 |  |
| Etusivu | In English | Palaute |
|
Aika: 1.10.2009 klo 09:00 - 14:30
Paikka: Digia Tower, Hiomotie 19, Helsinki
Mallilähtöinen ohjelmisto- ja prosessisuunnittelu / Model-Driven Design
Presentations Juha-Pekka Tolvanen: Moving from coding to model-driven development Abstract: Domain-Specific Languages and Model-Driven Development have moved from scattered successes, through industry hype, to increasingly widespread practical use. Well-attested benefits include raising the level of abstraction, improving productivity, and improving quality. The main questions are no longer what or why, but where and how. This session will demystify Domain-Specific Modeling, open the lid on the OMG's MDA, and shine a light on Microsoft's Software Factories and Oslo. We show which bits of Model-Driven Development work, which might work, and which don't, and explain where they can be applied - and where not. Most importantly, we show the steps to define your own modeling languages and generators to start using models effectively. The session concludes by highlighting the lessons learned from industrial experiences. Bio: Juha-Pekka Tolvanen is the CEO of MetaCase. He has been involved in model-driven approaches, metamodeling, and domain-specific languages and tools since 1991. He has acted as a consultant worldwide on modeling language and code generator development. Juha-Pekka has authored a book on Domain-Specific Modelling (Wiley 2008) and written over sixty articles for software development magazines and conferences. He holds a Ph.D. in computer science. David S. Frankel: Semantic Interoperability: The Next Integration Challenge Abstract: Macro economic forces that require B2B and A2A integration are making data transformations increasingly important. When the MDA initiative was launched early in this decade, its founders accurately predicted that model-driven techniques would make data integration less costly and error-prone. However, integration costs are still too high, consuming 30-40% of IT budgets. Now it is time to chart the next milestone on the data integration path: improvement in semantic interoperability. Data and message format definitions contain only one kind of machine-readable metadata, namely metadata about the syntax of the information. Machine-readable syntactic metadata makes it possible for current state-of-the-art data integration tools to generate data transformation code, once the integration analyst has entered a mapping from one format to another into the tool. However, this metadata does not provide integration tools with a basis to help integration analysts decide what the mapping between two formats should be. This is a key reason that integration work is still so labor intensive. Moreover, misunderstandings about the semantics of data and message definitions lead integration analysts to make subtle errors that can be costly and difficult to track down. Machine-readable semantic metadata helps to address these limitations. This presentation will explain the nature and structure of machine-readable semantic metadata that is now coming on line in standards for the manufacturing and financial services industries and is being increasingly used by ERP vendors, opening up possibilities for a new generation of tools that make integration analysts substantially more productive and accurate. Bio: David S. Frankel is a member of SAP's Standards Strategy and Management team, which is part of the Global Ecosystem and Partner Group. He focuses on standards for the financial services sector and for model-driven systems. David has 30 years of experience as a programmer, architect, and technical strategist. He is recognized as a pioneer and international authority on the subject of model-driven systems and semantic metadata. He has published two books and dozens of trade press articles, and has co-authored a number of industry standards, including UML and the upcoming revised version of the ISO 20022 core methodology for optimizing financial networks. Jan Wirix: Novaxess case study Abstract: This case study reports on the successful application of MDA in the implementation process of an integrated back office system for a telecom startup company in Amsterdam. The presentation shows how a functioning system, incorporating important business requirements as flexibility and time to market, has been built while using resources significantly below industry benchmarks. Both technical and human aspects of the project and lessons learned are discussed. The main steps of the project were the specification of a platform independent model for the service infrastructure according to the MERODE methodology; the transformation to a platform specific model (in this case VERSATA); the integration of third party business applications and the development of user interfaces on top of the service infrastructure. The case illustrates the importance of the so-called IT demand approach, where business sponsors work together with an architect role to produce “blue prints” of a solution that can be built by an IT supplier. Bio: Since 1981 Jan Wirix has assumed positions at development and management levels in integration projects in an international context for companies as Wang Laboratories, Tractebel Information Systems, MCI WorldCom and Novaxess. In 2002 he started as a self-employed enterprise software architect, focusing on consultancy for strategic IT alignment projects, enterprise architecture and development of an IT demand role in organizations. He also teaches the master level course “project management and advanced software development methods” at the Katholieke Universiteit Leuven (KUL) since 2009. Jan Wirix obtained a master in computer science (KUL 1978) and a master of business economics (KUL 1988). Matti Rossi: Are Domain-Specific Models Easier to Maintain Than UML Models? Abstract: While Domain Specific Modeling (DSM) languages have been adopted in industries such as telecommunication and insurance, they have not yet gained wide acceptance in practice because claims of increased productivity and ease of understanding have not been established by independent studies. Our study addresses this concern by comparing the maintenance performance of DSM and general purpose modeling languages. Maintenance in software intensive systems is critical because software often continuously evolves during development as well as after delivery, to meet ever changing needs of users. Proponents claim that a key driver of the DSM approach is the easier comprehension of the structure and behavior of systems, which should make both evaluating and maintaining the models easier. We investigated this through the following research question: Does DSM improve the maintenance performance of designers when compared to general purpose modeling using UML? Bio: Matti Rossi is a professor of information systems at Helsinki School of Economics. He received his PhD from the University of Jyväskylä. He has been the principal investigator in several major research projects funded by the technological development center of Finland and Academy of Finland. His research papers have appeared in journals such as CACM, Journal of AIS, Information and Management and Information Systems. He is a member of ACM and AIS. Panel discussion MDA vs. DSM: either or both? There is some debate whether organizations should utilize MDA or DSM approach. In this panel we discuss the pros and cons on both approaches and also if there is room for both approaches to co-exist in certain cases. |