I'm not defending the draft per se here. Maybe I should start a new thread.
I think we, as SIDR, need to revisit some basics and some 'axioms' that have been taken for granted, in particular, why are we using RFC 3779 rules here. So far we got: 1- because they were used in s-BGP 2- because they ware already there 3- because we understand them So far, this looks quite 'wobbly' too. This actually doesn't mean that the proposal of the draft is the 'right' one, but I think we need to re-discuss this before the RPKI gets more traction and people get burn scars. Wes's email is pretty clear, risk vs reward. cheers! -Carlos On 8/8/14, 10:55 PM, Randy Bush wrote: >> #1 - Certify the rightful holder of a prefix >> #2 - Help routers validate prefixes when deciding how to build their >> routing tables >> >> RFC 3779 validation rules are probably the best suited rules for #1. It >> seems clear now that whether RFC 3779 validation rules are adequate or >> well suited to #2 has not been discussed here in depth. > > i think you are a bit confused here. > > what geoff proposed changes nothing about 1 and 2. it's about how one > validates the correctness of the external data, irrespective of how it > is used. > > randy > _______________________________________________ sidr mailing list [email protected] https://www.ietf.org/mailman/listinfo/sidr
