A Solution for Sun’s OS Community Problems

Once again a Sun open source project is having a revolt in their community. Last December, Sun was accused of playing hardball with an ex-employee over the governance structure of OpenDS. Now it seems the OpenSolaris community is in turmoil due to Sun’s decision over the use of the OpenSolaris trademark and the resulting resignation of a Ray Fielding, a longtime Apache member who was trying to help the OpenSolaris community get started. Stephen O’Grady and Michael Dolan have some excellent commentary.

There is a very simple solution that I offer to Sun. Stop setting up community governance boards for your projects. They are causing confusion and setting the wrong expectations within the community you are trying to build. Instead, follow the MySQL/JBoss model and be a true open source company that employs all of the committers, sets the project roadmap, owns and protects all of the trademarks and is focused on making a profit from your open source projects.

There is nothing wrong with this. You actually spent $1 billion on a company that was very successful doing this.  Your communities and employees will appreciate the honesty. They will understand you own and intend to profit from the trademarks. Your employees will also know that their involvement in the project is due to their employment with Sun.

Right now it seems Sun is trying to be like Apache, Linux, or Eclipse but acting like MySQL or JBoss. Both models have proven to build great communities and successful open source projects. However, trying to be in between is just plain confusing.

11 thoughts on “A Solution for Sun’s OS Community Problems

  1. Man, I haven’t seen Dalibor’s name since the Kaffee days.

    Anyway… when Sun realizes that their shareholders include their community members and they aren’t just means to an end, things will change.

  2. I agree wholeheartedly with Dalibor on this, and I think it’s the core of the issue. Creating a governance structure (and constitution) while a fledgling community is forming seems a very, very unwise thing – let alone trying to shoe-horn in a governance model that has worked well for other communities.

  3. Glynn and Dalibor

    You might be right that bootstrapping a community has to be finished before the governance. However, I still think early in the process you need to be able to answer basic questions like ‘who can become a committer’ and ‘how are important technical decisions make?’

    Governance is also something that changes and evolves; in fact good governance should allow for and define how change happens.

    I really think Sun would solve a lot of their problems by being more upfront about a lot of these things.

  4. Sun is taking way too long to get many of those things done, in particular without setting the expectations accordingly, or getting the dependency chain of their goals figured out, and having the good sense to disseminate it widely. Goals change, priorities shift, etc. but if they do, talk about it with the community you are bootstrapping, don’t present it as a fait accompli afterwards.

    I’d hope that it’s just an issue of self fulfilling prophecy, where fear of massive public failure om every step in the way they open up their mission critical products is leading Sun’s management to a politician’s preferred strategy of carefully thinking things through, and picking the safest option: go for it slowly.

  5. Glyn, I think the core of the issue is a mix of many things:

    * institutionalizing governance before there is a community with a significant number of contributors outside Sun is a bad idea. Even if the governance arrived at is the best possible governance model ever devised by man, human nature dictates that contributors inside Sun will be tempted on every decision they make to take the short path of taking the discussion internally, rather than discussing it externally. Eventually including those decisions that matter to the community outside Sun.

    * finishing governance before there is a significant number of contributors outside Sun is a bad idea: it draws a line in the sand where everyone who was too shy, busy, proprietary etc. to contribute is declared to have missed out on having their interests represented. Anyone arriving later at the game then needs to figure out if they agree with the governance, or if they’ll take their time and effort to fight it.

    * You need a strategy to seed the community as widely as possible to include many different stake holders in the project, including those who pay Sun real, big money for open or closed Solaris. If your community and your customers don’t overlap at least a tiny bit, the community thing will be second priority to those that finance it.

    * aiming for 100% governance prior to aiming for 100% open source means you are looking for a community of members more interested in open governance than open source. You get the community you shop for.

  6. Though i am very late to comment, i would like to say one thing:

    Unless i misunderstand, many people (Glyn, Simon, Ian, etc.) are now sounding like there was a misunderstanding on SUN’s assumptions of Opensolaris and community’s expectations.

    I don’t believe Ray Fielding had accepted the constitutional work without understanding (read as: misunderstanding) SUN’s expections and promises about it.

    Its funny to see all these SUN employees starting to support this misunderstanding argument, *after* Ray Fielding has resigned!

Comments are closed.