As part of the next OSIS Interop, participants are contributing pain points that stand in the way of easy interoperation of user-centric components.
My personal list involves the pieces that I have found to stand in the way, when writing code for the Pamela Project. For those of you who are interested, the discussion is still going on, you could probably sneak in a few extra suggestions if you were do it soon. I would suggest subscribing to both the OSIS general mailing list and the new Google Group dedicated to this particular Interop (the new list is currently empty, but I don’t expect it to remain that way for long).
List items have so far ranged from usability issues to protocol underspecification, to debates over whether certain aspects of a component can be implemented securely. There are OpenID suggestions from at least two different participants, which is great! I think the final list is going to be an important body of work, and I personally would like to see that body of work turned into the beginnings of a set of recommended practices that new component writers can refer to as they work to become interoperable.
Stay tuned for more…