Thoughts on JCP, JCRs and Many things
With all the troubles in the JCP community((1), (2), (3), and (4)), I thought I would provide some of my thoughts on the JCP, JSRs, RIs, TCKs, etc.

I believe:
(1) RIs and TCKs should be separate from the JSR and the JCP.
(2) JSR Specs (not RI) and Test Specs (not TCKs) should be provided by the JCP.
(3) Control of Java (TM) and Patents directly effecting a JSR should be turned over to the JCP.
(4) Additional patents in an implementation (not spec) would then be part of the RI/TCK and not limited in compliance of the Specs. So then any patent issues would be against the implementors (not the JCP).
(5) Oracle (Sun) should not be given a permanent veto controlling leadership position in the JCP. The leadership role should be elected and voted just like everything else.
(6) JCP (or a Validation WG) can be a validating body if they so chose to take up that role. Implementors can provide their implementation for acceptance against the Test Spec (not TCK).
(7) Commercial interests addressed would be addressed by individual, groups, and companies with their own implementations, independent of the JCP with their own licensing for their own implementation. The public could then identify which implementation they like best in the market.
(8) Oracle/Sun still owns (an open source based) OpenJDK code base implementation.

Reference:

• (1) Is JCP Dead


• (2) ASF Resigns


• (3) JSR for Java 7


• (4) IP in Picture