3 ms·

What a strange sentence: "When IBM released its game-changing System/360 in 1964, the head of the development team, Frederick Brooks, decided that its punch ca

by Woodi 6y ago

What a strange sentence: "When IBM released its game-changing System/360 in 1964, the head of the development team, Frederick Brooks, decided that its punch cards and printers were not yet capable of using ASCII." Brooks decided or just stated a state of reality that currently available peripherial hardware is uncapable of using newly proposed standard ? Also how - as usual - convenient, that IBM keeped machines uncompatibe while proposing others to make their systems accesible :> Also 'ASCII: Origins' is more fashionable - more like what movies do :)

jecel 6y ago
There were changes in 1965 and the 1967 version is considered the official one. Companies who adopted ASCII before that, like Xerox and DEC, had a left arrow instead of "_" and up arrow instead of "^" in their equipment. When Smalltalk-78's internal character set was replaced by ASCII in Smalltalk-80, it was the 1963 ASCII that they adopted since that is what they knew at Xerox. Though the goal was to be friendly to the rest of the world, the fact that almost everyone else saw their assignments as underscore causes problems to this day. So it is possible that 1964 was too early for IBM to embrace ASCII. But to be fair, the 1963 x 1967 ASCII problems are trivial compared to any ASCII x EBCDIC.
BareNakedCoder 6y ago
FYI, the IBM S/360 did support ASCII[0]. A bit in the program status word indicated if the machine was running in ASCII or EBCDIC mode. Most customers ran EBCDIC because it was more compatible with their masses of punch card originated data (no translation required). For the S/370, for virtual memory, a bit in the program status word was needed to indicate addresses were real or virtual. IBM surveyed its customers and found almost no one was using ASCII so its bit was taken for virtual memory. [0] Page 70 of http://bitsavers.org/pdf/ibm/360/princOps/A22-6821-0_360PrincOps.pdf http://bitsavers.org/pdf/ibm/360/princOps/A22-6821-0_360Prin...

by Woodi 6y ago

jecel 6y ago

There were changes in 1965 and the 1967 version is considered the official one. Companies who adopted ASCII before that, like Xerox and DEC, had a left arrow instead of "_" and up arrow instead of "^" in their equipment. When Smalltalk-78's internal character set was replaced by ASCII in Smalltalk-80, it was the 1963 ASCII that they adopted since that is what they knew at Xerox. Though the goal was to be friendly to the rest of the world, the fact that almost everyone else saw their assignments as underscore causes problems to this day. So it is possible that 1964 was too early for IBM to embrace ASCII. But to be fair, the 1963 x 1967 ASCII problems are trivial compared to any ASCII x EBCDIC.

BareNakedCoder 6y ago

FYI, the IBM S/360 did support ASCII[0]. A bit in the program status word indicated if the machine was running in ASCII or EBCDIC mode. Most customers ran EBCDIC because it was more compatible with their masses of punch card originated data (no translation required). For the S/370, for virtual memory, a bit in the program status word was needed to indicate addresses were real or virtual. IBM surveyed its customers and found almost no one was using ASCII so its bit was taken for virtual memory. [0] Page 70 of http://bitsavers.org/pdf/ibm/360/princOps/A22-6821-0_360PrincOps.pdf http://bitsavers.org/pdf/ibm/360/princOps/A22-6821-0_360Prin...