Hi,

we probably have the TCK test in place because the spec explicitly
says the map is empty in those cases. see 12.3.2 Event Information.
on the other hand, I don't see a reason why an implementation shouldn't
be allowed to pass more information than required.

I'd say we create an issue for JCR 2.1 and propose a weaker contract.

Regards
 Marcel


On 08/07/14 17:52, "Michael Dürig" <[email protected]> wrote:

>
>Hi,
>
>Often it would be helpful if observation events would carry node type
>information. In Oak we added some initial support for this [1] by adding
>the node type and the mixin types of the associated node of an event to
>the info map.
>
>In JCR-3765 [2] I suggested to backport this functionality to
>Jackrabbit. However this will cause the GetInfoTest TCK test to fail.
>That test asserts that the info map is empty for all non move events.
>What do people think about adapting the test expectations of this test
>and allow for said entries?
>
>Michael
>
>
>[1] https://issues.apache.org/jira/browse/OAK-1661
>[2] https://issues.apache.org/jira/browse/JCR-3765

Reply via email to