27 Jan
1999
27 Jan
'99
10:59 a.m.
Gary Oliver <go(a)ao.com> wrote about 7-track magtape formats:
In "binary" mode with odd parity, all 6 bit
codes could be represented since
there would never be a field of all 0's. I've never found out why there
needed to be both of these modes rather than just use binary mode for
everything. Anyone know the reason for this?
I can hazard a guess. According to _IBM's_Early_Computers_ by Bashe et al.
(an awesome book, but unfortunately out of print), their early "business"
computers (702, 705, 705 III) used even parity, and their "scientific"
computers (701, 704, 709, etc.) used odd parity. (I hope I've got that right,
I don't have the book here to check.)
While the "scientific" computers could deal with aribtrary binary data,
the "business" computers only dealt with characters. They did arithmetic
using fields of numeric characters. So the business computers could easily
restrict their tape requirements to avoid the use of the all-zeros character
code, whereas the scientific computers could not.
At the time that the 701 and 702 were developed, there wasn't any tape
standard that they had to be compatible with, so they were free to make
arbitrary decisions. Apparently they didn't even care about tape interchange
between the 701 and 702.
Of course, this doesn't really explain *why* even parity was chosen for
the business computers. With the benefit of 47 years of hindsight, it now
seems like a stupid decision, but that is probably only because I don't know
what other factors influenced the decision.