psqt00south-1.gif (2282 bytes)

Software Dimensions and The International Institute for Software Testing

Present

PSQT/PSTT 2001 East

Win a Free Admission to PSQT/PSTT 2001 East

spacer.gif (849 bytes)
psqtleft.gif (1883 bytes)
spacer.gif (849 bytes)
calls.gif (2148 bytes)
spacer.gif (849 bytes)
Registration.gif (2258 bytes)
spacer.gif (849 bytes)
program.gif (2139 bytes)
spacer.gif (849 bytes)
winfree.gif (2459 bytes)
spacer.gif (849 bytes)
committee.gif (2228 bytes)
spacer.gif (849 bytes)
sponsors.gif (1588 bytes)
spacer.gif (849 bytes)
vendors.gif (1302 bytes)
spacer.gif (849 bytes)
archive.gif (1510 bytes)
spacer.gif (849 bytes)
softdim.gif (2173 bytes)
spacer.gif (849 bytes)
bottom.gif (1023 bytes)

 

Wednesday Afternoon Conference Sessions (3:45 - 4:45 PM)

PSQT PSTT
Requirements Management Process Improvement Process Management Quality Management Test Process I Test Process II Web Testing Test Automation

Using the Cost of Quality Approach for Software: A Progress Report

Herb Krasner

Collaborative Test Management (I)

Randy H. Hefner

White Box Testing in the Real World (B)

Oliver E. Cole

Your Site Goes Down – Will it Scale or Explode? (B)

Phil Hollows

Integrated test automation and test generation for non-GUI and embedded systems (I)Martin Gijsen


PSQT Feature Presentation (3:45 - 4:45 PM)

A Progress Report: Using the Cost of Quality Approach for Software
Herb Krasner

Concepts:

Cost of software quality (CoSQ) is an accounting technique that is useful to enable our understanding of the economic tradeoffs involved in delivering good quality software. Commonly used in manufacturing, its adaptation to software offers the promise of preventing poor quality, but unfortunately has seen little use to date. This presentation discusses the rationale and context for using CoSQ, then defines a basic CoSQ approach that differentiates the costs involved with: handling nonconformances due to a lack of quality, activities performed for the achievement of acceptable quality, and efforts to prevent poor quality.  CoSQ implementation issues and suggestions are also discussed.  The proliferation of this approach to eliminate the widespread and debilitating effects of poor software quality is also discussed.

Presentation Outline:

·         The Nature of Software and It’s Quality
·         Why is CoSQ Important Now
·         Adapting CoQ Principles to Software
·         Applying Cost of Quality Principles to Software
·         An Economic Model of Software Quality Costs
·         CoSQ: Data Found in the Open Literature
·        Elements of a CoSQ Program
·         CoSQ Program Purposes
·         CoSQ Programmatics
·         Initiating the CoSQ Effort
·         Quality Metrics Collection
·         Presenting CoSQ Information
·         Improving the CoSQ Program

Biography:

Herb Krasner is an independent software systems management consultant, specializing in troubleshooting problem projects and products, and, in coaching organizations to become excellent producers of superior software.   He has over 26 years of experience in computing/information systems and software engineering in both industrial and academic settings.  He is the founder of the Austin Software Process Improvement Network, Past Chairman of the Software Quality Institute at the University of Texas, and has been keynote speaker at several international conferences. He holds academic degrees in Computer Science from the University of Missouri at Rolla: BS ('73), MS ('75) and ABD/Ph.D. ('79).

Back to Top


PSTT Track Presentations (3:45 - 4:45 PM)

Collaborative Test Management
Randy Hefner

Concepts:

This presentation discusses problems that are common to formally testing large-scale systems, details a collaborative test management approach yielding numerous benefits and overviews a web-based Information Technology (IT) solution that has been applied to various NASA and Navy test programs.

Presentation Outline:

·         Introduction
·         Review current trends in software test engineering
·         Overview characteristics common to large scale DoD and NASA test programs
·         Discuss the need for Information Technology solutions to support geographically dispersed test personnel
·         Explore some of the benefits of adopting IT based test management tools.
·         Explain the TestManager 2000 (TM2000) Tool (free distribution for use on government projects)
·         Conclusions and Feedback

Learning Objectives:

·         Common problems confronting large scale test programs
·         Benefits of applying IT based solutions to test programs
·         A government developed application that aids geographically dispersed test teams Content requested by Military, Joint, and IEEE development standards
·         Success in applying TestManager 2000 tool to various NASA and DoD test efforts

Biography:

Randy Hefner graduated with honors from West Virginia University in 1986 with a Bachelor of Science degree in Computer Science. Mr. Hefner was hired by Intermetrics (Titan Systems Corporation, AverStar Group) in 1994 to lead IV&V computer infrastructure and tool development activities for AverStar contracts based in the NASA Software Independent Verification and Validation (IV&V) Facility. His primary focus has been the development of Information System (IS) tools to support, automate, and improve AverStar IV&V processes.  In recognition of work performed, more recently Mr. Hefner was designated as the IV&V Tools & Technologies Project Manager for AverStar.  Mr. Hefner has conceptualized and led the design and development of several IT based tools including the TestManager 2000 tool (previously called the Test Management Database) that is now in use on various NASA programs and by the Navy SPAWAR Systems Center.

