We're now entering the next phase of OpenSolaris. It's been a long time coming -- and we're not done yet -- but the availability of this much code is a huge step in the process.

It has been an enormous undertaking, and I am constantly impressed with what we've accomplished to date:

• We've created a new license that we think is business friendly and will help us build a thriving community and an even better product.


• We've created a website that helps us start an ongoing conversation with members of the community.


• We've started to drive cultural change within the engineering team. The entire engineering team needs to be part of the community, and they are helping everyone get there as quickly as possible.

I probably haven't said enough to convey how big an effort this has been, but I'm not sure I have the words to express how much everyone has done. I can say that we're just getting warmed up, though...

Our Pilot community has been so helpful and dedicated, but there was only so much they could do to build a community in an NDA-constrained program. It was definitely time to move to the next phase and do our first big code release. We've made countless priority calls to get to this point, so you may need to check back to see if your code/feature/forum is available after launch.

You aren't going to agree with all of the priority calls we've made for this release. I don't agree with some of them, but I stand behind every one of them. More importantly, I am the person who will take all of the blame and none of the credit for making these decisions. With this launch everyone has the opportunity to provide input on what our priorities should be going forward.

I should probably focus on all the positive things we've done for the launch, but I'd rather do some 'splainin' about areas you may find frustrating:

• Having only the core kernel, networking, libraries, and utilities code available
  We had to start somewhere, and our (admittedly biased) opinion was that the core was the best place to begin. We have lots of evaluation to do of our current, internal delivery mechanisms for all the other areas of Solaris, and you'll want us to do that before dumping lots more code into your laps.


• The amount and type of core code being released in binary-only form
  This is a starting point, and we're working as quickly as possible to make more code available. I know it seems straightforward, but the level of complexity involved here makes my head hurt. Thank goodness Bonnie keeps it all straight, and I hope she'll soon have time to tell you more about it in her own blog.


• A daunting development process/model
  While we absolutely want to make our development processes more "community friendly," we don't want to just throw away our current processes. Good, bad, or indifferent these processes have helped us manage 1,000+ developers delivering to a single code base. Having sat through the meetings, I'm sure Al Hopper will confirm that there is actually some thought that was put into each step. I don't think anyone would say that there isn't room for improvement, though.


• Limited website functionality
  We want to build something truly great with the website, and want to be sure that the tools we use match our development processes rather than the other way around. We need your input to help define the way the tools should work.


• Lack of historical technical information
  You'll see more and more historical information available over time. We need to scrub documents and other historical information for confidential materials before we can make them available. I certainly wish we could have people in the community help with this, but this is one of the areas that someone within Sun needs to do the work. You can certainly help us prioritize by asking for specific documents/information. We are already prioritizing documents for scrubbing, and your input will be helpful.

I may not have convinced you, but I hope you have a sense that we do "get" open source. The choices we made, imperfect as they may be, have the best of intentions behind them. We aren't trying to hide features, code, or our decisions, though sometimes we have legal reasons for holding back.

I know most of the team would have liked this initial release to be even more complete, but (again) this is just the next step. My belief is that releasing this portion of code now will help us get to the next steps.

Thanks to all the amazing people inside and outside of Sun that made this launch possible. It really has been a team effort, and I'm proud to be part of it.

You're probably wondering who the hell I am at this point, since I haven't written a weblog entry before. I'm the engineering manager for the OpenSolaris project. I work directly with Jim Grisanzio, Keith Wesolowski, Mike Kupfer, Stephen Harpster, and (of course) the larger Solaris development organization. Before that I was a program manager in the Solaris kernel group. The last project I managed was Zones -- working with David Comay, Dan Price, and many other engineers. Before then I worked on Solaris Resource Management projects working with engineers like Stephen Hahn, Andrei Dororfeev, Liane Praza, Joost Pronk, and many others.

Words can't express how amazing this group is. With OpenSolaris, you'll get to find out for yourself.