[gumby]

Nothing really these days. It was a thicknet (10BASE5) LAN similar to PUP with service names (strings) rather than port numbers. We used it at MIT pretty much only inbetween the AI and LCS machines (which were all in the same machine room and building) but was never really commercialized except that early machines from Symbolics and LMI were just CADRs which had chaosnet interfaces in them so if you had any of those early lispms you needed chaosnet to talk to them.

There were other experimental LAN systems around as well, some commercial some not.

It was all superseded in the mid-late 80s by IEEE ethernet.

It's a little bit of a nostalgia trip for me but if you never used it there's not much to learn in retrospect.

[dfrage]

It was used way beyond the 9th floor of 545 Tech Square. To the west, the longest cable run was to the NW16 Plasma Science and Fusion Center, I think there were only 2 taps on that cable, it was marginal. To the far south the Mathematics Department in Building 2, and it went east to Building 20 at minimum.

[gumby]

True, I'd forgotten that it went to Building 20, but Building 2? Someone wanted to use MACSYMA?

In any case by 83 or so they were building out Athena which was ethernet based.

[dfrage]

> but Building 2? Someone wanted to use MACSYMA?

As I understand it, they just wanted to be connected to the rest of the world, which their RJE connection to NCARs' supercomputers didn't accomplish.