Jaikiran,
Thanks for the confirmation.
My only concern is the functionality gap between Weld-int and EJB3.1 in JBAS - i.e. developers can create no-interface views but then they cannot inject them, which gives way to the perception that "EJB injection doesn't work in CDI". If we can close the gap by M4, it would be great - so that's the main reason why this would be a "very good thing to have". Anyway, as soon as the functionality is exposed in the resolver, I can make sure that it's integrated correctly on our side (if it needs to).
Btw, how do you plan on dealing with this? Supply the bean class as the "interface" of the EJB reference ?
Marius