Producers only called for subscribers to exact type, not any assignable type about otto

[JakeWharton]

Nice find. This is a bit of an edge case, but one we indeed thought of when designing @Produce. The reason for this is that while post() is a one-to-many operation we don't want register()-->subscribe() to produce multiple invocations.

Take for example this case:

class A {
  @Produce public CharSequence provideChar() { return "char"; }
  @Produce public String produceString() { return "string"; }
  @Subscribe public void onObject(Object object) {
    System.out.println("GOT: " + object);
  }
}

While very obviously an extreme case, you can see where the problem lies. If there are multiple @Produce-annotated methods which subclass @Subscribe types it becomes ambiguous which to call (or it requires multiple invocations).

This is why we require explicit 1:1 mapping.

[grodin]

Thanks for the explanation, that makes a lot of sense.

If I wanted to submit a pull request to make this explicit in the Javadoc and website, I should target the 2.0-wip branch, correct?

[JakeWharton]

PRs should be against master.
On Mar 5, 2014 4:22 AM, "grodin" [email protected] wrote:

Thanks for the explanation, that makes a lot of sense.

If I wanted to submit a pull request to make this explicit in the Javadoc
and website, I should target the 2.0-wip branch, correct?

Reply to this email directly or view it on GitHubhttps://github.com//issues/106#issuecomment-36736893
.