The answer to the “why” part is that generally “disk” is more American, and “disc” is more European - IBM (USA) brought out the hard drive, while Philips (Netherlands) and Sony (Japan) brought out CDs. Everything else really follows from there.

My Uncle, an old-school Cold War computer geek, was in the habbit of saying “Diskette” for most of my life. I haven’t heard him say it in a while because we really haven’t used Floppies of any sort for about ten years. I used to cringe when he’d say diskette because he would emphasize the second part of the word (disk-ETTE) whereas I always thought it was the Disk that gets the emphasis - hence the abbreviation.

[...] Disc or Disk? Link to the Post [...]

Showing my age, when I worked for Sperry UNIVAC (Blue Bell, Pennsylvania) back in the 70’s, and well before optical media, UNIVAC used the term “Disc”. Then again, they also used 1’s Complement math as well :-)

Bob Harris

disc = optical Media
disk = magnetooptical Medias
if only-readeble or also rewriteable doen’t value … just optical (c) vs. magnetooptical (k) :-)
cya

Scott is the only one who is correct here. “Disk” is short for “Diskette”. End of story.

The origin is debatable.

Possibly introduced when floppies went from 8 inch to 5.25 inch to emphasie the smaller size.

It is also thought that the term may be a marketing term, introduced when computers made there way into the home. “Diskette” sounding like “cassette”, a device home users were familiar with at the time. The use of a diskette is analogous with that of a cassette.

The term ‘disc’ refers to the shape of the substrate that is coated with a magnetic material in the original patent (# 2,690,913) for a “Magnetic Memory Device” filed by Jacob Rabinow on Oct 5, 1954 and later in the patent (# 3,134,997) for a “Magnetic Storage Machine” by the IBM team on May 19,1964.

I’ve been in this industry for over 35 years in both the military and commercial arenas and the short term ‘disc’ always referred to a non-flexible substrate magnetic hard disc, both removable and fixed. The short term ‘disk’ always referred to a flexible substrate magnetic ‘disc-shaped’ media like the 8″ and 5 1/4″ floppy disks. As previously mentioned ‘disk’ is short for diskette, a marketing term. Most people don’t know that the diskette was developed by IBM to load firmware and diagnostics into their main-frames.

Bob Harris scribbled:
when I worked for Sperry UNIVAC (Blue Bell, Pennsylvania) back in the 70’s, and well before optical media, UNIVAC used the term “Disc”. Then again, they also used 1’s Complement math as well :-)

In the early ’70s, I programmed the Univac 418-II (an 18-bit, load & store machine with 1’s-complement arithmetic in assembly language. For you guys who are still wet behind the ears, 1’s complement has two representations for zero — all zeroes and all ones. So, to test for zero after an arithmetic instruction with signed numbers, you had to test both cases. Which flavor of zero you got depended on the operation and the arguments. A particular form of pain inflicted on programmers by hardware designers cutting corners. (At least, that’s my theory.)

Years later, I heard Maurice V. Wilkes give a lecture at the Computer Museum, at that time housed in MR-2 (Building 2 in DEC’s Marlboro campus, which they bought from RCA after they quit the computer biz, in DEC’s first expansion from the legendary Mill back in Maynard, MA). Maurice Wilkes was one of the computer pioneers, from before the days of commercial computers. He and his students at Cambridge built one of the early research project computers, the EDSAC, which became operational in 1949.

From the early days, until the mid-60s, with 3rd generation machines, 2’s-complement was not universally used in computer designs like it is today. For binary math, two other representations were common: signed magnitude and 1’s-complement. Apparently, hardware designers chose them because they were less expensive (in those days, a single flip-flop or a couple of gates were a whole circuit card made of discrete components. Chips? Never heard of ‘em.) or easier to build.

During the Q&A after the lecture, I asked Wilkes why people built machines with 1’s-complement arithmetic. His answer: Because they were mad.

The EDSAC project broke a lot of new ground. The microcoded CPU was a real innovation. Wilkes also produced the first assembler language with macro instructions, called “synthetic orders” on the other side of the puddle. To the delight of everyone there, he showed a little movie from about 1951 introducing the EDSAC. Imagine the laughter from a room full of systems programmers with these two highlights.

