On Wednesday, April 25, 2012 01:53:35 PM Murray S. Kucherawy wrote: > > - The addition to section 4.5.1 isn't quite correct. Elsewhere we > > > > tell report senders not to assume different types of reports will be > > treated differently, so I don't think there's any need for receivers to > > update to do so. I think the most that can be said is that receivers > > ought to arrange for a reasonable default result if an unknown type is > > encountered. > > There issue is that we make it a MUST to accept all types listed in a > registry. How would an implementation do that? There's no protocol to > query the registry for new types, so it can't really be done live. The > IESG member is saying we need to explain to people what's involved in > satisfying that MUST.
We don't really define accept. If I "250 OK" to accept the message and then discard unknown types, that's a win from that perspective. I probably wouldn't do it that way, but that's not inconsistent with the requirement. I don't think it's necessary to have a list of all currently possible values to satisfy the requirement to accept all types. Scott K _______________________________________________ marf mailing list [email protected] https://www.ietf.org/mailman/listinfo/marf
