Basis of the Eclipse phenomenon
That's my response to a recent thread of blogs about the integration of Eclipse tools, whether to favor platforms over tools, and the future technical direction of the Eclipse Foundation.
Platforms didn't attract 50 million downloads. 50 million downloads attracted the platforms! Eclipse became popular simply because it was a powerful and free Java IDE. Everything else, while it certainly has some interesting potential, is and should be secondary to the power and capability of those user tools.
Case in point: NetBeans is becoming more and more popular because it is a free and increasingly powerful Java IDE. Sound familiar? Plus, it has a powerful and free GUI builder, a powerful and free profiler, and by the way it also does a fine job of Web and Mobile development. Plus it integrates them all together like a single unified product. These are all things that the users care about, and 99% of them will weigh the product solely on these criteria.
So what happens if the market share flips and something else is used by more developers than Eclipse? Plug-ins follow the users. So do books, articles, etc.. I'm talking about mindshare here folks. The corporate interests, the add-on developers, the people that use the platforms--they will follow. Don't worry about them so much.
Yes, good solid platforms are important. Just like good architecture, good API documentation, good OO design, and so forth. Eclipse has all that in spades. But the users don't see that. They see the quality and ease of use of the final product, not the guts that enable it. And the users are the important ones here.
Eclipse is at a precarious point in its life. It's hugely successful, but at risk of getting complacent and taking all those users for granted. Pride goeth before the fall as they say. The Eclipse community, especially the decision makers, should step back and look at what made Eclipse great in the first place. The tools! Keep doing more of that and you'll be fine.