IBM Information-On-Demand XML Highlights

I am writing this quick report from the airport lounge in San Francisco, as I am waiting for my flight to Frankfurt and onward to Vienna for a week of meetings at Altova GmbH's headquarters in Austria. You just have to love Wi-Fi access - how did we ever get anything done before the Internet, laptops, Wi-Fi networks, and blogs?

With a week of conference sessions and trade show work in Las Vegas behind me, here is a quick summary of what I perceived to be some of the highlights of IBM's Information-On-Demand show from an XML Aficionado's perspective:

IBM announced DB2 version 9.5 (scheduled to ship October 31st), which contains several feature enhancements to the pureXML functionality, including inlining of XML, compression, several performance improvements for transactional XML, and the first implementation of XQuery Update in a major database. The last bit is probably the most interesting, because of the lack of updating capabilities in the original XQuery specs, which was exclusively focused on queries (analog to the SELECT statement in SQL). XQuery Update provides the ability to insert, delete, or update any node, i.e. potentially just one single element or attribute in the database, which should provide a huge difference to current implementation, where the entire XML document typically needs to be written back to a column in the database.

Altova's Nick Nagel, spoke on XML-Driven Data Management in a Developer Den session on Wednesday, which was well-attended. Nick's presentation addressed a "top-down" approach to data modeling using XSD (XML Schema Definition language) as a data modeling language with implications for data storage and retrieval as pureXML in DB2. Nick spoke on how XSD in turn can drive process implementation serving as formal design document, and how XML facilitates process development by enabling automated data binding, data mapping, as well as storage and retrieval with XPath 2.0 and XQuery. He also showed several screenshots on how Altova's tools can make working with XML in DB2 easier for developers.

IBM's Berni Schiefer conducted a Birds of a Feather session on Performance Tips, Conundrums and Experiences with DB2, where he answered customer questions and spoke in-depth about performance tuning of DB2, including tuning for pureXML. He also gave a few other interesting talks, but I did unfortunately not have time to attend those.

There were also plenty of interesting customer talks about how they are using XML, as well as more in-depth sessions on various aspects of pureXML in DB2, but my flight is boarding in a few minutes, so I don't have enough time to report on those anymore.