Last week I had a long chat with Michael Meeks of Novell, he of the (in?)famous blog post about the stagnation of OpenOffice. The post itself has been chewed over and thrown around by so many other big dogs, I thought I'd go right to the man himself and ask him some questions. The biggest one was this: How is it that one of the biggest success stories in the open source software world (at least in terms
Last week I had a long chat with Michael Meeks of Novell, he of the (in?)famous blog post about the stagnation of OpenOffice. The post itself has been chewed over and thrown around by so many other big dogs, I thought I'd go right to the man himself and ask him some questions. The biggest one was this: How is it that one of the biggest success stories in the open source software world (at least in terms of recognizability) is getting so badly bogged down from within?
Michael cited a few things. The first is Sun's own internal development culture, which is something I've pounded on more than a few times myself. They're not exactly a Johnny-come-lately to open source, but so much of how they deal with open source on the day-by-day, detail-by-detail level is problematic. Sun's control over the OpenOffice project is too tight, and -- as other people have suggested -- they would have nothing to lose by spinning off OpenOffice into its own independently-managed foundation, a la Eclipse.
A second thing is what I think could be called the Too Many Evangelists syndrome. The way Michael put it, the major stumpers for Linux, like Alan Cox or Linus Tovalds, are themselves programmers. "With OpenOffice, the exact opposite is typically the case. Most the leads have had intangible contributions, and I think that's a big part of the problem with OpenOffice in terms of attracting developers -- that there are so many people who are not developers who are also very eager to tell everyone what to do." The code base isn't even the real issue, in his purview: every project has a potentially messy, outsized code base.
When Michael cited this as being an issue, I came back with a counterpoint: Maybe this is happening because the people who have their heads down, writing code, are not the ones most attuned to what users ask for and what they need or want. His take was a little different -- what Sun had created here was a community that didn't attract developers, but the next best thing (e.g., evangelists). Those who cut their teeth in the community by contributing code first can then graduate up to a position where they can evangelize in a truly informed way.
So why isn't this happening? For reason #1 above: Sun's control over the project isn't allowing the community to take the shape it needs to complement the work that needs to be done on it. Jeff Atwood said something analogous to this in his own discussion of Meeks's post: "...if you're having difficulty getting software developers to participate in your open source project, I'd say the community isn't failing your project. Your project is failing the community." From what I've seen, that sort of failure is a revolving door that goes in both directions.