Talk:Technology Square/Wollman

From Computer History Wiki
Jump to: navigation, search

When I started at LCS in 1994, NE43-901 was largely empty; most of the AI Lab computing infrastructure had been pushed into the southwest corner of the old machine room, and there was just raised floor. The two offices in the southeast corner -- which I never saw in person -- were apparently floor-to-ceiling with Carl Hewitt's junk. My office was then on the 5th floor, next to JTW, and I didn't have any reason to go up to the 9th but I did peek through the door once or twice. I recall the big old air conditioners were still there.

In the late 1990s, nearly all of NE43-901 was renovated to create a new headquarters suite for the AI Lab. During the renovation, all the raised floor was torn out, and Hewitt's junk was unceremoniously trucked to his house and left on the driveway, if I recall the story correctly. Ron Wiken's stockroom was in 908, and Bruce Walton was in 906. The southeast corner kept the raised floor and the air conditioner that was there. There was a CM-2 in the machine room, and I don't quite recall when it went out the door -- it might have gone the same time as Tom Greene's CM-5 from the 2nd floor, which went to a museum. I'm pretty sure it was already gone when I moved up to 911 in late 2003.

In 1997, when I left ANA and became the network administrator, I moved down to the 2nd floor. At that time, we had one big machine room in NE43-250, where the CM-5 and Arvind's Monsoon prototype machines were. The Chaosnet connection to campus ran in the riser column from 901 down to 250, where the TI Expolorer CPUs were, and then down into the basement telecom entrance room (which was located behind the cafeteria's grease trap). In 1999 after a burst of faculty hiring, NE43-250 was shrunk, and the CM-5 was winched underneath the freight elevator car to get it out of the building. The Monsoon hardware was dismantled, and offices were built around the south exterior wall of the building. (Mike Patton told me that this was the second time 250 had been shrunk: his office in the early 90s was in space that had been created from a previous shrinking.)

Our main machine room in the late 1990s was NE43-351, which had my network core, the dial-up servers, PSZ's Symbolics machines, Theory, the DNS and mail and everything else servers, W3C, and Zermatt. The Sun that took over for the VAX that took over for XX was located there. There was one air conditioner that barely cooled the room, plus another, altogether ancient one in the old CIA vault that apparently operated on a once-through cycle -- whenever it was turned on, the building management would go around looking for water leaks because the water bill had suddenly skyrocketed. The vault, which was originally used as an extension of 341, was converted into student offices after Zermatt and its disks were retired.

When we left Tech Square, we had five proper machine rooms and three smaller rooms: the rump 250, 351, the fourth-floor one (443?) that Switzerland and Dave Gifford shared, 504b, two air-conditioned closets on 6, and the AI rooms on 7 (where Boris's Lispms were) and 9. Kalman Reti somehow manages to keep Boris's Lispms running to this day. There was an awful lot of stuff that we simply abandoned in place when the day finally came: all the network equipment was obsolete, there was 40 years of antiquated cabling in the chases, old non-EIA racks, terminal servers, thick Ethernet (a DELNI in every office), thin Ethernet, it was a lot.

Near the end, the AI Lab had a tentacle in another Tech Square building, but they didn't have much the in way of hardware over there, just offices and some network equipment.