In this report from the field, SystemiNetwork.com blogger Bill Blalock shares his take on the EclipseCon conference last week. Eclipse, of course, is the open-source community dedicated to building an open-development platform based on a Java IDE.

The fourth EclipseCon took place in Santa Clara, California, last week, with an unofficial paying attendance of 1,350 and a noticeable absence of IBM products and System i talk. The gathering was nevertheless similar in other ways to COMMON conventions I attended in the late 1980s and early 1990s.

The structure of the programming tracks and exhibits was nearly identical to those of COMMON. Perhaps it was the luck of the draw and my interests, but the COMMON labs I recall were more challenging and had more impact, and I brought back more from them. The exhibit hall could have been from any of the System i conventions I have attended — with the exception of IBM's scaled-down presence.

IBM didn't exhibit any products but rather just sponsored a lounge and posted a schedule of when different IBM experts would be available. I didn't see a single product brochure in IBM’s space on the exhibit floor.

This isn't to say that IBM had a small representation. I would not be surprised if as many as 10 percent of the attendees, presenters, and support staff were employed by IBM.

IBM created the Eclipse organization and then helped it morph into the Eclipse Foundation. Although the foundation is independent, IBM has great influence over it. The company's financial support and participation on the board of directors are the tip of the iceberg. IBM has pervasive influence through its employees, who are paid to work on open-source projects and are also paid for volunteer time. IBM, at least for now, makes resources available to Eclipse-related open-source and commercial projects. If IBM isn't doing the work, it is supporting staff members who are doing so in their open-source personae.

I didn't hear anyone being told, "That is an opportunity for an independent business partner." I seem to remember something akin to this statement as a standard answer to why a feature or need wasn't met on the AS/400 in days past.

Jazz on the Way

Through the Eclipse project, IBM enlisted the open-source community to create a new software development paradigm in order to break out of a software development evolutionary dead end it had reached in the late 1990s. IBM legitimized the effort by unfolding it into the Eclipse Foundation to broaden the support base and keep the open-source community contributing. Now IBM is attempting to put the Eclipse experience and framework into a purely commercial project called Jazz. On Wednesday, IBM highlighted Jazz at EclipseCon with a full-court press.

The System i?

The System i wasn't discussed much at EclipseCon. I spoke with only one person among the vendors, presenters, and attendees who worked with the System i. Some Danish members of the convention remembered the System i as being marketed in Europe as a replacement for PC servers. It seems to me that an option the System i has for future viability is to co-exist and partner with Eclipse. Unfortunately, I failed to run into any like-minded people.

I searched for IBM staff members who worked on Rational WDSc for the System i, but I didn't meet any. I did learn that Rational WDSc for the System i is generally three months behind the rest of the Rational WebSphere product line. I also learned that the technical basis of the claimed increased performance in the new version is believable and logical and that WDSc 7.0 for System i is based on Eclipse 3.2.1.

I think there is a disconnect between how WDSc would be used in shops with large legacy code bases (if it worked) and the perception of IBM developers. WDSc has some cool features and helps shops understand complicated legacy code. However, I don’t think that the right people in WDSc development realize that it just doesn’t operate in large, complicated programs in which it could give significant benefit to the users. At EclipseCon I sought Rational developers with whom to discuss this but found very few leads.

Sad Conclusions

I am sad to conclude that IBM has decided to leave the System i in the software development evolutionary dead end it hoped the Eclipse project would change. System i shops that won't be modernizing applications do not demand Eclipse-based tools specific for the System i platform, which means they won’t be embracing useful parts of Eclipse — making it that much easier for IBM to do so.