[ih] Internet without entrenched factions?

[John Gilmore]

John Day via Internet-history <internet-history at elists.isoc.org> wrote:
> One of the issues that OSI had that the Internet didn't were very
> entrenched factions that were going to have their way no matter
> what.

I'm shocked that you would say that.

But I'd guess you've never tried to get a simple historical correction
through the modern IETF.

> The PTTs absolutely didn't want a Transport Layer or datagrams,
> when they couldn't stop it (the US wanted nothing else) they did all
> they could to restrict it and make it something to avoid.

This exactly describes what IETF is doing to IPv4.

They did actively try to kill it -- while it was carrying more than 80%
of the world's Internet traffic!  There was a whole "sunset4" WG!

I tried to get the IETF intarea wg to simply work on a draft that said
that IPv4 is a valid protocol suite that the IETF will continue
maintaining.  Half of the WG said "no way", the other half said "we're
already doing it so there's no need to say so".  The result was that no
draft was accepted, so no statement that IPv4 remains valid was ever
adopted by the wg or by the IETF.  The factional campaign to kill IPv4
continues.  The whole history is documented in these expired drafts:

  https://datatracker.ietf.org/doc/html/draft-ietf-sunset4-ipv6-ietf-01
  https://datatracker.ietf.org/doc/html/draft-schoen-intarea-ietf-maintaining-ipv4-01

Meanwhile this IPv6 faction shoots down every other simple proposal that
straightens out historical kinks in IPv4 that have already been fixed by
most implementers -- but never written down in a spec.  This has been
going on for years and years.  And the rest of IETF lets them get away
with it.

For example, the lowest address in each subnet was reserved back in 1983
because the 4.2 BSD UNIX release used it as a broadcast address, before
IP broadcast was standardized.  This was fixed in the 4.3 BSD release in
1986 to use the last address in the subnet (like everyone else).  But
despite the issue and its resolution being documented in RFCs 1122,
1812, and 3021, the lowest-address restriction was never removed in a
spec.  The lowest address has never since 1986 been used for broadcast,
nor for anything else.  A mythology grew up about it that the .0 address
somehow referred to "the whole subnet" and couldn't be used for a
unicast host.  This wastes many millions of perfectly good addresses
throughout the Internet, particularly in small subnets like a /29 (with
only 6 usable addresses rather than 7).

I and implementers wrote a draft that explained the history, proposed
standardizing the .0 address as unicast, and demonstrated that it's
upward compatible because distant hosts can't even tell the difference
(by definition, they don't know where your subnet starts or ends).  The
implementers have already straightened it all out and it's running in
production -- OpenBSD since 2010, NetBSD since 2016, Windows since 2021,
Linux since 2021, Fedora and Ubuntu since 2022, Android since 2022.
Amazon has been using these addresses in production since 2021; ping
44.192.0.0 or access http://ec2-reachability.amazonaws.com/ to check it
live.  But IETF still refuses to merely document that it works --
because "it would reduce the pressure for Internet sites to move to
IPv6"!!!  See the expired Internet-Draft at:

  https://datatracker.ietf.org/doc/html/draft-schoen-intarea-unicast-lowest-address-08

It's possible that IETF had no entrenched factions "back in the day".
But the very slow uptake of IPv6 has caused the people who designed IPv6
to become such a faction.  And the IETF seems to have no working
defenses against such factions.

	John
	
PS: In a completely separate incident, there was the 5-year delay in
standardizing DNS Security because one or more high-up people in the
IETF decided that "no standard that requires the use of Jim Bidzos's
monopoly crypto algorithm (RSA) is ever going to be approved on my
watch".  Despite having a free DNSSEC license for the patent and for the
RSA software, and the patent only having three years left to run at that
time.  Sure, there are no "get our way no matter what" factions in IETF,
and never have been!