On June 1, 2004, the Eclipse Board of Directors named Mike Milinkovich the executive director of the Eclipse Foundation. The hunt had been on since Eclipse became an independent entity earlier this year. Here JDJ's Eclipse editor, Bill Dudney, talks exclusively with Milinkovich about the direction he wants to take Eclipse and what his vision is for the community.

Mike Milinkovich comes to the Eclipse Foundation from Oracle, where he served as the vice president of OracleAS Technical Services. Prior to Oracle he was at WebGain as the vice president of World Wide Services, Training and Support Services. Milinkovich joined WebGain as a result of their acquisition of TopLink from ObjectPeople, where he was vice president of business development. Prior to ObjectPeople he was with Object Technology International (OTI) - the company that IBM acquired and that eventually built Eclipse, so this move is sort of a homecoming for Mike.

 
Mike Milinkovich Executive Director Eclipse Foundation

The interview generally covers four areas. First is SWT and its impact on the Java market. Second is Eclipse as an open source community. Next, the WebTools project, and finally I asked Mike about JSR 198, the JSR to define a common plugin API so plugins can work across IDEs.

JDJ: Does SWT bifurcate the Java market?
Mike Milinkovich: I know there is a perception in the marketplace that Eclipse is primarily a Java IDE. But in many ways Eclipse is and always has been about more than just Java - more languages and more platforms.

Part of the rationale behind supporting SWT and now the Rich Client Platform (RCP) was not to bifurcate the Java market, because our focus was on building tools that were of interest to people who perhaps were not interested in Java at all. We're interested in tools for other languages like what we're doing with the CDT for C and C++, and for other platforms and for other things like modeling and tools for the embedded marketplace. These areas are quite different from what you typically think of when you think of Java tools. So from that perspective, I think SWT is here to stay.

However, I don't want this characterized as a Sun Java versus Eclipse scenario because I frankly just don't see it that way. We are an open source project; we are trying to build a tools integration platform that can be used universally for many languages and many platforms. We are happy with what we are getting from SWT.

JDJ: SWT runs fast on Windows, but on other platforms (Mac OS X, Linux, etc.) the performance is not as good. What plans do you have to get the performance on other platforms more in line with the Windows platform?
MM: I'm too new in the job to bring a huge amount of perspective on how we got to where we are and what the processes are that are achieving the results we are getting. I think as a general comment, though, that we are looking for more involvement from the Linux people, be it Novell or Red Hat, and from other places like Apple to help us improve the tool set on those platforms. I think there's probably a lot of performance tuning that we could do but the best way to go about that is to get deeper involvement from the people who own those projects.

JDJ: That leads me back to the bifurcation of the marketplace: Apple and, I believe, Red Hat as well have invested heavily in making Swing run well on their platforms - now you are asking them to reinvest in SWT.
MM: It's actually news to me that Apple and/or Linux did anything on their platforms to optimize for Swing. I think they may have helped the Swing people improve performance on those platforms. I'm not aware of anything that, for example, Apple would have done to specifically improve performance of their platform to make Swing run faster. It might be something that I'm just not aware of.

JDJ: Doesn't the presence of the RCP bifurcate the market because of the requirement to build the user interface with SWT? For example, tools such as the Visual Editor (an Eclipse subproject) are having significant work poured into them to make the SWT aware. Doesn't the RCP/SWT combination bifurcate the market, even if that bifurcation is unintended?
MM: No, it doesn't. I'd like to point out that one of the key features of the 3.0 platform is SWT/Swing interoperability, and it was the investment on our side that made that happen.

JDJ: What about the rest of the platforms that do not support this SWT/AWT bridge?
MM: Sometimes it takes more than one release to solve all problems.

JDJ: Any other comment on SWT?
MM: I just want to reiterate what I said a little earlier. I believe that SWT is here to stay. I don't believe that statement should be characterized as trying in any way to split the Java community. We are an open source project doing what we need to do for our project to be successful.

JDJ: Moving on then, can you shed some light on what the status of the WebTools project is? At EclipseCon there was a lot of excitement about this project, but since then it seems to have died down.
MM: I'm too new in the job to give any history since EclipseCon. The fact that we are finally getting this project off the ground is very good news. That is certainly an area I intend to put some energy into, to make sure it gets off to a good start.

JDJ: Any kick-off date in mind? When might we see some of the output?
MM: No I don't know yet about the output part. I believe the kickoff has actually been posted on the Web site and the project is following the Eclipse development process. We are in a 30-day period of proposal review that ends some time in the next couple of weeks.

[Author's Note: The proposal was posted on April 27 and the 30-day period was extended by one week to finish assimilating feedback from the community. Look for the project to become official very soon. The WebTools Project (WTP) awaits a board vote.]

JDJ: We are 2.5 years into Eclipse. How do you feel the model is working for IBM and the other commercial entities that are involved?
MM: I think IBM is actually very happy with the payback they have received on the project. Of course I'm not an IBM spokesperson so I can't comment for them, but I think that they view Eclipse and the momentum it has in the marketplace as a very positive thing. I think that's a proof point of the notion that when it comes to open source and commercial participation in open source, the more you put in the more you get out of it.

JDJ: Do you see this as a trend in the industry? For example, BEA recently announced Beehive.
MM: I can't speak for other companies but I do believe that there will be more companies doing this. I believe they look at the success of Eclipse as an open source project driving market acceptance and see Eclipse as an exemplar. I would point out also that we are very excited about what BEA is doing with Beehive and we fully intend to be working with them down the road to make sure that the effort they are putting into their project will work well with Eclipse.