Back to Top


White Box Testing in the Real World
Oliver E. Cole

Concepts:

Black box testing has become an accepted practice. All software projects have some number of tests that are run automatically and for which results are checked automatically. This automated testing has proved to be a very cost-effective way to detect many types of bugs. However, studies show that black box testing alone will only uncover about 40% of the bugs in an application. Attention is now turning toward white box testing. White box testing is defined as performing tests with knowledge of the implementation, such as requiring that every line of code be executed at least once, or that each function be individually tested. Studies have shown that when white box testing is added to black box, quality is improved by 20% – a significant amount.  This presentation is designed for organizations interested in starting to use or expanding the use of white box testing.

Presentation Outline:

·         What is white box testing?
·         Basic white box tests
·         Methods of doing white box testing: advantages and disadvantages of each
·         Implementing white box testing
·         Personnel issues
·         Measuring success

Learning Objectives:

·         What white box testing is and why it is important
·         How to incorporate white box testing in their organization
·         What the most important white box tests are and what problems they solve
·         One simple white box strategy that can significantly improve testing
·         Issues to consider when doing white box testing

Biography:

Oliver Cole is the president of OC Systems, Inc., a software company that provides advanced software development and testing tools. He has over twenty years of experience in developing and testing software. Before founding OC Systems, Oliver Cole worked on a number of high reliability real-time systems for the U.S. military, including a sonar-processing system. He has extensive hands-on experience developing large-scale mission-critical software for a variety of organizations, including commercial and international interests. Oliver Cole speaks regularly at industry conferences and has had a number of articles published; most recently, "White Box Testing" in the March 2000 issue of Dr. Dobb's Journal.

Back to Top


Your Site Goes Down – Will it Scale or Explode?
Phil Hollows

Concepts:

Attendees of this presentation will learn how testing early and often will allow you to improve product reliability and how carefully tracking and measuring your projects will improve your productivity and help to minimize risk.

Learning Objectives:

From server transactions to network capacity, the one truth about today’s applications is that you have to be prepared for demand to blow through the roof. In the session you will learn about:

·         How testing early and often will allow you to improve product reliability
·         How carefully tracking and measuring your projects will improve your productivity and help to minimize risk
·         How can you design applications and networks with scalability in mind?
·         Are some technologies, OSs or protocols better than others?
·         Is outsourcing more flexible? 

Biography:

Phil Hollows joined RadView in October 1999, bringing with him over 12 years of experience in software engineering and management consulting. Phil previously worked as the Technical Director for Kronos, Incorporated, where he was responsible for technical strategy and implementation for the Timekeeper Systems Division. Prior to that, Phil was the Managing Consultant in New York for the executive information system consultancy Metapraxis, Inc. Phil had previously worked for Metapraxis in the UK as both a developer and consultant, moving to the US in 1993. Phil holds an MA in Physics from The Queen's College, Oxford, England, and a Certified Diploma in Accounting and Finance from the UK's ACCA.

Back to Top


Integrated Test Automation and Test Generation for Non-GUI And Embedded Systems
Martin Gijsen

Concepts:

Test automation for non-GUI and embedded systems brings different challenges than does test automation for GUI or web applications, mostly because of the different interfaces. But they do share several other complications, such as shortage of technical testers and maintenance sensitivity of the testware. This presentation will indicate how the TestFrame method for test automation can be applied to these systems to handle the complications. It will also describe how (requirements) modeling can integrate with it to generate a high quality automated test set.

Presentation Outline:

·         How to prepare for test automation
·         The TestFrame method for test automation
·         Requirements modeling and test generation
·         Conclusions

Learning Objectives:

For non-GUI and embedded systems:

·         How to check what testing can and should be automated
·         How to prepare for test automation early in the project (requirements phase)
·         What roles there are and what skills they require
·         How the 'action words' from the TestFrame method for test automation can prevent    several common test automation problems
·         How to model requirements and then generate a good, action word based test set from models

Biography:

After a university study in computer science, Martin Gijsen has been working as a professional software engineer for four years and a test automation consultant for three years. In early 1998, Martin became involved in test automation for unit testing. In late 1998, he was part of a project that started black box testing of telephone switches, which are huge embedded systems. He is currently working on extending the resulting test automation platform to cover more features of the switch. The material for the presentation comes from almost three years of experience with test automation for non-GUI and embedded systems, contributed by several consultants (test automators / software engineers). The experience exists of providing complete test automation solutions for complex systems of several clients, including telephone switching systems and multimedia devices.

Back to Top