Life at Eclipse

I get to join Ian at the Embedded Systems Conference (ESC) this week. I spent several hours today walking the exhibit floor looking for Eclipse-based tools and products. They’re everywhere! I am really miffed with myself for forgetting to bring my camera.

Just a partial list of companies showing Eclipse-based tools are: QNX, WindRiver, Aonix, AMD, DDC-I, ENEA, IBM, KlocWork, LynuxWorks, Mentor Graphics, MontaVista and TimeSys.

The icing on the cake was discovering that Eclipse was the cover story for the issue of Embedded Systems Design being distributed at the conference. We had no idea, so it was a very welcome surprise.

I know that many people associate Eclipse with enterprise development tools, and more specifically with Java development, but the number of companies in the Eclipse ecosystem engaged with embedded and device software is massive. For most of these companies, the ability to build their tool chain by extending the Eclipse Platform and/or the C/C++ development tools (CDT) is a great enabler for their platforms.

Many thanks for Doug Gaff (WindRiver), Anders Florin (ENEA) and others who helped staff the Eclipse booth today.

Last year for the Eclipse Community Project Spot Light panel, we collected questions from the community to ask the PMC leaders. If you would like to provide us with a question, please do one of the following:

1. Add it as a comment to the talk.
2. Send me an email at mike at eclipse.org
3. Write it on a piece of paper and give it to me when I’m wandering around. You can likely find me at any reception offering free beer 🙂

Last year’s panel was a lot of fun, largely because of the great questions from the community. Please take a moment to contribute.

So here are my personal EclipseCon 2006 highlights so far:

1. The Sun team promoting NetBeans out front. Hilarious! I’m proud to say that I’ve got my full allotment of NetBeans kit: one CD and one bottle of water.
2. Joel Spolsky’s keynote. It’s been a long time since I’ve laughed as much. What a great start to the conference.
3. Watching the woman from Microsoft work on recruiting the RadRails guys who won the Eclipse Award for “Best Open Source Eclipse based developer tool”. She was on them in about 30 seconds flat. Hey Ed, have you been recruited yet?

I have to admit that I really enjoy EclipseCon. OK, so I’ve only been to one of the two so far, but last year really was a killer conference. As of last night we were just about at 1200 registrations, so we know for sure that this year’s event is going to be an even bigger group.

But as good as the program is, the real reason I enjoy EclipseCon is just seeing everyone and getting caught up on what folks are working on. By far my favourite activity is wandering around just chatting with people. (Actually that’s my favourite part of any conference.) The Eclipse community is just such a fun place to be, and EclipseCon is where you get to see it all in one place.

So if we bump into each other at EclipseCon, please say “hi”.

I have seen more than a few references to Tim O’Reilly‘s post on IDE buzz. I actually thought that Tim’s post was well written and thoughtfully speculative. No one can reasonably take issue with statements like:

I don’t really follow the IDE wars, but this result surprised me, so I thought I’d ask my readers if you have any ideas about this result. It could be an artifact of the Yahoo! user base, for example, with very different results if we were able to run the same exercise against Google search volume…. So I’m not sure I’d take it as the final word on the popularity of the two IDEs.

But how I’ve seen Tim’s speculations used since have not exactly been to the same level of quality, if you know what I mean.

Two things to think about:

1. First off, Eclipse (the blue line) has now moved into the number one spot on the graph. The point here is not that Eclipse is #1. It’s that no one really knows what makes these indices go up or down, and as Tim himself points out, no one knows if there is any correlation between the numbers and actual usage. Buzz is good fun and all, but great technology, adoption, ecosystem strength, predictability and open governance are all examples of things which matter much more IMHO.



2. Secondly, as shown by a very small experiment we tried, many of these games are easy to game. For example, on the same page as the Buzz meter, you can find a pseudo stock chart for IDEs. A few days ago, we noticed that Eclipse was not the most highly valued of the stocks. Notice that it is today. Why? Because six of us at the Eclipse Foundation went to the trouble to register accounts and use our mythological $10,000 to buy Eclipse shares. We moved Eclipse into first place and each made a tidy “profit”. The fact that just six people can do that clearly means that this market is lacking in liquidity. Which is finance-speak for it’s a sucker’s game.

The point here is not pick on any particular metric, but to hopefully make folks a little more discerning the next time they’re reading claims of greatness.