I have a lot of sympathy for the points raised by Neil. I do believe, though, we actually have to work on these points, as opposed to just accepting the fact that some previous choices we made weren’t optimal.
I believe we can best do this work when we have a draft that clearly raises the issues we need to work on. E.g., is the distinction between “fully specified” and “polymorphic” indeed as clear-cut as the intro says? Can we maybe develop this definition further in such a way that we can actually derive benefit from this draft? Can we make reasonable transition recommendations with the old registrations still in place (*)? I think that the draft points out the general direction most of us think we want to move toward, but has lots of fine points that need work. We also seem to be able to muster energy to discuss them even during the new year's lull. Looks a lot like a valid WG -00 draft to me. Please adopt. (And keep in mind this really useful thread.) Grüße, Carsten (*) E.g., when offering an alg choice, offer the fully specified values alongside the polymorphic one, enabling an upgraded recipient to work with the fully specified values only, and a legacy recipient to continue functioning. In essence, the presence of at least one fully specified alg value that details a polypmorphic alg value takes the latter out of the game for the recipients that know about this specification. _______________________________________________ jose mailing list [email protected] https://www.ietf.org/mailman/listinfo/jose
