This bugfix is similiar to problem reported by WELD-1155.

In Java SE the following occurred:

• One thread (T1) is resolving an ordinary build-in-scope-proxy that has not yet been initialized.
• The production take place in another thread (T2). In our application this is caused by a special producer methodology. One transitive @Injection Point is a @SessionScoped bean (SB). This @SessionScoped-Bean (SB)is produced correctly.
• Then in the origin Thread (A) the code passes the proxy related to (SB). (SB )gets produced again for the identical session !

The problem is caused by a similiar fixed problem reported by WELD-1155.

AttributeBeanStore#attach is copying some Maps (in Thread A). This copies are not changed by production (in Thread B).

The bugfix for WELD-1155 has taken place in a http-related context that is not used in Java SE.

Is it possible to move-up the bugfix into the base-implementations that are currently also used with Java-SE ?