Microsoft Application Model
Clemens Vasters has given a presentation at the EMEA Architecture Forum in
Paris (01/14/02). This forum was organized by Microsoft. The slides can be found
As of today, there are basically two routes possible for BPM technologies.
The first one that most vendor took is to claim that BPM adds value to
Enterprise Application Integration and that BPM should be the focus of EAI. If
this is a valid approach, it is also cornering BPM limiting its applicability.
The question is rather isn't all business applications somewhat business process
based? Most enterprise systems from ERP, CRM, SCM, to PLM automate not only the
tasks performed by various roles in the organization but also coordinate (if not
orchestrate) these tasks across the organization. Funny enough no provision
exist -at all- to make these business processes first class citizens within
modern application models and languages (J2EE being the best example of this
lack of understanding of the foundation of business applications) all they can
deal with is ECA (event condition action) using the MVC (model-view-controller)
Microsoft commitment to BPM has been early and mostly focused on EAI. The BizTalk
server was announced in 1999 and delivered in its first version in early
Then came .NET. At first I must admit I was disappointed looking at the .NET
application and felt that they were just trying to do a better J2EE platform. I
also felt that this overemphasis on "web services" was something
fairly artificial yielding little benefits to an application model. Ironically enough, with the power of its desktop
application (Office, IE, Visio, ...) and other potential clients like the
excellent Groove application, Microsoft has
the ideal substrate to build a winning BPM strategy since all the complex user
activities can be performed in the most appropriate client not just IE.
When I came across the presentation from Clemens Vasters and the excellent
of Mark Lucovsky, I felt that all is now in place for a new application model to emerge which could seriously challenge J2EE.
(Please see Clemens slides
for the details. The second presentation on J2EE interoperability is
more technical and offers greater details.)
Of course IBM is not letting Microsoft go alone in this direction as pointed
out by this article: https://www.infoworld.com/article/03/01/27/hnj2eemove_1.html
( I did not see any blueprints relative to that. )