template attribute definition possible bug?

[Cantor, Scott]

On 11/3/15, 4:17 PM, "users on behalf of Cantor, Scott" <users-bounces at shibboleth.net on behalf of cantor.2 at osu.edu> wrote:


>
>Hmm. I'll have to see what the dependency pulling code is doing on a non-existent attribute ID. Perhaps that is the bug. And of course if you have a dependency with only a subset of the attributes in it, that would be equivalent. I didn't follow your meaning, now I think I do.

I can see what it's doing, and I think it explains the behavior. I don't know if V3 does this, but there's that whole compatibility issue.

The trick is that the SourceAttribute ID is only relevant if the dependency is a data connector. If it's an attribute definition, the SourceAttribute doesn't matter, it simply pulls the values from all the dependent attribute definitions and merges them together as the value of the SourceAttribute.

That's what you're seeing. The gist is that layering Template on top of *any* AttributeDefinition is pathological.

My use case locally is on top of a DataConnector. In that case, the SourceAttribute ID is applied to pull that *specific* attribute out of the connector. Which is what you expect to get.

That's the "bug".

I don't know if it's fixable, but Rod and I might need to huddle on this in the context of V3.

-- Scott