> Third, I think the implementers guide is absolutely essential. > (Interestingly enough, the discussion of white spaces vs. commas in > yesterday's thread has effectively started this work.) In my opinion, > this item must be carried in parallel with others. I wonder if this > should be tied up with use cases. The use cases drive the protocol > definition and then effectively become the testing tool for the > protocol. Thus, for every use case, the guide could show an example > implementation.
This is what I'd had in mind when I brought up the implementers guide. I think we're seeing some common collections of stuff, profiles based on real use cases. So a person saying "I want to build a PR server that supports the webserver grant profile and the display extension", or someone saying "I want to build a client for the Facebook collection of stuff" will have a document that tells them exactly what to do at each step. The spec itself splits things out into their normative places, but someone building a component is going to want to just know what arguments to put on each request. -- Justin _______________________________________________ OAuth mailing list OAuth@ietf.org https://www.ietf.org/mailman/listinfo/oauth
