detail to make sense, and try to organize them into clusters.

Analysis and assessment of software architecture
Rick Kazman and Naftaly Minsky

Plenary session

Presentation of paper by Magee, Pree, and Atlee

Working session For the working sessions we will spend the first session trying to decide what one can analyze an architecture for (i.e. to produce the beginnings of a state-of-the-art report on the spectrum of analysis techniques and their objectives)

We will spend the second session discussing what information needs to be represented in an architecture to enable its analysis for those properties that we identified in the first working session.

In both sessions we will distinguish between what is being analyzed (the architectural description (AD) or the underlying system (S)), and between the nature of the analysis appropriate in each case, as follows:

1: Analysis of the architectural description (AD) for:
a: consistency
b: performance
c: modifiability
d: reliability
e: security
f: liveness
g: . . .

2: Analysis of the underlying system (S), with respect to the AD.
a: verification that the AD is satisfied by S.
b: enforcement of AD on S
c: Synthesis of aspects of S according to AD..
d: (anything else??)

Models of Software Architectures
Don Batory and Ric Holt

Format: presentations will be 20 minutes each, with 10 minute discussions

Authors will answer questions specific to their work, plus some general questions:

What problems does your approach solve?
When should it be used?
Why should it not be used?
Why should practitioners be interested?
Why should academics be interested?
If your approach is formal, why should practitioners be interested?
If your approach isn't formal, how does your approach contribute to the Science/principles of SA?

Plenary Session

Stuurman, "Software Architecture and Java Beans"
Hirsh, "Modeling Software Architectures and Styles with Graph Grammars and Constraint Solving"

Working Session #1

Canal, et. al. "Specification and Refinement of Dynamic Software Architectures"
Medvidovic and Rosenblum, "Assessing the Suitability of a Standard Design Method for Modeling Software Architectures"
Hoffmeister, et al., "Describing Software Architecture with UML"

Working Session #2

Discussions/themes raised in the plenary and WS#1 will continue in this session.

23 February -- Tuesday

Product Lines
Christine Hofmeister and Henk Obbink

Plenary session

Introduce two discussion topics, define terms (5 min)
1. Evaluation of Product Line and Reference Architectures

A Domain Engineering Method
Felix Bachmann, Len Bass, Gary Chastek, Sholom Cohen, Pat Donohoe, Linda Northrop
(evaluation as an integral part of the domain engineering of a product-line architecture)
Flexibility of the ComBAD-architecture
N.H. Lassing, D.B.B. Rijsenbrij and J.C. van Vliet
(using SAAM to evaluate a general-purpose arch: discuss which qualities can be evaluated for a general-purpose arch)
discussion (10 min)

2. Evolution of Product Lines

"Evolution and Composition of Reusable Assets in Product-Line Architectures: A Case Study
Jan Bosch
(how does evolution affect product lines)
discussion (10 min)

Group session

1. Evaluation of Product Line and Reference Architectures panel (1 hr -- Each panelist gets 10 min to give an overview of their architecture.)

Seidersleben experience paper (a reference architecture for business info systems)
Savigni technical paper (a reference architecture for monitoring and control systems)
Pronk experience paper (a product-line architecture for medical diagnostic imaging systems)
Remaining 30 min for questions & discussion on the topic:
- What's special about evaluating a product-line or reference architecture; how is it different from evaluating an architecture for a particular system?
- How were the panelists' architectures evaluated (or how could they have been)?

2. Evolution of Product Lines (1/2 hr)

Dealing with systematic evolution of productlines is a very difficult topic. Many of the changes during the lifetime of a productline could have a significant architectural impact. Analizing the impacts of the following changes:

Change in Domain
Change in Function
Change in Technology
Change in cost structure
Change of interfaces

could be very useful. The goal of this working session is to investigate the practices that exist to deal with these changes in a systematic way.

3. discussion, preparation of summary points

Style WG
Dewayne Perry and David Rosenblum

Plenary Session

Architectural Styles as Adaptors
Don Batory, Yannis Smaragdakis (University of Texas at Austin), and Lou Coglianese (LGA Inc.
Attribute-Based Architecture Styles
Mark H. Klein, Rick Kazman, Len Bass, Jeromy Carriere, Mario Barbacci, Howard Lipson
Modelling Software Architecture Using Domain Specific Patterns
Jan Peter Riegel, Christoph Kaesling, and Martin Schuetze

Style WG - Discussion

We will cover the questsions:

What can we do with styles and patterns?
What cant we do with them?
What research questions should we look at?

We will summarize our findings for the working group as a set of research issues and goals - ie a research agenda for architectural patterns and styles.

Techniques and Methods Working Group
Frances Paulisch and Will Tracz

Plenary Session

This WG will focus the processes and players (stakeholders) that support the creation, use, and maintenance of software architectures.

1) State the scope of the working group as described above.
2) Summarize the relevant points in the 5 experience papers and 1 paper that were deemed by the PC to fit in our WG.
3) Propose/reuse a framework of stakeholders and issues (using Boehm's Table 1 as a start) and sketch out a "super-process" model/picture that shows where styles, models, ADLs, analysis, interoperability, product lines all fit together.
4) Introduce the objective of our working group session (see below).

Working Group Session:

During our 2.5 hours we will address 5 or 6 carefully prepared and partially answered questions related to:

What academia can do for industry, and what industry needs to provide in the next 2 to 4 years to advance the state of the practice and art in support of the creation, use, and maintenance of software architectures.

Note: this includes introduction of new tools, processes, training, etc.

Dewayne E. Perry - This information last updated January 1999