Models and Measures

By Peter Coffee  |  Posted 11-27-2006

Announcements this week from Sweden's Telelogic and Alfresco Software in San Francisco address both ends of the process that brings the next generation of Web-based applications to reality.

Telelogic announced Nov. 27 its release of Version 3.0 of its Tau model-driven design tool. I had a chance to review the details of the new version in advance of that announcement, and I was struck by the way that the demands of mainstream enterprise application development have caught up with the things that Tau was designed to do.

The target market for Tau has traditionally been among builders of complex, distributed, heterogeneous real-time systems—think "intelligent battlefield" or "pay-at-the-pump retail"—rather than the comparatively uninteresting world of vanilla terminals whose response times could vary up to several seconds before an application would be said to be broken. The world of Web services is now a highly message-driven environment in which different types of device use a complex mesh of communication channels to provide prompt access to anything from financial market transactions to traffic alerts and updated driving instructions.

The abstraction of model-driven development, combined with Tau 3.0's enhancements aimed at accelerating the coding of Web services interactions, represents a timely aid to the hottest spot in development today—or at least to the developer topic that's tied with that of addressing Vista compatibility issues, but that's another matter for another time.

Later this week, we'll see another announcement from Alfresco Software and several partner companies concerning the joint development of an open-source benchmark for JSR-170 content repository performance. Getting away from application-specific file systems to more comprehensive, far more robust content management foundations has long been a goal that I've wanted to see the industry pursue: An open-source benchmark, whose details we'll see soon, offers us a common and transparent means of characterizing the capability and especially the scalability of alternative approaches to that end.

Read the full story on eWeek.com: Models and Measures