ZapThink: Native XML Data Storage Will Evolve
Page 1 of 1
With the promise of another year around the corner, research firms are caught between wrapping up what has come before and looking ahead to what will be. Analysts at XML and Web services consultancy ZapThink Friday made the prediction that the native XML database (NXD) niche no longer exists as a separate market.
XML.com defines a native XML database as possessing a model for an XML document -- as opposed to the data in that document -- and stores and retrieves documents according to that model. The model must include elements, attributes, PCDATA, and document order. It also claims an XML document as its fundamental unit of storage, just as a relational database has a row in a table as its fundamental unit of storage. It is not required to have any particular underlying physical storage model. For example, it can be built on a relational, hierarchical, or object-oriented database, or use a proprietary storage format such as indexed, compressed files. Apache has an open-source version of a native XML database called Xindice.
ZapThink Senior Analyst Ron Schmelzer told internetnews.com XML-enabled relational database (RDBMS), content management, and integration vendors are best suited to offer general-purpose XML data store solutions, while XML database pure-plays are offering more focused XML data storage solutions. Schmelzer believes XML database features will eventually become incorporated in an increasing number of major software packages, including those offered by Microsoft, Oracle and IBM.
For example, XML and Microsoft's SQL don't match on a number of levels. They use different data types, and XML's ability to change structure in the middle of a document does not mesh with a relational database's rigid table structures. Also, when an XML document is stored in a relational table, information can be lost, such as element ordering and the distinction between attributes and elements. Native XML databases are remedies to this problem.
"We don't think this means the end for NXD vendors, it means that they will be increasingly selling "purpose-built" data storage solutions where the XML data store is just a component of an overall solution," Schmelzer said.
The analyst said proof of this testimony comes from market evidence. Schmelzer cited Progress' purchase of eXceLon for the benefit of its Sonic unit, Ipedo's focus shift to content management, Coherity's new data integration focus, Sybase's immersion in the native XML database game by developing a native XML indexing solution in ASE 12.5. Moreover, he said, there is evidence that Microsoft, Oracle, and IBM will soon follow suit.
"With the RDBMS vendors now selling native XML indexing, it begs the question as to whether there really is a separate market for NXD solutions, or whether it's a feature of a much larger system," Schmelzer said. "We now believe the latter to be true. Pure-plays need to choose whether or not they want to continue selling native XML databases and standing alone, or building stuff on top of them."
So, will the Microsoft's, IBM's and Oracle's scoop up a Xyleme, NeoCore, or AGiLiENCE?
"They might," Schmelzer said. "They are definitely adding XML indexing to their products. I'd expect to see [native XML databases] in Microsoft's Yukon and IBM's next DB2 release."
Not all is peaches and cream, however. Standards have yet to be established for NXDs and until they are there is bound to be confusion in the market over their purposes.
The study, which covers major pure-play XML data store vendors and XML-enabled relational database (RDBMS) vendors, is structured into two parts: a part that discusses the requirements for XML storage and the directions that the market is heading, and a part that profiles specific solutions and how they meet those requirements.