A Short History of Chaosnet (2018)

#1

Post by whocansay » Mon, Mar 25, 2019, 5:57 AM UTC

A Short History of Chaosnet (2018)
twobithistory.org

Re: A Short History of Chaosnet (2018)

#2

Post by teddyh » Mon, Mar 25, 2019, 9:21 AM UTC

Oddly, absolutely no mention of the current use of the CH class in DNS: The “version.bind.” CH TXT query, which often gives the server version of the Bind DNS server.

Re: A Short History of Chaosnet (2018)

#3

Post by segfaultbuserr » Mon, Mar 25, 2019, 9:32 AM UTC

Its quite interesting to observe that the trace of Chaosnet can be found in Richard Stallman's original GNU Manifesto, written in 1985. Though it's not a surprise since he worked extensively at the MIT AI Lab at that time.

> GNU will be able to run Unix programs, but will not be identical to Unix. [...] We will try to support UUCP, MIT Chaosnet, and Internet protocols for communication.

https://www.gnu.org/gnu/manifesto.html

The mention of UUCP at a separate protocol in parallel to IP was also interesting. Since the 90s, UUCP is running on top of TCP/IP which runs the Usenet, but in the 80s, UUCP was used on point-to-point dialup links.

Re: A Short History of Chaosnet (2018)

#4

Post by lukeh » Mon, Mar 25, 2019, 12:14 PM UTC

Did anyone actually deploy Hesiod? I haven't heard about it for a long time. I remember implementing experimental support for it in lookupd when I was an intern at Apple in 1998 ;)

Re: A Short History of Chaosnet (2018)

#5

Post by cpr » Mon, Mar 25, 2019, 2:30 PM UTC

While the LispMs designers might have originally intended a network-based file system over Chaosnet and a network-based shared terminal configuration, in the end the LispMs each had sufficient local storage for standalone operation, and dedicated displays. (Though using long cables so that the machines could be in a machine room.)

Moon wrote that the Chaosnet routing scheme “is predicated on the assumption that the network geometry is simple, there are few multiple paths, and the length of any path is quite short. This makes more sophisticated schemes unnecessary.”

In the end, the whole network at MIT was effectively a single physical run (I managed the EECS portion) of the old-style thick coax. What a nightmare to manage and debug across a whole campus! (Memories of crawling in ceiling cavities with a TDR...)

Re: A Short History of Chaosnet (2018)

#6

Post by a893754 » Mon, Mar 25, 2019, 3:04 PM UTC

This was previously discussed at: https://news.ycombinator.com/item?id=18107136

Re: A Short History of Chaosnet (2018)

#7

Post by icedchai » Mon, Mar 25, 2019, 7:37 PM UTC

segfaultbuserr wrote: post #3

Its quite interesting to observe that the trace of Chaosnet can be found in Richard Stallman's original GNU Manifesto, written in 1985. Though it's not a surprise since he worked extensively at the MIT AI Lab at that time. > GNU will be able to run Unix programs, but will not be identical to Unix. [...] We will try to support UUCP, MIT Chaosnet, and Internet protocols for communication. https://www.gnu.org/gnu/manife…

I was around during that time period. UUCP was used over dialups well into the 90's. Many early ISPs offered UUCP email and news feeds.