Program Chair's Remarks

Note: These remarks and associated ViewGraphs in PowerPoint format framed the three technical plenary sessions of QW'97. The commentary here summarizes the comments made at QW'97.

Introduction to QW'97

Statistics about the Attendees

800+ people registered at QW'97 in various categories

375 companies represented
31 Vendors at the QW'97 Exhibits

34 US states represented
17 European countries represented
4 PacRim/Latin America countries

About the QW'97 Advisory Board

The QW'97 Conference papers and presentations are all reviewed by the distinguished International Advisory Board. Composed about 1/3 from Europe and 2/3 from the United States, the Advisory Board members represent Industry, Government and Academe.

Special recognition to Advisory Board members who have agreed to monitor the presentations as Track Chairs is very important:

• Mr. William Bently (Bayer Corporation), Applications Track Chair
• Dr. Robert Birss (Intuit), Management Track Chair
• Dr. Boris Beizer (Analysis), Technology Track Chair
• Prof. Leon Osterweil (University of Massachusetts), QuickStart Track Chair
• Mr. Otto Vinter (Bruel & Kjaer), Vendor Track Chair

Why Is Quality In The Marketplace Important?

Quality: THE Distinguishing Factor in Software

People in the quality management field understand that the ultimate product distinguishing factor is perceived and internal software quality. Everyone in the QA/QC community has hoped for this for many years, because it will act as an impetus in our field, but it is clear from recent events that NOW is the time!

The Quality/Test Marketplace: ~$1B by 2000

From various companies' business plans and from sources through the industry the "experts" say that software testing and quality will be a $1B/year market by the year 2000.

Growth Areas:

We know that there is tremendous growth in these areas, but growth is not limited to just these:.

Industrial Process Management
WWW/Internet
Java
E-Commerce
Y2K
...more...

Quality Week #10 Reflects These Focuses

QW'97's theme Quality in the Marketplace reflects there interests and goals.

What Can Be Done About Software Quality?

Many Tools and Technologies Available

There are many tools and associated quality techniques that can be applied, but very often they are applied unevenly and ineffectively. The chart shows some 20 different QA/QC methods arranged roughly according to their position in the software life cycle:

• Some are inexpensive to use, some are moderately expensive to use, and some are very expensive to use (vertical position on chart).
• Some are lightly used, some moderately used and some heavily used (size of dot).
• Some are ineffective, some are moderately effective, and some are very effective (color of dot).

The message of this chart is that it appears that the most-widely used methods are the ones that are least effective, and vice versa. Apparently people choose quality control methods that "feel good" or are "inexpensive to apply" over those that do the job right!

Best & Brightest Attracted to "The Problem?"

Yet we know, looking at the titles of those attending QW'97, that the best and brightest apparently ARE attracted to the problem area. That is very, very good news!

Errors Abound!

And, the other message is that in spite of a variety of methods, and the services of the best and brightest, errors still abound!

Clearly, there is work for the future.

QW'97

We hope QW'97 reflects these trends, and addresses these issues.

Unsolved Problems In Software Quality

The best of our efforts notwithstanding, Software Quality Technology has many gaps. Here are a few of them.

Automatic Test Data Generation

In spite of years of work, there is still no practical automated test data generation.

Practical Data Flow Coverage

Even though it has been demonstrated to be effective in the labs, data flow coverage hasn't been made available by any commercial products.

Automated Semantic Update (Branch Merge)

Even with the major advances in sophistication of configuration control methods, there is only minimal assistance for merging two "branches" in a source tree.

Software Reliability Measures

The technology is reasonably mature, but you can't find "reliability figures" for software products that are realistic.

Self-Correcting Software

We have the technology for software products to repair themselves, but this kind of technology hasn't made it to the market.

Continuous Testing

Mentioned last year in Prof. Osterweil's keynote talk, the notion of continuous testing, while attractive, does not seem to be catching on.

What Will The Future Bring?

Y2K The Wakeup Call?

The Year 2000 Problem may be the general wakeup call for software quality issues!

Much More Innovation

We can expect to see a LOT of innovation in the coming years in the software quality area. It will be exciting.

Phenomenal Growth in the Quality Marketplace

As mentioned, there is likely to be a BIG growth in the software quality market!

Life-Cycle Quality Management

And, more and more, we will see software quality as a MAIN thrust overall in the life cycle.

QW/Europe, November 1997, Brussels, Belgium

We will expand Quality Week to Brussels in November 1997.

QW'98, May 1998, San Francisco, California

Be sure to schedule May 1998 for QW'98 here in San Francisco!