Typically, telecommunications services are implemented in software. Feature interaction is the term used to describe interference between services or features; most attention is given to cases where the interference is undesirable, ie. there is an incompatibility. In telecommunications, control and data is distributed and on such a large scale that software development is by numerous disjoint teams; by its nature, therefore, this software experienced the feature interaction problem first. But, while the workshop focuses on communications services, the subject has relevance to any domain where separate software entities control a shared resource. An important aspect of this and past workshops is the mix of experts in the area. A synergy is obtained by mixing academics and industrialists, vendors and operators, engineers and computer scientists, practitioners with theorists. At Glasgow, the workshop tradition is maintained, so in addition to fully refereed papers, time is also devoted to invited speakers, discussion sessions and posters. The afternoon discussion sessions and the posters have key roles.
Both allow a degree of exchange that presentations cannot permit: the posters on a one-on-one basis, the discussion sessions on a broader platform. Of course the presentations are an important framework for the workshop by introducing novel ideas and reporting on experiences. In addition, three well-known workers within the field are invited to give their perspective on the problem. What better way to test out the offered approaches than by having a contest? This 6th workshop maintains the contest started at the last workshop in Sweden. Groups enter the contest during the months preceding the workshop, but the winner is kept secret and only announced during the workshop. We are entering an exciting new phase for feature interaction. New generation networks are offering huge opportunities for growth and change. For example, a momentum built upon CORBA and DCOM is accelerating as consortia such as PARLAY and JAIN offer open public standards. No longer do we have isolated features and services unaware of each other controlling shared resources; the signalling limitations of POTS and IN are giving way to a regime where services may communicate directly and negotiate.
Balancing this new paradigm will be the increased possibilities of interaction: simply consider an accelerating number of features and services across an increasingly disparate range of networks, within an increasingly deregulated market. Consequently, enormous new possibilities for feature interaction detection, resolution and of course avoidance will be revealed; never has feature interaction research been so necessary.