SAST Stockholm Q1-möte 2006-02-16

SwedenSTAR

EuroSTAR 2005 i Köpenhamn hade 830 deltagare vilket var rekord. 169 kom från Sverige – flest av alla. Åtta svenska talare fick vara med i programmet. SAST är glada att sex av dem har tackat ja att hålla sin presentation på Q1. Det har blivit en tradition att första SAST-mötet för året erbjuder medlemmarna att ta del av föregående års bästa svenska presentationer i ämnet Software Testing, Analysis & Review. En unik chans som Du inte kan missa!

Plats:Bonnierhuset, Torsgatan 21, Stockholm

Mötessponsor

Sammanfattning av föredragen

Sammanfattningarna är hämtade ur EuroSTARs program.

TE1 - Knowing What's Right, Doing What's Wrong

Sofie Nordhamren, St. Jude Medical AB, Sweden

As testers and test leaders/managers, we are convinced that testing adds value to the project and to the product. When assigned a new project, we burn with fervour to do the “right thing” (using the appropriate testing techniques, best practice etc), and show the world how much good testing can provide! However, sometimes we cannot achieve that “right” for different reasons, even if we know how it should be done: which can be really frustrating.

Based on experiences from recent projects, this presentation aims to give some advice on how to get closer to what is “right” and how to add value even if you can not do exactly the right thing. Also, you will be given some thoughts on how to handle the “wrongs” and on how to survive and sustain your vision for what is right even when you cannot achieve it.

Key Points

» How to deal with obstacles that prevent you from running the perfect test project
» How to add value and get as close as possible to what is right even in the imperfect project
» How to deal with the frustration, and motivate yourself and keep doing a god job when “the going gets tough”

W3 - Understanding System Requirements

Anders Claesson, Enea Systems, Sweden

To be able to test a system you need to have a clear understanding of what the system is supposed to do. You need clear, complete, consistent, unambiguous and testable requirements. This is rarely the case in real life why you as a tester must learn how to interpret the requirements and know which questions to ask to obtain a clearer picture of the system from the users point of view.

This presentation will give you some guidelines how to get enough information in order to select the most important Test Cases.

Key Points

» Learn how to analyse requirements and ask the right questions for missing or unclear information.
» Learn how to analyse requirements and ask the right questions for missing or unclear information.
» Learn how to identify the amount of testing required to obtain both sufficient requirements coverage and an acceptable risk level in the finished product.

Experiences of Design Based Testing With TTCN - 3

Conor White, Ericsson AB, Sweden

Experiences in piloting and deploying of standardized test framework based on TTCN-3 in GSM BSC. TTCN-3 was piloted due to the need to improve the design based unit test phase due to poor reliability and the cost of maintaining an own built test framework. TTCN-3 was chosen based on performance and scalability and support for code-orientated style of testing. The solution supported OSE Delta unit/load module testing. Key requirements were support for automating translation of message types between the product and the test product whilst also supporting very short build times. The pilot started with a small-scale deployment. Wider deployment started after improvements in TTCN-3 environment reduced average build times. Integration with run-time analysis tools for code coverage and memory profiling then followed.

Key Points

» The value of a standard scalable test framework in design based automated testing.
» Deployment characteristics of an SQA improvement in the design process.
» The value of combining a solid test framework with run-time analysis tools.”

Improvement By Advanced Data Analysis

Rikard Edgren, Spotfire AB, Sweden

When developing software products a lot of data is produced regarding quality issues, e.g. test results on all levels, test coverage, bug reports, support incidents. This data is very rarely used in the best way possible. There is more to learn about the data than the initial reports delivered by the separate systems.

By looking for trends, patterns, outliers and unanticipated relationships in the different data sources, and by cross-referencing the data sources; a lot of new knowledge can be gained. This analysis will improve the product, and the way you produce it.

Key Points

» How to organise software product data (e.g. bug reports, test results, support incidents)
» Ways to analyse the data
» Ideas on how the results of the analysis can be used to improve testing and product quality

Model-Based Testing Using UML 2.0

Per Blysa, Telelogic AB, Sweden

With UML 2.0 and the UML Testing Profile model based testing can now be integrated into the development process. UML 2.0 provides one modeling language for systems, software and test engineers - improving overall quality of the end product, enhancing productivity and communication, and providing better visibility of progress and traceability.

Using UML requirement models that are automatically transformed to test architecture and sequence diagrams, the test engineer creates the detailed test in the easy to use sequence diagram format. These are automatically transformed to test executables used for early requirement verification, unit testing, regression testing or final system testing.

Key Points

» UML 2.0 and the UML Testing Profile is now ready to used in real projects
» Improves quality, productivity and traceability
» Automates the process of development of test for requirement verification, unit testing, regression testing or final system testing

Verification of Critical Telecommunication Systems

Mats Larsson, Ericsson, Sweden

Critical telecommunication systems are a challenging area within system verification as such system often is designed with a high level of fault tolerance and is expected to operate without failures for long periods of time. This leads to large and complex systems and also a huge challenge for the verification process to ensure their quality. This presentation will focus on both architecture and verification issues of complex fault tolerant telecommunications systems, throughout the presentation real world examples will be used from an existing architecture currently developed by Ericsson in operation throughout the world today. The presentation will cover today’s “State of the Practice” within telecommunication.

Key Points

» How fault tolerance could be built into telecommunication architecture with real world examples from a currently used system.
» How the increased complexity of a fault tolerant system increases the demands of the verification process and how this complexity is targeted.
» How the complexity and size of the architecture reflects in special problems encountered during the verification process.

Bonusföreläsning – Bygg Ditt personliga varumärke inifrån

Bitte Slättebol, StratEthics