/* Google Analytics ----------------------------------------------- */

Tuesday, December 2, 2008


I went to the Paris CASEWISE User Group.

The first speaker made a presentation on how to manage complex project (especially in finance) and benefiting from a tool like casewise. He also provided some key hints on how to make a project succeed.
1/ Management support
2/ Investor support
3/ know your client (or find it)
4/ manage the change and train the trainer
5/ use technology

He also stated that to ensure success you need to be on TOP (technology / Organization / Process).

I made a presentation on IS cartography: how to migrate from mega to Casewise.
The key idea is whatever the tool to keep your EA framework. So you can define objects, views and diagrams to describe views. The EA tool is then seen as a facilitator.

We also had the presentation of the new Casewise EA modeler based on Silverlight 2 and called WebModeler. Technologically impressive, web based and very adapted to today's distributed team and way of working. I like the SaaS approach that can also be used for Compliance as a Service.

A great user conference ... Only one concern, Casewise is not offering goodies ;)

Sunday, September 28, 2008

My Publications

20091. I was invited to publish a guest post on SOA governance concerning how you could categorize your services. Take a look at : Art of software reuse blog.
2. IS Rating: French article here and English article here. The official web site is here. Check the spreadsheet and let us know if it works for your organization
3. The article I wrote on Cloud Computing (InfoQ): here
4. I answered some SOA questions in a virtual pannel published on InfoQ, here.
2003 “MDA Guide Version 1.0.1”, OMG document, June 2003 (Download).

“Model Driven Architecture for Agile Web Information System Engineering”, El Kaim W., Struder P., Muller PA., 9th International Conference on Object Oriented Information Systems (OOIS’03), 2-5 September 2003, Geneva, Switzerland. LNCS Volume 2817, August 2003, pp. 299 - 303. (Download
2002 “Outils, Méthodologies et Technologies de produits MDA pour la
génération et le déploiement automatique de systèmes d’information Web“,
El Kaim W., Burgard O., Muller PA., Génie Logiciel n. 61 (ISSN
1265-1397), pp. 23-33, Edited by GL & IS, July 2002. (Request copy here)
2001 OMG paper, “Model Driven Architecture - A Technical Perspective”, by OMG Architecture Board MDA Drafting Team, July 2001. (Download)

“MDA Compliant Product Line Methodology, Technology and Tool for Automatic Generation and Deployment of Web Information Systems”, El Kaim W., O. Burgard & P.A. Muller, Software & Systems Engineering and Their Applications (ICSSEA) Conference, December 4-6, 2001.

“Petri Nets for Systems Engineering: A Guide to Modelling, Verification, and Applications”, Girault, C. and France Valk, R., El Kaim W. in Chapter 21, pp. 433–470, Springer Verlag, 2001., 606 pp.
(Order the book)

Editors: Juan C. Dueñas, William El Kaim, Cristina Gacek, Style, structures and views for handling commonalities and variabilities. ESAPS Deliverable (WG 2.2.3), Eureka 2023 Programme, ITEA project 99005, 2001. (See ESI web site)
2000 “Managing variability in the lcat split/daisy”, W. El Kaim, In Proceedings of Product Line Architecture Workshop, The First Software Product Line Conference (SPLC1), pp. 21–32, August 28, 2000. (Download)

“Applied technology for designing a PL architecture of a pilot training system”, El Kaim W. & al. International Conference on Software Engineering (ICSE), The First Workshop on Software Product Lines Economics, Architectures, and Implications, Limerick, Ireland, June 10, 2000. (Download)

“Domain Analysis and Product-Line Scoping: A Thomson-CSF Product Line Case”, S. Cherki, W. El Kaim, P. Josset, and F. Paris. In Proceedings of the International Workshop on Software Product Lines: Economics, Architectures, and Implications. IESE-Report 070.00/E. (Download)
1997 PHD Thesis: “Structuration, placement et exécution de composants logiciels dans les applications reparties ou paralleles : mise en oeuvre avec des applications construites selon le paradigme client-serveur sur des architectures materielles hybrides” (Download PS files).

"Interoperability and Performance of an Object Distributed Application Based on Corba", Castiel S, El Kaïm W., “Proceedings of the International Conference On principles of Distributed Systems (OPODIS 97)”, pp. 269-272, Chantilly, France, December 1997. 
1996 "VODEL : Un langage de description d'applications parallèles à granularités variables", El Kaïm W, Menaud J-M, Actes de l'Ecole Thématique CNRS & PRS Placement Dynamique et Répartition de Charge, Presqu'ile de Giens, France, 1996.
1995 “H-COSTAM: a Hierarchical Communicating State-machine Model for Generic Prototyping”, W. El Kaïm and F.Kordon, In Proceedings of the 6th IEEE International Workshop on Rapid System Prototyping (RSP 1995), number 95CS8078, pages 131–137. IEEE Computer Society, 1995. (Download on IEEE Xplore)

“Outils pour la vérification des VADIS dans IRENA”, A. Diagne, W. El Kaïm, K. Foughali, C. Girault, F. Kordon, and S. Romand. Technical report, Rapport d’avancement (année 3) du projet IRENA, Juin 1995.

Guide du doctorant.
1994“An Integrated Framework for Rapid System Prototyping And Automatic Code Distribution“, William El Kaim and Fabrice Kordon, in proceedings of the 5th International Workshop on Rapid System Prototyping, pp. 52-61, 1994. (download on IEEE Xplore)

F. Kordon and W. El Kaïm. The CPN-Tagada User Manual, June 1994. Now called Framekit

Monday, January 28, 2008

The end of Enterprise Architecture as we know it

Did you already be part of an enterprise Architecture (or city planning or IT urbasnism) program in your company?

If yes, then you already experienced the numerous meetings around which framework to use (TOGAF, DODAF, etc.), the architecture views to propose, the tool to find, buy and deploy (Mega, Casewise, etc.), the documents to produce, the web site to generate, etc.

And when everything is ready (tool, process, organization) then the most difficult part has to be done: data in context should be loaded in the tool by stakeholder or automatically through specific interfaces. And this part is taking time.

Making different people do cartography and following the same modeling techniques, the same naming conventions, the same granularity is very difficult. Of course, you can do it. It takes time and money and when it's done ... it is generally no more accurate.

Loading data in a tool is not the issue. making drawing to show some particular view of the system is also not an issue. The main problem arise when you try to be exhaustive in linking objects (a server to an application, a server to an environment, etc.).

I think that this way of doing enterprise architecture is a dead end! The end of Enterprise Architecture as we know it.

We should think of enterprise architecture as a social network. The idea is to create process and tools to let users create the global view by collaborating and bringing for each of them part of the reality. But what is new, is that you do not force the user to follow a particular format or view. The user describes its vision of business processes or IT in its own "Domain language".

Behind the collaboration tool, a semantic engine, should look at all created models and data and try to build the semantic relationships. Then you can leverage knowledge management and look at enterprise architecture from an associative point of view (instead of declarative and normative).

Of course today no tools exist on the market. But When I see what "web 2" is bringing and how it changes how people work, I do think that this will also change the way we make enterprise architecture. So instead of buying monolithic proprietary tools for enterprise architecture, I will rather like to have a great new set of collaborative tools that let me relate information in context and let the tool build the network of relationships and dependencies. Displaying those dependencies can be done using SVG or rich application framework like Flex and Silverlight.