Workshop: Architectural Concepts and Programming Languages
November 9th and 10th 2009, Munich
This workshop is focussed on the inclusion of architectural concepts in programming languages. This includes questions on
- the relationship between different architectural views and programming languages (including formal refinement and verification of certain consistency properties)
- the definition of extra-functional properties and abstractions needed to predict such properties, like performance or reliability
- the support of software evolution by explicitly defined traceability, design decisions and rationales.
Participation in this workshop is by invitation only. See below for a list of confirmed participants.
We do not plan to have a formal proceedings volume nor to run this workshop as a mini conference. Much more, we only will have short position statement presentations to dedicate most time to discussions. However, a written position statement (less than 10 pages) may be submitted. We will collect these statements and distribute them before the workshop. To be included in this collection, please submit your position statement before October 15th to rhode(at)ipd.uka.de.
- Sven Apel, Universität Passau
- Colin Atkinson, Universität Mannheim
- Steffen Becker, Forschungszentrum Informatik (FZI) Karlsruhe
- Erik Burger, Universität Karlsruhe
- Gregor Engels, Universität Paderborn
- Holger Giese, Universität Potsdam
- Martin Glinz, Universität Zürich
- Michael Goedicke, University of Duisburg-Essen
- Thomas Goldschmidt, Forschungszentrum Informatik (FZI) Karlsruhe
- Gerhard Goos, Universität Karlsruhe
- Jörg Henß, Universität Karlsruhe
- Jean-Marc Jezequel, Université de Rennes
- Christian Lengauer, Universität Passau
- Raffaela Mirandola, Politecnico di Milano
- Oscar Nierstrasz, Universität Bern
- Frantisek Plasil, Charles University Prague
- Iman Poernomo, King's College London
- Arnd Poetzsch-Heffter, Universität Kaiserslautern
- Wolfgang Pree, University of Salzburg
- Andreas Rausch, Technische Universität Clausthal
- Bernhard Rumpe, RWTH Aachen
- Wilhelm Schäfer, Universität Paderborn
- Wolfram Schulte, Microsoft Research
- Axel Uhl, SAP AG
Towards Synergies between Architecture and Code
As we have learned in recent years, architecture is one of the most dominant artefacts in the development, in particular, of large software systems. Therefore a huge amount of research has been carried-out directed to architecture. A good example is the work on architecture description languages (ADLs). So far, however, it is an unsolved question, how architectural descriptions say by ADLs or general modelling languages find their way into programming language code, since so far there are not too many concepts from architectures that have found the way into programming language concepts. As a result in many software development projects, even in those where architecture is carefully defined, the architecture is not explicitly present very much in the final code. Apart from the often found deficiencies in the description of architecture, the missing information in the code makes it very difficult to use the architecture in the analysis and further evolution of software at the code level. In particular, architectural information cannot be used in the analysis of code in automatic code inspection or performance analysis in many other important ways to deal with code. It is an interesting scientific challenge to go a step towards a stronger and deeper integration of architectural description and code as a basis for improvement of the software development and software evolution processes.
The challenge is whether and how it is possible to inject architectural information into the code with the goal that as the result we get a structured description of large software systems, they go beyond the simple structuring into file structures as we find it today. An integration of architectural issues and additional architectural documentation in the code that can be used at compile time but also at runtime for performance analysis and runtime optimisation would be one of the goals. From a model-driven software engineering perspective, it is to be investigated, which architectural elements in a language make other redundant, as specific abstract concepts can be used to generate code (e.g., abstract connector specifications). The idea would be to have a coherent architectural concept that also supports the distributions of parts of the software onto distributed hardware architecture as a result, a smooth integration of a number of execution concepts is necessary going away from purely sequential execution models at the level of components. In this respect many problems have to be considered and to be solved which can only be done by a community of scientist that understand the various aspects involved. To see whether such an approach is possible and finds enough support it is suggested to carry-out a workshop of about two days where a number of people are invited and should discuss among them chances to do steps and to an integration of architecture information in code.
We suggest reservation as early as possible. Please use the Hotel Reservation Service (FAO Ms. Pontzen: Tel.: +49(0)89 323 840, Fax: +49(0)89 323 841 00, E-Mail: H3679(at)accor.com) and consider referring to the password “ArchiPL” in order to book one of the rooms we reserved at Hotel IBIS Muenchen Garching, which is situated in the Garching-Hochbrueck business district. The Campus can be reached from Hotel IBIS by underground in about 9 minutes. Prices are very reasonable.