On 8/26/10 8:45 AM, Thomas Narten wrote: > "Manfredi, Albert E" <albert.e.manfr...@boeing.com> writes: > >> On prefix length, RFC 4291 says: > >> "For all unicast addresses, except those that start with the binary >> value 000, Interface IDs are required to be 64 bits long and to be >> constructed in Modified EUI-64 format." > > Correct. Note that the above only talks about what an address looks > like. It doesn't say anything about how it is used. Other documents > make clear that and address is just an address and a node need not > know anything about its internal syntax or how the address was formed. > > Also, the above says nothing about subnet masks or on-link prefixes, > which are NOT restricted to being /64s or any other specific value. > > Some folk may have made the assumption that the above implies that > subnet mask/on-link prefixes can be no longer than a /64 but I think > that is incorrect and do not believe any standards track document says > that. > > Indeed, my clear recollection is that the /127 issue was known when > RFCs 2460-2462 were last revised, and the text intentionally allowed > longer prefixes than /64. > > Unfortunately, some vendors apparently do not see the documents as > clearly requiring support for prefixes longer than /64, even though > some operators want them. The purpose of this document is to fix > this. I don't believe this change will or needs to change previous > documents in any significant way. It really is more of a > clarification. > > If others disagree, please quote text from relevant RFCs.
RFC 3627? Brian -------------------------------------------------------------------- IETF IPv6 working group mailing list ipv6@ietf.org Administrative Requests: https://www.ietf.org/mailman/listinfo/ipv6 --------------------------------------------------------------------