Juno Performance

There has been a lot of conversation this past week about the state of Juno’s performance on the new Eclipse 4.2 platform. The original thread was kicked off by Thomas Hallgren, there’s been a flurry of comments on bug 385272, and a nice summary of the concerns in Andrey Loskutov’s blog.

I think it is apparent there are a lot of very valid concerns that will need to be addressed, and we as a community will need to pool our resources to ensure we succeed. However, there are two issues I’d like to address: 1) moving to a major new release is always hard, and 2) we need the community to help solve the problem.

Moving to a new major release

For a little historical perspective, anyone who was around when Eclipse 3.0 shipped in 2004 can likely tell you that moving to a major new release is never ever easy. It is absolutely impossible to do a major rewrite of a platform used by millions without pissing somebody off. Nor is it possible to achieve perfection the first time around. That is not said to dismiss any of the concerns that have been raised. We need to have the feedback, bugs and patches to move things forward. However, we do need to move forward.

My personal summary of where we’re at is that Juno is on 4.2, and it’s far too late to talk about switching the release train back to 3.8. Kepler is going to be based on 4.3. There will never be an Eclipse 3.9. Those decisions were made quite some time ago by the Eclipse PMC and the Planning Council, after long discussions about the pros and cons. That’s where we’re at. So if you want to help move Eclipse forward, please focus your energy on providing input, feedback, patches and tests on the Eclipse 4.x stream because that is unequivocally our future.

Significant community help is required

The performance test were turned off because the Eclipse platform team has a serious resource issue. The simple fact of the matter is that the Eclipse platform team is stretched well beyond what it can reasonably be expected to accomplish. This is not a new problem. It has been discussed in many forums for at least the past three or four years. Unfortunately, very few people or organizations have stepped up to make significant contributions. Perhaps this will serve as a catalyst for people to step forward to contribute more to the needs of the platform. Let me stress how important it is for this to happen.

There have already been a couple of very positive developments that have come out of this conversation. Google has stepped up and donated $20,000 towards the creation of a brand-new testing infrastructure for eclipse.org. A few people and companies have quietly approached us about where they can best put their resources to help. The entire staff of the Eclipse Foundation is going to pitch in to help resolve these issues as much as we can. However, there are thousands of companies and millions of developers that make use of Eclipse every day. We need more of these companies to come forward to start participating in the core Eclipse platform. Google’s contribution is a perfect example.

The Eclipse 4 platform has a large number of important new features and capabilities which are going to allow us to do many more cool things in the future. The Eclipse platform team has done a great job bringing it this far. It is not perfect and we have the SR1 release coming up in a couple of weeks that will address some known performance degradations and memory leaks. I expect we will continue to see more improvements in SR2 and Kepler. The critical issue is that the entire community needs to be part of the solution. Now is the time to step forward and give back to the Eclipse project that started it all.