Subject: Re: v6 (was Re: -current sendmail cancer in IPv4-only kernel)
To: Robert Elz <kre@munnari.OZ.AU>
From: Sean Doran <email@example.com>
Date: 05/09/2000 00:18:16
Robert Elz <kre@munnari.OZ.AU> writes:
> CLNP as a protocol is in many ways quite like IP, so
> obviously isn't awful .. what is awful is its addressing
> plan, which is truly atrocious.
Oh, no, actually I thought that decoupling the transport
syntax (bits on wire) from the abstract syntax used by an
addressing authority to define the structure of their
respective addressing domain was a great feature of
IS 8348/AS 2 and I was sad to see it die. (Although the
argument over 6 bytes vs 8 bytes or more vs 2 bytes or
less was very silly and typical of design-by-committee.)
The variable length addresses were also a great feature,
although the maximum length was too short.
There were some really truly awful things in OSI --
the preferred decimal encoding of binary syntax NSAp
addresses, the IONL, the BRP, ISO 8881, the acronyms --
but CLNP itself was OK, once you wrap your way around how
address binding works to actually route packets (tricky!).
RFC 1629 is actually kinda neat, if you ask me, and is the
best overview of CLNP addressing and how to do CLNP
routing I know about.
> That's what killed TUBA - its one big drawcard was that it could leverage
> on what CLNP was actually deployed (in existing code bases, if not
> actually very much in use). But to gain that advantage the addressing
> plan, and precise (ie: unaligned) packet formats, and its checksum algorithm,
> had to go along with it, and all that baggage was simply impractical.
All that baggage... whee!
I know you were there, but my memory is a little
different. What killed TUBA was the combination of
anti-ISO folks and anti-IETF folks strangely acting in
concert. One side was so repulsed by the OSI brigade that
they did EVERYTHING differently as a matter of principle
(witness OSPF) and rejected TUBA because it did not. The
other side was so exercised by the idea of upstart IETF
*CHANGING* an International Standard on its own, that it
became nearly violent about any suggestion that IETF could
improve upon CLNP, and maybe feed back changes through
people who played in both ISO and IETF.
The alignment canard was absolutely destroyed by Tony Li's
FPGA demo. That forwarding speed is quoted as a reason
for CLNP's and variable-length addressing's failure is a
tragedy. I think people lost track of things during the
heat about the compromise of "profiling" CLNP.
Personally, I'm just upset that every attempt to carry
forward with CATNIP was sniped at by some SIP/SIPP
partisans, and so seems to have been abandoned. :(
I think IPAE was very nearly the most disgusting thing
that was wrung through the IPng process.
> TUBA came much later, and was one of the candidates for
> the IPv4 replacement (along with PIP, and SIP, which
> turned into SIPP, and then IPng).
RFC 1752 is a really interesting document, in retrospect.