|
Your quote even says there IS a bean:
that the second BEAN is never instantiated or called
Still this doesn't say anything about BeanAttributes. Btw, there is still a certain mixup between Bean<T> and it's contextual intance left in the spec...
There is also another issue open which affects this behaviour. In ProcessBeanAttributes you can change the attributes in a way that the specialization chain gets changed. And thus we NEED to invoke it. If this isn't done in Weld, then you might have a hole in there. At least in theory, as I'm not sure if there are many extensions who rely on such nerdy tricks ![]()
So I really think this is minor minor minor, but still the TCK should not test this. The spec simply doesn't say anything specific about it.
|