Evaluable criteria

[Cantor, Scott]

On 6/1/15, 10:45 PM, "Cantor, Scott" <cantor.2 at osu.edu> wrote:


>
>It wouldn't work, but subclassing the existing plugins to add support for 
>an additional Criterion subtype should work, as does adding functionality 
>to the existing base classes.

Related point: by the strict reading of the Java policy I posted last 
week, the metadata resolver abstract base classes and implementations are 
both impl classes we can change freely.

Now, you can take the view that we need to change the policy in light of 
people using OpenSAML in their own projects because a good chunk of the 
real useful code in it is implementation by that definition, as we have 
discussed. Myself, I advocate we do what's best for Shibboleth because 
that's where our funding is.

-- Scott