TechnologyDriven

Already last year Hasso Plattner had a lecture at his “Hasso Plattner Institute” (HPI) about the so called composite applications, applications “that sit on other applications” as he said, with no own data storage.

He gave an overview of what composite applications are and illustrated this with the example of Duet, a joint project of Microsoft and SAP to integrate Microsoft Office and SAP applications. I would say this is more an example of an integration of two systems than an example of a composite application, but at least it shows how relatively easy applications can be built with the help of standardized services, on top of web-based platforms (NetWeaver and .Net).
Technically an integration has always been possible Plattner outlined, but nowadays it is easier because of the standardized protocols (e.g. SOAP) and the systematization of application integration.

Nevertheless, there are some issues to consider when composite applications are built (and just as well when systems are integrated):

• User-experience consistency: which UI-guideline is to use, e.g. that of Microsoft or that of SAP?
• Data model overlaps: the data of which system is to use, when there are overlaps?
• Process overlaps: every software has some functionality, perhaps overlapping with other systems, and will obtain new functions in the future. So only the core-processes have to be combined.
• Joint configuration: to stay in the example of Duet: MS Office has extensive configuration options, and there are thousands of customizing tables in SAP. To bring the two systems together the configuration has to match.
• Life-cycle synchronization: every software has a release-cycle, perhaps one system changes and the other does not change accordingly.

Finally, as he likes to do very often, Plattner brought an example from the automotive industry, pointing the way for the software industry: in earlier days there were radios with their own brand in cars, nowadays the radios have the “user-interface” of the cars, but the car-manufacturers still don’t produce radios on their own. This would also occur in software-industry, Plattner predicted, software would consist of services (= components) of different manufacturers, and there would be a consistent UI on top the respective application-suite.

A video of the lecture (in German) can be found here:
http://www.hpi-potsdam.com/index.php/2006/07/05/summer-term-2006-composite-applications/

Technorati: sap blogger

A few weeks ago Hasso Plattner, co-founder and former board member of SAP, did a lecture at the “Hasso-Plattner-Institut” at the university of Potsdam, an institute for IT systems engineering he founded and is financing. The topic was “Design and Innovations in Enterprise Applications”, you can find the full version of the speech here and the highlights here.

In his lecture he made some interesting notes about SOA and Gartner calling it “the next big thing”: SOA would surly be a good concept and will be widely adopted (if it not already is), but there would be no “plug and play” or “putting lego-bricks” together, believing this he noted would instead be “the next big trap”. “We can do more”, but nevertheless complexity of software systems would increase, like seen in other disciplines of engineering (e.g. cars or airplanes).

Other parts of the speech focused on “people-centric design” and the ignorance of engineers towards the needs of users or consumers. Concerning this he brought the example of German car-vendors that disregarded the need for coffee-cup holders in cars for the American market for a long time.

It’s always nice to hear someone like him, with such a big experience, insight and passion for the software industry, no matter if you agree with all aspects or not.