ERCIM Working Group on Software Evolution
by Tom Mens
In November 2004, the ERCIM consortium approved a new Working Group on Software Evolution. The purpose of this ERCIM Working Group (WG) is to build and maintain a network of ERCIM researchers within the particular scientific field of software evolution. In this sense, the WG will serve as a natural successor to an existing Scientific Network called RELEASE that will continue to be financed by the European Science Foundation until autumn 2005.
The inaugural meeting of the WG was held on 2 October 2004 in the Universita Degla Studia in Rome. It was co-located with ICGT 2004, the International Conference on Graph Transformation. The goal of the meeting was twofold: from the point of view of research, it served as a workshop for researchers to present recent results from a broad range of activities that fit within the general topic of 'Software Evolution through Transformations' (for more information, see http://wwwcs.upb.de/cs/ag-engels/ag_engl/Segravis/Events/SETra04/); the second goal was to start up a new ERCIM Working Group on Software Evolution.
The meeting was very successful, with thirty participants from eleven European countries attending. Twelve participants were members of the ongoing European Science Foundation 'RELEASE' research network, and representatives were present from eight ERCIM institutes.
Following the success of the meeting, the ERCIM Board of Directors decided to approve the proposed Working Group as an official ERCIM WG. At the time of writing, members of over 25 different research groups in European Universities, fourteen of which are ERCIM partner institutes, have already expressed their interest in becoming members of the WG, and this number is growing. If you are interested in joining the WG, please consult our membership policy and follow the procedure explained on our website (see below).
The scientific motivation for the WG is that numerous scientific studies of large-scale software systems have shown that the bulk of the total software-development cost is devoted to software maintenance. This is mainly because software systems need to evolve continually to cope with ever-changing software requirements. Unfortunately, existing techniques and tools for the support of software evolution have many limitations, including being (programming) language-dependent, not scalable, difficult to integrate with other tools and lacking in formal foundations.
The principal goal of the proposed WG is therefore to identify a set of formally founded techniques and associated tools to aid software developers with the common problems they encounter when evolving large and complex software systems.
The WG intends to address a diverse array of evolution problems. We will not only attack the technical aspects of software evolution (the how), but also try to understand and model the fundamental principles behind software evolution (the what and the why). Following is a tentative and inevitably incomplete list of topics to be addressed:
- specification or analysis of the evolution of software artefacts in a broad sense (including, but not limited to requirement specifications, architectures, designs, models, metamodels, programs, components, tests, documentation, bug reports, version control information, log files, release histories, language descriptions, APIs and protocols)
- re-engineering and reverse engineering
- software restructuring, refactoring and renovation
- model-driven engineering and model transformation
- co-evolution, consistency maintenance and inconsistency management
- impact analysis, effort estimation, cost prediction, evolution metrics
- traceability analysis and change propagation
- version control and configuration management
- run-time adaptation and dynamic reconfiguration
- family and product-line engineering
- methods, processes and tools for managing software evolution
- development of a formal theory of software and systems evolution.
Activities in 2005
As for all other ERCIM WGs, recurrent activities include proposing new European research projects, enhancing intra-European collaboration at both research and teaching levels, collaborating with other WGs, and organising scientific events (eg international workshops).
In 2005, the WG will organise the following activities:
- a symposium on Software Restructuring, 6 January 2005, Gent, Belgium (co-financed by FWO/FNRS, Belgium)
- the annual WG meeting, to be held in co-location with the international conference ICSM 2005 in Budapest, September 2005.
Other activities may be added to the calendar; an up-to-date list can be found on the WG website.
Tom Mens, Institut d'Informatique, Université de Mons-Hainaut/FNRS/FWO
Tel: +32 65 37 3453