Thursday, March 24, 2005

“Where does the newborn go from here? The net is vast and infinite.” (from my Borland blog)

These are the final lines of the popular Japanese anime “Ghost in the Shell”. This seems appropriate to describe the momentum, interest, and uncertainty around the Eclipse project. David Intersimone and I spoke recently at EclipseCon 2005 on this topic.

Our presentation “The IDE is Dead, Long Live the IDE” was well received with people sitting in the aisles and standing in the back and the doorway. (Memo to self: Always get your session put in a small room so it looks exceedingly popular).

David gave a smart overview of where we have come from with IDE’s over the past 20+ years and made at least one super-duper smart point. When you take a step back and look we have retooled the underlying tool framework of the IDE almost every time we have changed from one software engineering “epoch” to another. A good example is the OO environments we had on the LISP machines, or for Prolog and Smalltalk. They were aptly built and customized for the emergence of the object-oriented age. As “OO” ceased to be the defining invention, as we moved into perhaps the “distributed epoch” – the OO vendors did not necessarily have the economic oomph, nor the point of view necessary to incorporate these dimensions into their products. A new generation of IDE’s emerged that were more optimized for database or distributed development, and which unfortunately ignored a number of the features of use from the previous epoch.

David made the optimistic point (with which I concur) that the Eclipse project represents a significant chance that we now have a tool framework which will outlive our current engineering epoch and allow us to tool into the next epoch without rebuilding the entire foundation. Don’t misunderstand us – PRODUCTS have survived epochs before, for example Delphi, JBuilder and Visual Studio. However, whilst the products moved forward across epochs their underlying foundations were rebuilt multiple times.

Next item to address is the fear that “Oh my god the sky is falling – Eclipse is commoditizing the core IDE functions and now Borland will have nothing left to sell”. Not true at all. In case you missed it we have grown up the stack since the days of Turbo Pascal. Download it and take a look, how could there of been anything left to build after that? Same thing with Eclipse – there is a lot left to do to make commercial products that support the individual’s productivity, team productivity, and align the projects for business success.

Another thing about commoditization is its several possible outcomes; it either drives capital investment to the sidelines, which is bad, or it causes capital investment to chase higher value solutions. I think if you look at the activities of thousands of entrepreneurs, privately held companies, and publicly held companies innovating on top of Eclipse, you don’t see capital in slothful repose.

As we see the emergence of tooling foundations like Eclipse and Visual Studio as platforms, the challenges falls to Borland to help customers bridge the gap between disparate platforms and find ever higher levels of value.

Domo arigato. Oyasuminasai.