In-Depth

Q&A: Best of breed? It depends

• By Michael W. Bucken
• April 1, 2004

Editor-in-Chief Mike Bucken recently spoke with Mike Gilpin, VP and research director at Forrester Research Inc., about the state of development tools. This Q&A is part of our ongoing coverage on the State of Tools.

Mike Gilpin

Q: Is there still a place for best-of-breed tools [from unaffiliated companies like Borland and many smaller firms] or are platform vendors about to take over the tool world?

A: The answer varies depending on the type of developer you're talking about. Broadly, there are three communities of developers, with gradations in between and overlaps: Group 1 is code-centric, more technology-savvy; Group 2 is business results-oriented, less technology-savvy; and Group 3 is model-driven, engineering/ architecture-oriented (technology-savvy, but not in the same detailed way as Group 1). Given this, for each: Group 1 still has plenty of room for independents like Borland, these developers are very independent and like to "roll their own" dev environments, sometimes even at the command line. Group 2 developers will, in the future, get most of their tools from their primary platform vendor with occasional exceptions -- a tool vendor provides an exceptionally productive environment that is well aligned with the ecosystem of the customer's primary platform vendor. Group 3 has some increasing dominance by the platform vendors, but there is still some room for innovation by more independent vendors, helped by the fact that RUP and UML are standards. The rate of growth or shrinkage of the populations of each of these groups will have a big effect on the fortunes of the independents. I can see there still being room for Borland in the future, but they will increasingly target the first group, with some secondary consideration for the third.

Q: What is Java's future as its proponents fight among themselves?

A: I think Java's future is assured for other reasons: its importance to major platform players like IBM, and the impact of the schism in the Java tools space between JTC and Eclipse will not overly impede the overall health of this market. The other drivers for its success are simply too strong, such as being perceived as more open and standards-based, and providing a wider choice of platforms. The feuding has had a small braking effect, but as long as great tools like BEA's WebLogic Workshop keep happening, Java vendors will maintain good forward momentum overall.

Q: Do Web services eliminate the need for traditional dev tools?

A: No! The paradigm that those tools offer has changed, and will change more in the future, but it is more evolutionary overall.

Q: Can Java and .NET developers co-exist in the same IT shop?

A: Yes! They do, and will continue to do so. Our research shows that most enterprises still anticipate more, not less, overlap and collaboration between these worlds, and they have been expecting this for years. The answer varies with respect to the size of the company: the larger the company, the more likely to find both.