> Behalf Of Dave Crocker > Some IETF work items involve narrow topics that are > well-understood and have a core of interested folk who are > clear about both of these facts, and clear about their goals. > > Other work items are not so lucky. So it would appear to be a good idea to hold a BOF to decide which of these categories this particular problem falls into. > As part of our efforts to ensure timely, useful output, we > should work pretty hard to assess likely risks for IETF work, > from before its inception (and on an on-going basis, of course.) That is the type of decision I am happy for the Area Directors to have made. > Some items are clearly *very* important and *very* risky. > They warrant pursuit but they are prime candidates for > wasting the IETF's time. So they require extra efforts to > find ways to make things be both timely and productive. So far it appears that at least 15 of the 50 or so people most closely involved in this space over the years (SAML, Infocard, Identity 2.0) will be attending the IETF for the sole purpose of attending this particular BOF. At least another 10 were planning to attend anyway. That looks like critical mass to me. > At the very least, a wasted BOF consumes an expensive slot > and delays the topic by at least 4 months. 15 * $650 = $9750, should be enough to cover the room hire. > Note, however, that the opportunity costs associated with a > BOF that is a public debacle can, in fact, have strategic > impact, essentially poisoning the IETF waters. Given that this community is already familiar with working in OASIS, Liberty and OATH and a workshop is planned in W3C it would appear to me that far from being premature that the IETF would risk missing an opportunity here if it does not act. In retrospect it seems that the reason we did not address these issues in the SAML group was that we did not feel we had sufficient leverage in that forum to deploy a uniform identifier. _______________________________________________ Ietf@xxxxxxxx https://www1.ietf.org/mailman/listinfo/ietf