In its recent quarterly results Oracle announced that its database and middleware division had grown by 17% in terms of new licence revenues. What are we to make of this? Well, it sounds (and is) very impressive. However, Larry subsequently commented that the middleware component of this had grown by more than 80%, a fact of which he was particularly proud because it means that Oracle has overtaken (at least in revenue terms) BEA to become the second largest purveyor of middleware, after IBM.

But, if middleware has been growing so fast what does that say about growth in the database portfolio? Well, it is difficult to be precise but one has to assume that database growth is certainly no more than in single figures. Moreover, the database part of this equation includes add-ons such Real Application Clusters, the Automated Storage Manager and Enterprise Manager, all of which the company claims to be increasingly popular, and which will take a substantial proportion of what growth we have left after taking out middleware. And finally, while relatively small parts of the database group in Oracle, the remaining figures will also include sales of TimesTen and BerkeleyDB.

In addition, bear in mind that the value of the dollar has been going down. On a constant currency basis (as opposed to the dollars that Oracle reports in) it would not surprise me if growth in 10g sales per se has been stagnant at best. Certainly, any growth must have been small (albeit it coming from a very large installed base). So, is this surprising?

I don't think so.

To begin with, Oracle must be facing increased competition from IBM, especially with the latter's new XML capabilities. Secondly, it is being attacked by Microsoft at the low end. Then there is EnterpriseDB which, while it may be a pinprick to Oracle, has picked up 100 or so potential or actual Oracle customer installations over the last year; and finally, there are the data warehouse appliance vendors, such as Netezza, that have won a number of major deals against Oracle (and others).

And then there's user concern. Sony Picture's CIO commented at a conference in Australia recently that "In fact I can fairly, confidently, frankly, say that the customer service component to working with those software vendors since the Oracle acquisition, each one of them on their own, has not helped. They've gotten worse. The service has gotten worse. The products have become somewhat stagnant and it's a real concern for me." Now, he was talking about the company's application portfolio and not its database but it would be a worry for Oracle if these concerns were more widespread. Fortunately, Oracle announced at Open World that it would be investing even more resources into the service side of its business, so this should not be a concern in the future even if it is (for Sony) now.

So, is Oracle doing anything about this?

Well, at least on the surface, not a lot. It has tinkered with its pricing model and it has announced a partnership with Sun in the data warehouse space. But this is just tinkering at the edges: prices can be discounted anyway and data warehousing appliances aren't just about pre-installing the software.

I think that part of the problem is that many of the companies that have invested in Netezza or EnterpriseDB have Oracle enterprise licences that allow them to install as many instances of 10g as they like—the fact that one of two of these possibilities fall to other vendors is neither here nor there as far as Oracle is concerned since it doesn't affect the enterprise licence. In other words, the problem is at least partially hidden from Oracle, at least in terms of the bottom line, which explains why the company has not been more aggressive in reacting to these threats: it simply doesn't see them as being of significant magnitude. Time will tell if Oracle is right.