A couple of years after the incident at Bell I left the telco to try my hand at system administration. This was around 1986 and things were different than they are today. Computers were pretty scarce compared to now. It wasn't economically feasible to have one on every desk. Servers used a lot more power than they do now, they cost a lot more, and they were a bigger pain to manage.
I got my first admin job through a friend. It was at a hospital outside of Chicago. I didn't really have a lot of applicable skills, to be honest, but I knew my way around a mini and I thought I could figure out the Unix server, since the ESS systems were based on Unix and I'd had a class on those. I was supposed to manage the network too, which I didn't mind at all since I had already spent years doing similar things with Bell. The hospital used 10Base2 for some of its stuff, but most of the computers were still not networked at that time.
I got the hang of the job pretty fast. It was busier than Bell. Telephony equipment used by Bell was designed for reliability, our stuff was amateur in comparison and would fail often. I liked a challenge and didn't really mind the workload. Plus I got to spend a lot of time at the nurses' stations where the networked minis were. I enjoyed that part of the job, if you catch my drift.
I was under the desk at a nurse's station replacing a coaxial cable that had been chewed through by a high-heel when I noticed something strange. The desk was right up against a wall, and the wall had a plate on it which held the network port for the station. That part was normal. What was strange was that the middle of one side of the plate was pried up, and a yellowish cable poked through it. It was just a couple of inches long, and at the end of it was a serial connector.
The process of rewiring a building's network was the same back then as it is today. You tie one end of the new network cable to the old network cable after you've detached it from its plate or port or whatever, and you fish the new through the conduit by pulling out the old. If there's no old cable in a conduit you use a fishing line. The serial cable I was staring at was probably ten years old at least. It was yellowed, dirty, and ignored. For whatever reason it had been left there when the new network was installed. None of the other plates I had seen had a serial cable sticking out from behind them, just that one.
Being the old Bell head that I was, I had to investigate. The hospital was open 24/7 so I couldn't simply stay late to have some alone time with the serial cable. I had to devise a ruse. The computer on the desk above the serial line was used for station logs. I didn't want to interrupt the flow of data at a hospital so I waited until a slow day, and unplugged the station's coax down in the basement. They probably wouldn't notice for awhile so I hurried things along by moseying up to the station and asking how their computer was working. The nurse said it was fine. She was doing her nails and didn't even look up. I had to coax her into checking the connection by telling her that some of the stations were having problems. Finally she let me behind the desk.
The serial cable in question had an DB-25 connector, and was probably RS-232. I'd brought along some tools and adapters. One of the adapters I had with me could connect the mini at the nurses station to the serial port. I connected it up and loaded a terminal emulator from a floppy I also had with me. I was right on with my guess about the interface.
I tried a couple of defaults for Unix systems and they didn't work. I asked the nurse to come over and try her login and password, and she did it without batting an eye. Hers didn't work either. Her and I messed around with it for awhile before she mentioned that there was a station login that they used before the nurses had individual accounts. She tried that one, and it worked.
1) Station log
2) Room Status
3) Log out
I was pretty darn curious at that point. Not only was the serial port still hooked up to some system, it was hooked up to a functional system. My first thought was that our mainframe downstairs had some backward compatibility built in, for the old network, that had since been forgotten. But I'd been all over that machine and I knew that it didn't. I asked the nurse to 'test' the system. She typed a '2' and then a room number.
*** Rm. 306 ***
PT: MARGRET CHUANG
ADMITTED: 0300 06MAY1975
REASON: CHEST PAIN
I ignored the rest of the screen. My eyes were fixed on the date. That was a decade prior. The nurse was confused too. She said:, "That's not the room I entered. And that's not who's in room 306." I told her that the system must still be malfunctioning, then disconnected the mini from the serial port. I went back downstairs and hooked it back up to the thinnet.
There was only one other guy who worked on the network in the hospital. His name was Rob and he'd been there for a long time. I asked him about the serial connector up at the nurse's station and he just said, "We used to have dumb terminals at the stations, connected down here to the old mainframe. I guess they forgot to pull that cable out." He assured me that there was no way the serial cable was hooked up to anything. I asked him where the other end of the serial cables were, and he said they'd cut them off and pulled out the cables when they did the new network install. I asked him if there were any other mainframes or servers anywhere in the building. I already knew the answer: there weren't.
Later that day I walked back up to the nurse's station and asked the nurse how the computer was working. She typed on it a little bit and said that it was fine. I asked her if the information was accurate now and she said yes. I asked her if she could look up Margret Chuang, "Just to verify our old data". There were two of them in the system. Narrowing down by date we found the one we had seen earlier.
According to the system, that Margret Chuang had indeed been admitted early morning on May 6, 1975. She had died that same day in surgery.