State of javac support for lworld-values.

[John Rose]

On Mar 26, 2018, at 6:25 PM, Srikanth <srikanth.adayapalam at oracle.com> wrote:
> 
> Could you elaborate ? If the vm were to treat assignment of nulls as though it was assignment of default value for value types (migrated or brand new) - I was not thinking of an alternate representation with extra storage for nullness encoding, only effectively bringing about assignment by default value - why would we end up with three kinds of classes ?

See my previous P.S.

> Wouldn't it be just Objects and object-like values ?

I think we'd end up with three kinds of types:  Objects
and two kinds of values, those which promote from
null and those which don't.  Since ints already set
the precedent of not promoting from null, I'd rather
shut down the other alternative, just in the interests
of simplicity.  Maybe promoting null to default will
turn out to be useful in the future, but let's not
implement it before we are forced to.