Forgot your password?
Cyhist Mar 04 1997 A
Document Actions
Date: Tue, 4 Mar 1997 08:11:25 EST
Reply-To: "CYHIST Community Memory: Discussion list on the History of
Cyberspace" <CYHIST@MAELSTROM.STJOHNS.EDU> Sender: "CYHIST Community Memory: Discussion list on the History of
Cyberspace" <CYHIST@MAELSTROM.STJOHNS.EDU> From: keith reid-green <kreid-green@ETS.ORG>
Subject: CM> First e-mail message
______________________________________________________________________
Community Memory: Discussion List on the History of Cyberspace ______________________________________________________________________
When I worked at Hercules in Salt Lake City I was part of a small team that put together a system that tracked the production of Minuteman missiles. The missiles were basically built outdoors in bunkers made of mounds of earth, and they were moved from one place to another during assembly. We put a terminal at each assembly point so the operator could report outgoing and incoming missiles (-: not in the military sense :-) and so management could get the status of the whole assembly process at any time. This was completed in 1963, and the computer that kept track of this stuff was an IBM 7740, of which almost nobody has ever heard. A side benefit of this system was the ability to send text messages from terminal to terminal or to broadcast a message from the host to everybody on the network. This was in 1963.
The IBM 7740 was a bitch to program. The only arithmetic instructions were Add and Complement, and the four working registers were one byte each, called, W, X, Y and Z. If an add to Z resulted in an overflow, the ovfl bit was set and the next Add also added the ovfl bit. This was nice if you wanted it to happen and not nice if you didn't.
The assembler for the 7740 ran on a 1401. The 7740 was the first 8-bit byte machine I ever ran into. I remember having to write a subroutine to find the address of a disk track, given the sector address. There were 200 sectors on a track and the sector address was six BCD digits. Sounds easy enough, you say--Just divide the sector address by 200. Right, but there was no Divide instruction. I had to convert the BCD number to binary,
shift it right one place (thereby dividing it by 2) and convert the result back to BCD, dropping the least significant two digits. Voila! Division by 200. Took 61 instructions...
I wonder if I had a reunion of 7740 programmers if anybody would come.
Keith Reid-Green
KReid-Green@ets.org
Princeton, NJ
______________________________________________________________________
Reply-To: "CYHIST Community Memory: Discussion list on the History of
Cyberspace" <CYHIST@MAELSTROM.STJOHNS.EDU> Sender: "CYHIST Community Memory: Discussion list on the History of
Cyberspace" <CYHIST@MAELSTROM.STJOHNS.EDU> From: keith reid-green <kreid-green@ETS.ORG>
Subject: CM> First e-mail message
______________________________________________________________________
Community Memory: Discussion List on the History of Cyberspace ______________________________________________________________________
When I worked at Hercules in Salt Lake City I was part of a small team that put together a system that tracked the production of Minuteman missiles. The missiles were basically built outdoors in bunkers made of mounds of earth, and they were moved from one place to another during assembly. We put a terminal at each assembly point so the operator could report outgoing and incoming missiles (-: not in the military sense :-) and so management could get the status of the whole assembly process at any time. This was completed in 1963, and the computer that kept track of this stuff was an IBM 7740, of which almost nobody has ever heard. A side benefit of this system was the ability to send text messages from terminal to terminal or to broadcast a message from the host to everybody on the network. This was in 1963.
The IBM 7740 was a bitch to program. The only arithmetic instructions were Add and Complement, and the four working registers were one byte each, called, W, X, Y and Z. If an add to Z resulted in an overflow, the ovfl bit was set and the next Add also added the ovfl bit. This was nice if you wanted it to happen and not nice if you didn't.
The assembler for the 7740 ran on a 1401. The 7740 was the first 8-bit byte machine I ever ran into. I remember having to write a subroutine to find the address of a disk track, given the sector address. There were 200 sectors on a track and the sector address was six BCD digits. Sounds easy enough, you say--Just divide the sector address by 200. Right, but there was no Divide instruction. I had to convert the BCD number to binary,
shift it right one place (thereby dividing it by 2) and convert the result back to BCD, dropping the least significant two digits. Voila! Division by 200. Took 61 instructions...
I wonder if I had a reunion of 7740 programmers if anybody would come.
Keith Reid-Green
KReid-Green@ets.org
Princeton, NJ
______________________________________________________________________
