Eliot Lear wrote:
It should also be noted that SIPP is already a union
of SIP and PIP, though Paul Francis, who proposed PIP,
was, in the face of SIPP, saying "PIP is dead".
So, SIPP was developed highly politically, though its
address is still 64bit long.
While *perhaps* SIPP shared some principles with Pip I don't think that
either the addressing or routing architecture was one of them.
https://www.potaroo.net/ietf/idref/draft-ietf-sipp-spec/
o Changed name from "SIP" to "SIPP" as part of merger with Pip.
IIRC, some routing header was added to SIP to politically claim
it is merged with PIP.
My
recollection was that the 128 bit address came to be because of concerns
that we'd end up having to do All Of This *again* if there were a
shortage.
As it is obvious even at that time that 64bit is a lot more than
enough, that was a poor excuse for political merger, I'm afraid.
You can say that IPv6 was the result of committee thinking, and
I would probably agree, but I'm not sure what the alternative would be.
As I wrote:
>> So, SIPP was developed highly politically, though its
>> address is still 64bit long.
SIP or SIPP was an alternative.
As an afterthought, it were even better if very complicated
functionality of SIP that it has optional headers, which
was denied by long and extensive operational experience
of IPv4, which makes SIP and IPv6 not simple at all, were
removed. Fragmentation could and should have been handled
more elegantly than PMTUD.
Masataka Ohta
