On 01/05/2011 12:53 PM, Maurizio Cimadamore wrote:
On 05/01/11 11:23, Alan Bateman wrote:
Hi maurizio,
In my opinion, XlintWarn should also test invokeGeneric().
Moreover, it will be better to have one test by invoke*() instead of
having one
test for all invoke*()
Rémi
Hi Remi
I agree on your comments; however note that the aim of the XlintWarn
test is to check that the compiler issues warnings in relation to
old-style usage of 292 target typing - since the old syntax is going
away (see related cleanup CR: 6992698). In other words, the aim of
this changeset is to make tests pass again (with b123) awaiting for a
more precise cleanup (which is needed since InvokeDynamic and other
292 features should be removed from the JDK 7 repo).
Not all syntaxes introduces by JSR 292 should be removed
MethodHandle.invokeExact()/invokeGeneric()
will stay. Otherwise, java.dyn package will not compile anymore.
And in my opinion other features should forward ported into lambda
workspace,
but that's a different story.
Maurizio
Rémi
