Knox dev's - We need to start locking down the release for 0.7.0. In preparation of this, Sumit created a branch a week or so ago and we should start considering the creation of a release candidate.
I believe that I have to update the CHANGES file with an entry for a patch that I cherry picked into 0.7.0 branch and I will look into that shortly. Standout features include: KnoxSSO for WebSSO, HA support for numerous services, diagnostic commands for KnoxCLI, regex based identity assertion, better control over thread pool, connection queue and request/response buffers. The ability to proxy Hadoop UIs, CORS support for cross origin request sharing and more. As well as a number of important bug fixes. We do have an important feature coming from the community - specifically from Jérôme that will be committed in coming days. KNOX-641 adds a federation provider that integrates pac4j in order to add: OAuth, Facebook, CAS, SAML, OpenID Connect. I think that this is an exciting integration that will require a bit of testing before it can be merged into a release branch. In my opinion, the set of features and improvements that are currently in the v0.7.0 branch more than justify a new release and delaying that any longer would be less than ideal. Concentrating on defining and testing the usecases that the pac4j provider will bring to the table post 0.7.0 and coming up with a compelling story for that feature set can be used to justify a release of its own. I think that we should target a feature release which we'll call 0.8.0 for now for a mid January timeframe. So, discussion points: 1. Should we move forward with the 0.7.0 release once the CHANGES file is updated? 2. Thoughts on holding the pac4j provider out until an early 2016 release when the main usecases are better defined and tested? thanks, --larry