1. In those days, there was no OS or run-time enviornment; programs ran on the bare iron. After the tweed-jacketed programmer consulted with someone who needed to get something done on the EDSAC, he designed a program and coded it, all the while puffing on his pipe. If he needed a function from the subroutine library, he went to a cabinet with little drawers, like an old-school card catalog, and pulled out the appropriate piece of paper tape for the clerks to splice into the program.

2. After paper tapes had been prepared for the the program and data, the user would write up a little job ticket and clip it and the paper taoes together with a clothespin and hang the whole mess on a wire. This was called entering a job into the queue.

It was really cute seeing how just how literally the terms we knew as abstractions were realized in physical terms.

Oh, speaking of bogosities from the 418s, each of the peripherals used a different character code:

* Card-reader & punch:6-bit XS-3 (excess three. Don’t ask.) characters. The cards themselves were punched in a code derived from the systems Herman Hollerith built for the 1890 US Census. The binary values exchanged between the CPU and the devices used XS-3 for its error detecting karma.
* Line printer: Fielddata. Kinda, sorta like 7-bit ASCII but not quite. I think Fielddata was created for some military project, so you can blame this one on DoD.
* Mag Tape: a 6-bit ANSI code which was a precursor of ASCII. (The tape-drive’s head had 7 coils for eight tracks, the seventh bit being used for parity.) Lower-case letters? We don’t need no stinking lower-case letters.

In those days, a byte was a field of consecutive bits in a word, where the field length was a divisor of the word length. Most-machines resolved addresses to words, There were instructions to squish a series of small values, like characters, into a word or extract them. So a 36-bit machine might have support bytes that were 6, 9, 12 or 18 bits long. These architectures were created before the IBM S/360 which had instructions both for binary (32-bit words) and decimal arithmetic on 4-bit BCD digits in a field of 8-bit bytes. S/360s were byte-addressible, but you had to worry about alignment: word instructions required addresses that were aligned on word boundaries, i.e., 0 (modulo 4) which is a fancy way of saying evenly devisible by 4, half-word instructions’ operands on half-word boundaries, i.e., even addresses, and double-word … — well, you get the idea.

Ah, the good old days. May they never come again.

Hey, Bob! Remember the Unisys FASTRAND? (For you youngsters, it was an early magnetic drum device — weighed more than two tons and was about 5 ft. high, 10 ft. long a couple of feet deep, if I recall correctly. And noisy. Loud hum from the motors and massive rotating drums, a big whoosh of air from the cooling fans and then the clickety-clack knocking from the comb assembly that held the heads as it zipped back and forth seeking to particular tracks. You never forget your encounters with hardware like that. Oh, yeah, The FASTRAND II had two long counter-rotating drums, one mounted directly above the other. That’s ‘cuz the original FASTRAND was inclined to walk across the floor — its single drum was a gyroscope. D’oh! As impressive a machine as it was, Its capacity (roughly 90 MB formatted as we measure things today) and speed were less than what you can get with a litlle flash drive today.

Well, this is pretty long, but Bob’s comment sparked a whole bunch of memories.

Two quick notes following up on Tim Stoddard’s comments,

1. My first encounter with floppies was just as he says, the microcode iloader for the IBM 3330 (not entirely sure I got the model number right) disc controllers. The floppy diskettes were about 10″ in diameter, and really did flop because they were not in paper sleeves. That was OK because the microcode rarely changed, so there was no need to make it easy or convenient to swap diskettes. Starting up the university’s S/370-168 was a real trip with the whirring and clacking from all the floppies in the various controllers scattered around in the dinosaur pen.

2. Jacob Rabinow was a prolific inventor. He designed the first linear or straight-line tracking turntables and started a company in 1968 to build and market them. My dead Rabco ST-4 sits in a box in the basement — can’t bear to toss it. The tone-arm design was ingenious and worked well, but the turntable itself was pretty funky, so Rabco didn’t do too well. Harmon/Kardon bought the company from Rabinow and used his tonearm design with their much more reliable turntables.

Computer Maintenance Tips…

I couldn’t understand some parts of this article, but it sounds interesting…