Forgot your password?
Cyhist Apr 09 1997 B
Document Actions
Date: Wed, 9 Apr 1997 15:22:47 EDT
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> 1401 architecture
______________________________________________________________________
Community Memory: Discussion List on the History of Cyberspace ______________________________________________________________________
John Ahlstrom wrote "I believe there was an 8th bit (in the 1401) the record mark..."
I gave my 1401 manual to the Archive for the History of Computing in Manchester, but I kept a bunch of handy-dandy pocket references, including 1401 and 1410. The 1401 ref that I have is not specific enough, but the 1410 has a table of character codes. This is how it appears to have been--in core storage, the 1401 was a seven-bit character and the 8th bit was used for a word mark. The record mark was a character itself, equivalent to the punched card column containing punches in 0-2-8, that is, rows 0, 2, and 8. The seven-bit character was encoded as CBA8421, where C was the odd parity bit, bits 8, 4 , 2 and 1 were directly translatable from the punched card, bits in B and A were generated from the 12-punch, a bit in B but not in A from the 11- punch, and a bit in A but not in B from the 0-punch, but only when the 0- punch was accompanied by one or more other numeric punches. For example, the character "P" was punched as 11-7, which translates in memory to B421, and since the number of bits is even, C was also "on".
I'm pretty sure the encodings for the 1401 and 1410 were identical, because we used a 1401 to create mag tape input to a 1410 and I don't recall any translation issues. Somebody is likely to point out that the addressing schemes were not the same in the two machines and this is true. The 1401 used three characters to encode a five-digit address, whereas the 1410 used five. I can't imagine that anybody would want to know about this piece of arcane trivia, so if you want the lecture you'll have to let me know.
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> 1401 architecture
______________________________________________________________________
Community Memory: Discussion List on the History of Cyberspace ______________________________________________________________________
John Ahlstrom wrote "I believe there was an 8th bit (in the 1401) the record mark..."
I gave my 1401 manual to the Archive for the History of Computing in Manchester, but I kept a bunch of handy-dandy pocket references, including 1401 and 1410. The 1401 ref that I have is not specific enough, but the 1410 has a table of character codes. This is how it appears to have been--in core storage, the 1401 was a seven-bit character and the 8th bit was used for a word mark. The record mark was a character itself, equivalent to the punched card column containing punches in 0-2-8, that is, rows 0, 2, and 8. The seven-bit character was encoded as CBA8421, where C was the odd parity bit, bits 8, 4 , 2 and 1 were directly translatable from the punched card, bits in B and A were generated from the 12-punch, a bit in B but not in A from the 11- punch, and a bit in A but not in B from the 0-punch, but only when the 0- punch was accompanied by one or more other numeric punches. For example, the character "P" was punched as 11-7, which translates in memory to B421, and since the number of bits is even, C was also "on".
I'm pretty sure the encodings for the 1401 and 1410 were identical, because we used a 1401 to create mag tape input to a 1410 and I don't recall any translation issues. Somebody is likely to point out that the addressing schemes were not the same in the two machines and this is true. The 1401 used three characters to encode a five-digit address, whereas the 1410 used five. I can't imagine that anybody would want to know about this piece of arcane trivia, so if you want the lecture you'll have to let me know.
Keith Reid-Green
KReid-Green@ets.org
Princeton, NJ
______________________________________________________________________
