On Tue, 16 Jun 2026 07:37:23 GMT, Alan Bateman <[email protected]> wrote:

>> On the contrary, I would argue for making requireIdentity a reflective 
>> preview API - this requirement makes sense for programs running out of 
>> preview too, that they want something with a monitor instead of null or new 
>> kinds of monitorless objects.
>
> I wonder if Roger has any notes on "why" the requireIdentity methods are 
> normal preview, and this one is a reflective preview API. If there are 
> frameworks or library looking to have early support for value classes they 
> might need want to use requireIdentity. When virtual threads were in preview 
> then we noticed the frameworks were accessing the preview API reflectively 
> (as they want to build/run across a range of releases) and thus it didn't 
> matter if the API was normal or reflective preview.

Is there a resolution here, or do we need a follow-up, post integ bug filed ?

-------------

PR Review Comment: https://git.openjdk.org/jdk/pull/31123#discussion_r3544757005

Reply via email to