WORTH READING the undead urban myth of the LOC/EID split

[David Farber <dave () farber net>]

Begin forwarded message:

From: "John Shoch" <shoch () alloyventures com>
Date: November 5, 2008 6:43:55 PM EST
To: <dave () farber net>, "ip" <ip () v2 listbox com>
Cc: "John Shoch" <shoch () alloyventures com>
Subject: RE: [IP] Re:   the undead urban myth of the LOC/EID split

OK, I've been biting my tongue during this discussion -- but if we are  
delving into "what might have been" done with addressing, let's give  
full credit to Yogen Dalal and Bob Printis for their paper, "48-bit  
Absolute Internet and Ethernet Host Numbers."  It has recently been on  
a number of lists for "my favorite papers."
People thought we were crazy when we argued for 48-bits in Ethernet  
addresses, and then 48-bits in Internet addresses.

The work emerged as a pre-print from Xerox PARC published around July  
1981:
http://ethernethistory.typepad.com/papers/HostNumbers.pdf
The published version came out later in 1981, in ACM SigComm:
http://portal.acm.org/citation.cfm?doid=1013879.802680  [ACM login  
required]

They presented the full, complete case for a very large flat address  
space for Internet addresses.  [Although not specifically identified  
in the paper, this was the basis for the Xerox Network System (XNS)  
protocols, which were implemented and sold as part of the Xerox  
product line.]

The paper is an easy read.  My favorite part:

"48 bits allow for 140,737,488-million physical hosts and multicast  
IDs each.  We chose this size based on marketing projections for  
computers and computer-based products, and to permit easy management  
of the host number space."
"An estimate of the number of computer systems that will be built in  
the 1980s varies....  The US Dept. of Commerce, Bureau of Census  
estimates that in 1979 there were 165 manufacturers of general-purpose  
computers, producing about 635,000 units...  There were also about  
992,000 terminals and about 1,925,000 standard typewriters built!"
"We expect that the production of microcomputer chips will increase in  
the decades that follow, and there will be microprocessors in  
typewriters, cars, telephones, kitchen appliance, games, etc. ... For  
example, when a car containing a microprocessor chip needs repairs, it  
might be plugged into a diagnostics system thereby putting the car on  
a communications system."

I sure wish I had written that, 28 years ago.  They had a lot of  
foresight, and had worked out the details:  unstructured addresses,  
network hints, clearinghouse lookup, multi-cast, etc.
How much would we pay to have a 280-trillion element flat name space  
today?
If only I could figure out how to rewind the video-tape and start  
over......

John Shoch
Alloy Ventures

-----Original Message-----
From: David Farber [mailto:dave () farber net]
Sent: Wednesday, November 05, 2008 2:01 PM
To: ip
Subject: [IP] Re: the undead urban myth of the LOC/EID split

> From: Tony Lauck <tlauck () madriver com>
> Date: November 5, 2008 3:06:50 PM EST
> To: Dave CROCKER <dcrocker () bbiw net>
> Cc: "'David Farber'" <dave () farber net>
> Subject: Re: [IP] the undead urban myth of the LOC/EID split
>
> If it had a few more bits the one serious problem with IP v4 would  
> have been pushed out past our lifetimes. Unfortunately, Vint Cerf  
> did not take me seriously back in 1978 when I suggested that the 32  
> bit address was going to be too small if he were successful. :-)
>
> As to the ID function, I think you are on the right track in  
> starting with existing end to end security mechanisms such as TLS.   
> I haven't thought through any details, beyond the possibility that a  
> private key / public key pair can serve as an ID without requiring  
> new global infrastructure.
>
> Tony Lauck
> https://www.aglauck.com
>
> > That is what motivated my selections, above, starting with Tony  
> > Lauck's  observation about complexity:  The network layer really  
> > does work well.  I mean *really* well. As "experiments" or  
> > "prototypes" go, IPv4 has  not done all that badly.  Not too many  
> > technologies scale over that many years and that many orders of  
> > magnitude of population size and  performance speeds.




-------------------------------------------
Archives: https://www.listbox.com/member/archive/247/=now
RSS Feed: https://www.listbox.com/member/archive/rss/247/
Powered by Listbox: http://www.listbox.com