Ok, I am fine for trying to target graduation for the reasons Alex has outlined.

My main hesitation was that Felix is in a pretty stable state right now, so I thought it would be worthwhile to get a snapshot released to help grow the community for those who find SVN to be a barrier for experimentation. As part of this, I was basically in feature-freeze mode for the framework, since I didn't want to start hacking some new features that would make it unstable before trying to do a public release.

So, we can just forget about a release now altogether and just keep progressing on new features. Another possibility, I guess, is to make a svn branch of the current stable state of the framework that could be the basis of a 0.8.0 release after graduation.

Of course, it seems that much of this depends on the time frame and actual possibility of graduation too. We have to defer to our mentors to know whether or not we are in a state where we should actually push for graduation. Also, we need guidance to know what we need to do to prepare for it as well.

-> richard

Alex Karasulu wrote:
Richard S. Hall wrote:
Upayavira wrote:
It just means that whatever you release has more standing. It is an
official Apache release. More people thus are likely to try it, so
you'll get better feedback. I guess that's the point in it.

Yeah, I guess I understood that a little.

What I was wondering, is how this changes our focus? Right now we am focusing on creating a release. If we decide to think about preparing for graduation, do we stop focusing on that and start focusing on something else? If so, what is that something else?
The incubator release process is more involved.
=================================

Well targeting an incubator exit does not necessarily change your primary focus of getting a release out. Graduating first may make it a bit easier to get that release out since our PMC (once formed) would vote on the matter rather than the incubator PMC. Being closer to the code and understanding the functionality I think this community can make a better quicker evaluation than the incubator PMC.


Graduation will delay an initial release but not by much.
========================================

Graduation may however delay a release a bit since we would have to wait until the next board meeting to present our proposal as a resolution for the board to ratify. Again as Upayavira stated, the value of your release after graduation would be greater than if you did release from the incubator. If there is some doubt we can always just take a vote on what to do here. Also you can still focus on tying up loose ends for a release even if we do shoot for an incubator exit.

Either way in the grand scheme of things I don't think there will be a net difference to Felix in the long term. Personally I don't think it's worth the effort now to release if the release is not an official Apache endorsed product. I'd keep working on taking care of loose ends while trying to graduate this project. Those who want to build with Felix can still do that using sources from the repository and can await graduation then shortly there after an official release which will be imminent.

Other projects cannot *easily* use an unofficial incubator release.
==============================================

Also I'm a bit biased here because we at Directory would like to use this release (official release hopefully) to release a Felix based OSGi version of ApacheDS in the coming few months. Otherwise we have to start long threads on the incubator list discussing the legal aspects of whether or not we can release ApacehDS with dependencies on an incubator release of Felix. No project has yet released a product based on an incubator release of another dependent project so it's uncharted territory which would require a good amount of discussion. Also the outcome is undefined since this is a relatively new situation for the incubator to establish policies around. The incubator release will take time and so will this conversation. Graduating Felix first then releasing it would be a much clearer path for those who wish to build upon this initial Felix release. This shows just how important an Apache endorsed release is verses an incubator release.

Alex


Reply via email to