Shibboleth Developers

["Scott Cantor" <>]

> I believe we showed our implementation to Chad.  It works for us.

As I understood it, it was somewhat postgres-specific, but I'm not
precluding anything out of hand.

But to the extent that people want a toy implementation that has the
illusion of reality to it, they probably want a self-contained Java
approach, and making it real, to say nothing of clusterable, becomes one of
those things people will all need to do, but few will.

Much like your comment about ARP management...I have a million things to get
to before automating that is even on my radar screen here. It's far from a
requirement for production. So adding features to make manual support
simpler is a big win, even if that sounds stupid.

-- Scott