Can IBM Make Data More Accessible Through SOA?

IBM launched multiple new products at its Information on Demand conference this week, as well as announcing more industry-specific deployment frameworks and partner programs. All are aimed at making data more accessible, in most cases (though not necessarily) through SOA or Web-based applications. The theory is that while SOA can do a good job at making executable processes accessible to other systems, databases are still relatively hard to access. That means users need either an XML-based database or a layer of ESB-like middleware that will convert existing databases to Web services.
On the database side, IBM announced DB2 data server 9.5 (codename Viper 2.) The current release 9.0 ("Viper") was the first version to use native XML formats, and 9.5 builds on this with improved encryption, management and compression, the latter being particularly important for XML files. The main competitor here is Oracle, though open-source and Microsoft options are gaining in popularity., and the move to standardized XML helps level the playing field. In what looks like an attempt to counter some of this, IBM is also launching the Data Studio, a free Eclipse-based IDE for XML databases that can be used with other vendors' products as well as DB2. Both will be available on October 31.
On the middleware side, IBM announced new functionality for its Information Server and a new MDM (master data management) product. The new features in Information Server are mostly incremental improvements, including better support for SAP and IBM's own zSeries, as well as replication features taken from its recent acquisition of DataMirror. MDM is a new area for IBM, essentially caching data from multiple databases and making it more easily accessible to users or other systems. The main competitors here are specialist vendors like Initiate Systems, which have a head start.