Peter, First let me say that where this open source project is concerned, my expectations of Oracle changing its attitude have come to zero. And what one expects from Oracle will play a huge role as to how the external community of individuals and 3rd-party companies organize itself.
I think OpenSolaris is a great operating system and I would like to see it flourish as a true open source project which provides a great platform for collaboration with the end result of creating a high quality product. This is my 100,000-feet view of how I think things should be. We are not sure of what Oracle's plans are for OpenSolaris but they have given everybody signs that their closed Solaris product is the main focus of the activities. So I'm ignoring here any attempts to "fix" this project as a whole. That's not to say I wouldn't be open to actively participate in fixing it, just that I have a hard time creating a plan to convince Oracle of something it has clearly showed not desire to consider. That being said, IMO please, I think the external community has a few options: 1) Keep depending on Oracle We are already on this route and things aren't working from a dozens different points of view. I will not list all the drawbacks of this here since this has been discussed ad nauseum. 2) Create a new project which closely tracks OpenSolaris.org The goal would be to create a OS based on current OpenSolaris.org source code. This project could initially keep a code repository consisting exclusively of patches and new code that doesn't interfere radically with that provided by the upstream repository in core areas which have not been flagged for replacement. What I'm trying to say is that we could, for instance, have a new installer but would we want a new device driver layer if that's going to make merging code one order of magnitude harder ? Closed binaries would have to be accepted initially but with the medium-term goal of replacing them with open source alternatives. I'm sure that there are some device drivers which aren't easily replaced but I wouldn't be too radical in getting ride of them until decent alternatives show up. In the long term, if this project is to reach a critical mass of contributions that afford this, it could decouple itself from the upstream repository and work more like the BSDs which have a separate repository, different philosophies but are not prohibited to share code when one side thinks it's beneficial. I'm sure people would pride themselves if this happens but we have to be realistic that it could take a decade. Please note I'm not using the "distribution" terminology here because, at least to me, that only means the distribution of a different set of packages. But since we currently don't even have the development builds available anymore, starting off as a distribution would help to get momentum before slowly finding our path. The overall idea is to slowly start walking to be able to run one day. 3) Fork completely Here we would take the code as it is and not care anymore about the OpenSolaris.org code base. People would be freed to change code as they pleased (provided they following strict guidelines that we can't possibly discuss now) and would not worry if it would be easy or not to incorporate changes back to/from OpenSolaris.org. Personally I think we would be setting ourselves up for failure. We don't have the resources to fork and keep the same pace of innovation that other more mature community have (e.g.. Linux and BSDs). -- I see no desire from Oracle to do anything but to slowly kill OpenSolaris and exclusively focus on Solaris in a closed fashion. My point is that if I were Oracle with great open source plans for OpenSolaris and people spent a single day spreading FUD about it, I would not wait a second to come forward and show them how great my plans are. If I had no plans at all, I would have to options: come forward and state that I didn't have 'em or keep quiet and let people figure it out by themselves. Oracle is clearly doing the later. To state it clearly: I think option #2 is best suited for our reality. Start small, be humble, be realistic, do our best and have fun! If I'm entirely mistaken about anything, please let me know. Contrary to some people that have recently posted to opensolaris-discuss@, I don't even remotely expect to be right always. And I'm not trying to shove anything down people's throat but simply offer my view and learn something back. Regards, Giovanni Tirloni gtirl...@sysdroid.com -- This message posted from opensolaris.org _______________________________________________ ogb-discuss mailing list ogb-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/ogb-discuss
