The objective of the experiment is to evaluate the benefits of using a formal change control system supported by configuration management tools.
The proposer, Sysdeco Innovation AS, is a vendor of software development tools, for the 4GL and GIS markets. Sysdeco's products are available on many operating systems, windowing systems, databases, and networks. With many delivered configurations - the numbers of licensees are in the ten thousands - it is important to have a change control system which ensures that the products are correctly configured. It is important to be able to reconstruct an old configuration, correct reported problems, and control the propagation of changes to other configurations. When configuration errors are detected in the system, often a number of people are involved such as sales persons, customer support engineers and developers. The costs are accordingly high.
A formal change control system (Clarecase) will be implemented, including organisational changes and more formal processes. Change control will be applied to all parts of the products, e.g. source code, object code, executable code, documentation, media, test plans, test data, make files, compilers used.
Sysdeco has a software experience database that will be used to measure and validate the effects of the experiment. The data in the database will be studied and several metrics will be defined and used to measure the effect of the experiment.
The formal change control system will be used in the development of the next release of a specified product and in the maintenance of existing products. Measurements will be done and compared with the data from the existing database.
EXPECTED IMPACT AND EXPERIENCE
In order to keep ahead of the competition the vendor must continually improve its software development and maintenance process. Introducing a formal change control system is believed to improve the process, resulting in 5% shorter time-to-market and a 10% reduction in error reports for the software products.