4.0460 Computers: Natures and Names (2/40)
Elaine Brennan & Allen Renear (EDITORS@BROWNVM.BITNET)
Thu, 6 Sep 90 17:59:09 EDT
Humanist Discussion Group, Vol. 4, No. 0460. Thursday, 6 Sep 1990.
(1) Date: Fri, 31 Aug 90 08:45:01 MDT (22 lines)
From: koontz@alpha.bldr.nist.gov (John E. Koontz)
Subject: Re: 4.0440 Nature of Computers
(2) Date: Tue, 28 Aug 90 15:14:00 CDT (18 lines)
From: "Michael S. Hart" <HART@UIUCVMD>
Subject: Re: 4.0415 Computer Names
(1) --------------------------------------------------------------------
Date: Fri, 31 Aug 90 08:45:01 MDT
From: koontz@alpha.bldr.nist.gov (John E. Koontz)
Subject: Re: 4.0440 Nature of Computers
I am sorry if my description of the IBM 1620 was confusing. I thought I
had made it clear that it used 4-tuples or 5-tuples - I don't remember
which - of bits to represent decimal digits. (The extra states beyond
ten had special functions.) In other words, it did not use special
ten-state devices; it used collections of two-state devices. The basic
addressable unit was the digit (call it a 4 or 5 bit word, if you like).
Numbers were represented by marking the high order digit with a special
code called a flag. The number extended from the addressed digit to the
flagged digit. The manuals contained sections explaining the underlying
binary basis of the system, but assembly and machine level programming
were at the decimal digit level. There was no way to manipulate the
bits directly. Machine language commands consisted of a two digit
number followed by two fixed length fields of digits representing
addresses or constants. Addition and subtraction, etc., were performed
in memory - there were no user accessible registers for arithmetic. I
doubt there are any functional 1620s anywhere today, but the manuals are
still available. I may have some buried somewhere. In its time the
1620 was a very important machine.
(2) --------------------------------------------------------------25----
Date: Tue, 28 Aug 90 15:14:00 CDT
From: "Michael S. Hart" <HART@UIUCVMD>
Subject: Re: 4.0415 Computer Names
Well, the names the manufacturers put on them may have diminished in
quality, but the names the users label them are just the opposite.
If you need examples . . . .
Michael S. Hart
PS For those who don't know what I am talking about, look at the names
in the relay systems shown in the headers. I receive from at least
several of the seven dwarfs, have worked on repairing Darth Vader,
and the time it takes for a name to appear (such as Sugarbombs in Calvin)
on a computer is sometimes apparently less than a femtosecond.)(btw - I
finally tracked down the names for information larger than megabytes and
gigabytes and terabytes - who else knows what lies between them and the
large sized whamtobytes?)