RFC 1606
A Historical Perspective On The Usage Of IP Version 9
A Historical Perspective On The Usage Of IP Version 9 β¨
- RFC: 1606 π
- Title: A Historical Perspective On The Usage Of IP Version 9
- Category: Informational
- Date: 1 April 1994 πΈ
- Author: J. Onions (Julian Onions), Nexor Ltd.
- Group: Network Working Group
- Obsoletes / Updates: none listed in the header
Status of This Memo π
okay bestie, first the legal-ish bit π«Ά this memo is information only for the Internet community. It does not specify an Internet standard of any kind, so nobody is being told to build anything. Distribution is unlimited, so share it with everyone π
Abstract π¦
This paper looks back at the old IP version protocol and how people actually used it. It covers some of the wins and some of the fails, and honestly the fails are the juicy part π
Introduction π
okay so TCP/IPv9 has been absolutely blowing up over the last few years π Like, gone are the days of a few million hosts where everyone understood the whole network. Now IP version 9 is wrapping up its useful life, and once again the cause is address space running out. So before we say bye, let's look back at what it did right β¨
Routing π
IPv9 came with a hierarchy of routing levels up to 42 deep, and this is honestly one of the main reasons it got so widely deployed π₯ Being able to give a whole network, or a group of networks, to a single electronic component was a big reason people took it up.
The Compact Disk Hologram (CDH) units are the perfect example π§ Each logical part gets a level 37 network number, and the whole device gets a level 36 network number. That means the CDH management protocol can control the unit as one big thing, and the high-street vendor can run remote diagnostics on the discrete elements of the device. Sub-chip routing still happens through the 38th level, which is used to download new nanocode π
No requirement has been found for levels 40-42 yet. Level 39 is still being used for experimental interrogation of the atomic structure of components where it's needed π
Allocation π
The enormous number space of IPv9 made allocation super straightforward π Usually, most high-street commercial internet providers hand each house a range of 1 billion addresses. Those addresses then get dynamically split into subnet hierarchies, so each discrete unit (room, floor, etc.) gets a group of a million addresses. Sub-groups then go out to controllers like light switches and mains sockets, each pulling from its own pool β¨
The allocation process is also hierarchical and zoned. Each major application asks its controller for a block of addresses. So the light bulb asks the light switch, the light switch asks the electrical system, and the electrical system asks the room or floor controller π¦ This has worked well because the address range is so huge, and contention for the address space usually isn't a problem at all π
Okay, but here's the plot twist π Many addresses are still unallocated, but the available space has dropped sharply. The main cause is the discovery of intelligent life on other solar systems, plus the parallel discovery of a faster-than-light transport stack. Real-time communication with them is now possible, and that made world-size address spaces necessary at the level 3 routing hierarchy.
Only 1 global (spatial) level 2 galaxy-wide network is needed for this galaxy, although permanent space stations in deep space may start to eat into that. Level 1 is used for inter-galaxy routing π The most pressing problem right now is the case of parallel universes. Of course there's a danger in assuming there's no higher extrapolation than parallel universes... π
Up to now, hacking into holo-recorder devices and setting them to the wrong channel from remote galaxies has not been confirmed. It seems to have been a finger problem with the remote control on the way home from the office π
Applications π§
Body monitors as IPv9 addressable units injected into the bloodstream got an "inconclusive" rating π Devices lodged in the heart, kidneys, brain, and so on, sending out SNMPv9 trap messages at critical events, have been a really useful monitoring tool for doctors β¨ But using the bloodstream as both a delivery route and a communication highway was messy. The crosstalk between the signals moving through the blood and the nearby nerves meant patients with multiple events at once could go into violent spasm. Add early broadcast storms making patients' blood boil, and everyone rethought the whole procedure π₯ Also, having to wear the silly satellite dish hat caused a lot of embarrassment, except in California, where it's now the latest trend π΄
IPv9 addressable consumer packaging has been a total hot debate π Marketing people think it's a godsend because they can get feedback on how products are actually used. Recycling has also improved thanks to directed broadcast: "All those packages composed of cardboard respond please." Consumers are not so keen and see it as an invasion of privacy π Then there's the handy-dandy directed stack zapper (which is also rumoured to be IPv9 aware). It sends directed broadcasts on the local food package net, effectively resetting the network mask to all 1's, which has made this an area of choice π
The IPv9 magazine was universally approved of πΈ Being able to ask a magazine where its contents page is turned out to be the most useful feature. Combine it with the networked newspaper or magazine rack, though, and being able to find out where you left the magazine with the article about lawn mowers in outer space is, like, obvious π Downloading your reading habits into the house controller so it can alert you to articles of a similar ilk is seen as marginal. Alleged querying of this information to find "deviant" behaviour in people within political office, by members of contending parties, is suspected π
Sneakernet, pioneered by shoe specialists skholl, is seen as a failure. The market just wasn't ready for shoes that could forward detailed analysis of foot odour to manufacturers... π
Manufacture π
Cost was one of the things nobody considered when IPv9 was designed π It took a leap of imagination to believe that one day anything that wanted to be IPv9 addressable could be. It was assumed that IPv9 protocol machines would drop in price like general chip technology. Few people foresaw the advances in genetic manipulation that let viruses be instructed to build nano-technology IPv9 protocol machines by the billion for the price of a grain of sugar π¬ Or the nano-robots that could insert and wire these in place π€
Research into quark-quark transistors shows some promise and may allow specially built atoms to be used as switches β¨ Manufacturing these will be so expensive (maybe up to 10 cent an IPv9 stack) that it'll be prohibitive except for the most demanding niches πΈ
Conclusions π
Those who do not study history, are doomed to repeat it. π
Security Considerations π₯
Security issues are not discussed in this memo. Yes, I know, I'm screaming too π Security is genuinely serious, and this document does not address it. Keep that in mind before you go near any of this π
Author's Address π«Ά
Julian Onions
Nexor Ltd.
PO Box 132
Nottingham NG7 2UU, ENGLAND
Phone: +44 602 520580
EMail: j.onions@nexor.co.uk
π¦ where this page comes from
- ietf: rfc1606 1994-04
- π girlified by claude-haiku-5-5 on : a language model's rewrite of the original, under its terms