On 26/02/2020 09:35, Stewart Bryant wrote:
On 20 Feb 2020, at 02:18, John C Klensin <john-ietf@xxxxxxx> wrote:
We do need Governments to ban the selling of new IPv4-only
domestic devices (CPE routers, TV's, game boxes, etc.).
We have been there and seen that or variations on it tried. In
some alternate reality, the last attempts were so successful
that TCP/IP and all other competing solutions died out, leaving
the world running entirely on an updated version of the OSI
stack (either connectionless or connection mode) today. In the
reality in which we (or most of us) actually live, government
attempts to advance networking by requiring some technologies
and banning others mostly lead to technological paralysis and
increased costs for all concerned.
Although ISO 8473 would have been a better platform than the IPv6 as we eventually settled on.
- It could support multiple address types, of variable length, which SR is teaching us that we need.
- It had a better checksum than that available in the transport layer, and would have meant that we could have avoided compulsory UDP c/s
- The protocol suite supported connection oriented which MPLS taught us that we needed.
Before my time, but was IPv4 designed before or after the Internet was released from the government to the public?
Way before, if I understand your question aright.
I see the start of the public internet as April, 1995, when commercial
activity, over and above applying for NSF grants, was permitted. This
enabled ISPs as we now know them.
IPv4? I date to RFC791, September 1981 although much of the technology
was fixed before then.
Tom Petch
Much of the success of the mobile phone industry arrises through standardisation, and if you read Stephen Temple’s book you understand the influence that the regulator (using spectrum availability as a leaver) had on forcing harmonisation.
So maybe governments do not always get it wrong.
- Stewart
