On 2/23/06, Mike Kienenberger <[EMAIL PROTECTED]> wrote: > On 2/22/06, Sean Schofield <[EMAIL PROTECTED]> wrote: > > In response to #1 I would say we do not need to be in the business of > > ensuring developers can rely on our public API's. From my perspective > > we are in the business of providing a JSF implementation and a series > > of components and addons for use in any JSF implementation. > > And how are we going to develop components? By reinventing the wheel > every time? > > There's certainly a class of common building blocks that can and > should be considered a public API. > > Why should every single component developer have to reinvent fetching > localized messages? Or evaluating value bindings in a consistent > manner?
No, this is exactly the reason why there will still be a (new) myfaces-commons lib in the future. Now, closer looks at the code and current state of the myfaces commons subproject have shown that the legitimate expectations (stable API, clear separation of API and impl) cannot be fulfilled now. We had some discussions before on calling this subproject "commons" or "shared". We decided wrong at first. I admit that I did participate in this wrong decision as well. But we have learned our lesson and now go back to the start. What is now in commons will be moved to myfaces-shared. Myfaces-commons will be flushed. What is worth will be moved back to commons ASAP. But with more caution and not before 1.1.2 release. Hope, everyone feels ok with this plan. Manfred
