[jmm-dev] ECOOP, JVMLS

[John Rose]

On Aug 3, 2014, at 5:36 AM, Doug Lea <dl at cs.oswego.edu> wrote:

> * Everyone seems to strongly favor the idea of replacing final-field
> specs with those that amount to guaranteeing release fences in
> constructors.  Still lots of details (some depending on core model) to
> work out though.

This change makes it easier to define interesting variations on final variables.  Such finals would be set some time after their constructor exits.   Scala-style lazily evaluated fields are an example; also no-copy builder APIs for immutables.  Finals could also be elements of monotonically initialized constant arrays.  A final set outside of a constructor (or final array element set) would be followed immediately (or on exit from enclosing method?) by a release fence, giving safe publication for the final value.  (A reader would have to test carefully for the empty state, if the object was published beforehand.)

— John