Alternative IdP Working Group

[Tom Scavo <>]

I finally figured out what was bothering me about the Alternative IdP
Strategies and Assessment Criteria table. [1] There's significant
redundancy in that table, such that investigators will end up
reinventing the wheel multiple times.

Take the "Support for ECP" column, for example. Any row that relies on
Shibboleth will come to the same conclusion regarding ECP. How
Shibboleth is deployed doesn't alter the fact that Shibboleth the
implementation supports ECP. Same is true of any other SAML
implementation---either the implementation supports ECP or not.

Now repeat the above argument for every "implementation detail"
referred to (directly or indirectly) in the column header.

It seems our focus should be on deployment strategies (not
implementation details). InCommon already has a set of Software
Guidelines [2] so I'm not sure we want to go there. In any case, I
think the deployment characteristics of each IdP strategy are most
important.

Tom

[1] https://spaces.internet2.edu/x/FgrkAg
[2] https://spaces.internet2.edu/x/R4HPAg