Burkhardt acknowledged Ingres' need to regain credibility that was lost during a period of decline. The database was created in 1973 as an academic project at the University of California at Berkeley and at one point rivaled Oracle as the leading enterprise relational database. The source code was spun out commercially and formed the basis of several databases, including NonStop SQL and Sybase. The best-known at the time was Ingres, which was popular in the 1980s. Ingres was bought by ASK Corp. in 1990, by which time usage had started to decline. It was sold to CA in 1994.

"Clearly there are other open-source databases with more mindshare than us now, but they are starting from a more recent technology base," he said. "If I'm a CIO with a mission-critical need, I wouldn't want to invest in a database proven only for read-only Web applications."

Curt Monash, a database consultant in Acton, Mass., and a Computerworld columnist, said that while Ingres tops open-source rival such as MySQL in terms of features, it lacks momentum. Moreover, Ingres' lacks the latest crop of features touted by the large relational database companies. "I don't see why many new customers would adopt the product, until it rivals MySQL in low-cost administerability," Monash said.

But Burkhardt argued that Ingres' low upfront costs will attract many companies that are not looking for the latest and greatest feaures.

"It's true that Ingres doesn't have the bleeding-edge features. It's also true that there's a very large set of users who are paying through the nose for Oracle to support databases that are 10 to 12 years old," he said. "New applications generally don't demand cutting-edge functionality. Three to five years ago before application servers were established, people did a lot more in PL/SQL. Now people want a more standardized approach."