[ih] TCP/IP routing

[william yeager]

A little bit more about what occurred at Stanford re TCP/IP, and other protocols. Details of my work can be found at: 

https://www.computerhistory.org/collections/catalog/300000119/search/people:yeager-william-bill-person-author/ <https://www.computerhistory.org/collections/catalog/300000119/search/people:yeager-william-bill-person-author/>

The initial implementation of the multiple protocol router I wrote was on a diskless, PDP-11/05. In 1980 it routed PUP, for Xerox PARC D-machines, altos, TENEX on Dec system 10s, printers, etc. The same code also supported an EtherTIP for LAN dial-in connectivity. 

In 1981 I acquired an IP protocol document via my director from SRI/NIC. For the fun of it I added ICMP/echo between my routers and EtherTIPs. They already had PUP echo. I think the Vaxs running BSD Unix and TCP/IP arrived in the 1982-83 time frame. The CSD installed these OSs for us. By that time my routers ran on Andy Bechcolsheim’s Motorola MC68000 mother board and supported both 3mbps and 10mpbs ethernets using 3COM’s 10mbps interfaces and Andy’s 3mbps interfaces. And the IP suite (ARP, RARP were required on 10mbps ether nets), Xerox PARC XNS (10mbps nets), and PUP and were used by the Stanford University LAN.

When the Internet came on line in 1983 Geoff Mogul acquired a ButterFly gatewy from BBN which he named Golden. This gave the routers connectivity across the to the Internet from Stanford’s LAN. When the Texas Instruments explorers arrived circa 1984, I added MIT’s chaosnet routing. Bill is correct that Chaosnet did not require ARP because the host/net fields in the packets were one byte each on 3mbps networks and the encapsuation of the packets was one byte dest host, and one byte source host. The routers boot time configuration had each interfaces source net address for directly connected networks. When the explorers moved to 10mbps networks, they did not support ARP, and the routers kept translation tables for each local net that had these systems. The details are in the paper.  

The routers supported a number of other features and the list is very long. A cute one was to respond to Xerox D-machines on 3 and 10mpbs breath-of-life requests. The response was the code for their boot block that allowed them to boot their OS from servers not on their local net. In any case, all of the details are in the 53 page paper above, and as Bill knows, he is mentioned therein for his contributions to the Stanford LAN.

The above is all from memory of course. The paper was published by the Computer History Museum in July, 2025 and was a two year effort. I still have the sources on my desktop and excerpts from the code are in the paper. To this end I’ll close with an interesting anecdote:

When the CHM was ready to publish when, the senior curator, Dag Spicer wondered about the Stanford copyrights that I included in the paper as well as the code excerpts. To clarify this, I queried the Stanford Office of Technology Licensing. They were not sure and so created a case and a case worker, Mike Spaid. On checking he found that Cisco was given a lifetime copyright by OTL in 2000. We discussed this for a couple of weeks because the software was written under NIH grants which insisted on public access for non-commercial use. Otherwise, OTL must grant a license for commercial use. Thus cisco based their router software on mine when they licensed it in 1987. The upshot was that the director of OTL granted copyright to the CHM for the paper with the requirement that each code excerpt have the following in the references:

<ref> © 2025 The Board of Trustees of the Leland Stanford Junior University 

Bill
————打🎾————
马年快乐🐎
<Happy year of the horse>
保佑众生 🙏🐘🌲🦋🏃‍♀️🐳
<Bless & Protect all living things>