Matej Novotny

The JIRA accessibility issue should now be solved - it can be viewed again even without logging in. It still (intentionally) remains read-only.

Ladislav is right that `CDI.current.get()` shouldn't be your go-to hot path for resolving beans. Standard way for dynamic resolution, assuming you cannot use classic `@Inject Foo`, is to use `@Inject...

> Maybe, you can help with this question: Is there another way to access service bean methods from entity beans? Thanks in advance. Hmm good question, the injection doesn't work...

> I was not expecting that this tiny shortcut feature request is even worth to have a long discussion since it does not harm or change anything in the architecture....

> Of course, my expection was that CDI.get(Foo.class) throws an unsatisfied dependency exception when the argument (here: Foo.class) does not exist. > without prior checking whether the instance is resolvable...

I do not have Mac to verify this but if it's indeed the case, I see no harm in adding a small note into the doc. Feel free to send...

Notably, CDI bean names are not always linked to EL (but can be). As stated in https://jakarta.ee/specifications/cdi/4.0/jakarta-cdi-spec-4.0.html#bean_discovery_steps_full: > A bean may have a bean name. A bean with a name...

> @Ladicek think the key point is even if there is a SPI the integration points should be consistent and there CDI uses a hack assuming it will be alone...

> So have several options here. First of all, we should decide if we really want to use CDI events or some other kind of API (listener interface, Vertx eventbus,...

> The only downside I can see is that for "close" events the connection object is basically unusable and most of the operations would result in an exception. Yes, at...