Ladislav Thon

Personally, I wouldn't expose the assignability rules directly, as that's too low-level. What might be interesting is to expose the resolution algorithm in a more fine-grained manner. For example, `BeanContainer`...

Just thinking about the use case here, an alternative solution could be: ```java /** * Obtains the {@linkplain Context context object} of given class for given * {@linkplain jakarta.enterprise.context scope}....

I used `ObserverMethod` in the proposal sketch above with a bit of hesitation, as indeed it contains way more information than necessary. If the method only accepted `Type, Annotation[], Type,...

I believe the original use case still requires the knowledge of event types and perhaps qualifiers, and implementing the `ObserverMethod` interface where most methods throw `UnsupportedOperationException` is pretty straightforward :-)...

Sure. We can either add an interface, as I mentioned above, or just accept types and annotations. Agree about the delegate injection points, that can't be practically useful.

The javadoc in my `getContexts` proposal above directly says "all context objects, active and inactive, for the given scope".

If we ever wanted to unify CDI's `TypeLiteral` and JAX-RS's `GenericType`, we'd have to solve that problem too, as `GenericType` can be created out of a plain `Type`. It's not...

I think that was meant in the opposite direction :-) A constructor accepting `Type`.

I'm not proposing this for 4.0, just figured we should have a tracker for it.

I was thinking about this and realized that we need to deal with a few things. ## Enablement - https://jakarta.ee/specifications/cdi/4.0/jakarta-cdi-spec-4.0.html#enablement Currently, a producer is enabled if the declaring bean is...