JDJ: Do you see Eclipse's dominance of the market as good or bad for innovation?
MM: I think it's wonderful for innovation. Eclipse is open source and therefore open platform. I can't see how, by growing the community of people who are contributing to an open source project, Eclipse could possibly be stifling innovation. I think there is a lot of interest in Eclipse; there is a huge amount of effort going into not just adding to the open source platform as it exists today but growing entirely new open source projects.

Just as interesting is building a vibrant commercial ecosystem of people building products on top of Eclipse or building plugins that work with Eclipse. People are making businesses building software in that way.

I think Eclipse is a platform for innovation and in no way stifles innovation. And I think that is part of the original vision for Eclipse. From the beginning, Eclipse was about making a common open source platform that solves problems people have been solving repeatedly in proprietary commercial implementations. By creating an open source and open implementation of the previously commercial, proprietary implementations, it allows people to go and innovate and solve interesting problems instead of building yet another text editor or yet another code browser.

JDJ: In which ways do you see the Eclipse market broadening? What's the most important area for it to broaden into?
MM: In my own mind I don't have any one particular area. I would go back to what I said earlier, that Eclipse is and always has been a universal platform for tool integration that goes beyond just solving Java developers' problems.

Where would I like to see Eclipse grow? First of all, we will continue to be very active as a Java IDE and you will see that in Eclipse 3.0. This next version will solve the problems that Java developers want solved. We will continue to grow in that area.

When I look forward into other areas where I want to see the technology grow, embedded systems is certainly one area I'm interested in seeing Eclipse grow. If you look at the board of stewards, three members - QNX Software Systems, MontaVista Software, and Ericsson - are on our board because they are interested in seeing Eclipse grow in the embedded space. The fact that Wind River joined as a member is another example of the embedded space companies joining to see Eclipse grow in that market. So I believe that growing Eclipse as a standard tool set for developers in the embedded market place is an area that we are very interested in.

Other technology areas will include the area of modeling and model-driven architecture. These are areas where Eclipse has got some traction today and I'd like to see it grow quickly in the future.

This list is not in any particular priority. As I mentioned before, the WebTools project is an area where I really see that Eclipse can and should grow to make sure it has a complete tool set for the J2EE developers who are out there. By the way, I'd add that in and of itself this is a pretty impressive list of things that we are already into that we will now be trying to grow.

JDJ: At EclipseCon there was concern that in adding the WebTools project you'd be cutting into some of the margins and feature sets of some of the commercial members of the Eclipse community. Do you have any comment on that?
MM:: I've spoken to a number of the add-in providers that are out there. I believe the consensus is (I've not spoken to anyone who disagrees with this) that Eclipse will continue to grow; it needs to continue to grow. We can't say that because a particular add-in provider is in a particular market segment or solving a particular problem that we can never have an open source project moving in that direction.

What there is a strong desire for, and I believe this a very fair request and one that I certainly intend to honor, is that there be visibility and communication about the direction Eclipse is going in. If you read the by-laws and development process there is a huge emphasis on having a very transparent strategy-setting and planning process. There is visibility from the add-in provider community on where Eclipse is going so they can make forewarned and rational business decisions about what they want to do and what they want to focus on.

JDJ: Do you have any particular areas you want to see Eclipse grow in that we have not already discussed?
MM:: It will focus on the areas we have already discussed. The one point I want to make is that right now Eclipse actually has excellent brand awareness and a lot of recognition in the marketplace. But my understanding from the people I've spoken to is that if you go out into the IT shops, the enterprise development shops, there are people out there who don't necessarily know that much about Eclipse. One of the areas we will focus on is developing marketing programs to target those developers. I want to see these developers become more aware of Eclipse and its benefits.

JDJ: How do you see the relationship between Eclipse and Sun in the future? Do you have a particular direction you'd like to see this go? Are there currently any ongoing talks?
MM: There's nothing in particular in the works right now. I think to a large degree those talks were understandably on a hiatus while the executive director was being recruited and the Eclipse Foundation re-created itself as an independent entity. There are no talks that I'm aware of.

That said, from my perspective the door is always going be open. If there is something Sun wants to talk to us about, I will always respond to their phone calls and have good-faith conversations about what we could be doing together. I personally think the entire conversation about Sun versus Eclipse is kind of missing the point. Sun is doing what they are doing in support of their shareholders and in their role as the steward of the Java community. Eclipse is doing what it is doing in its role as the steward of the Eclipse Open Source community. And to the degree possible, where we have congruent goals, we should be working together and I'd be thrilled to do so.

JDJ: Do you see a place for you to go after Sun to get some cooperation between the Eclipse Foundation and Sun?
MM: I certainly plan on making a few phone calls and introducing myself. Ultimately it takes two to tango and Sun is going to have to decide whether it's in its business interests to reopen the conversations with Eclipse. That will be their decision to make.

JDJ: IBM is on the JSR 198 expert group but there is currently no one from Eclipse. Are there plans for a member of the Eclipse community to join the JSR 198 expert group?
MM: Let's bear in mind that, up until the formation of Eclipse as a not-for-profit corporation, under the JCP process there was no way for Eclipse to have a representative on that JSR expert group. I don't know about trying to get someone on that group - it will be something that I will have to look into after I've been on the job for more than a week!

JDJ: Thank you Mike for talking with the world's leading i-technology magazine. We'll be closely watching the continuing story of Eclipse as a universal platform for tool integration.
MM: We'll always be here to answer whatever concerns and questions the technology community may have.