On flip-flops, what does J, K, and Q mean?

[Freddy1X]

Since like forever, any tine I have worked with flip-flops, I will encounter
inputs and outputs such as S, R, D, J, K, C, T, and Q / !Q. For most of the
connections the letter to name designation has been given but I've never
seen a reason given for why they carry the labels J, K, and Q. There might
be other letters used depending on what the flip-flop is designed to do.

S =Set
R =Reset
D =Data
J = ?
K = ?
C =Clock
T =Toggle
Q = ?

So why use J, K, and Q labels?

[Quadibloc]

Post by Freddy1X
Since like forever, any tine I have worked with flip-flops, I will encounter
inputs and outputs such as S, R, D, J, K, C, T, and Q / !Q. For most of the
connections the letter to name designation has been given but I've never
seen a reason given for why they carry the labels J, K, and Q. There might
be other letters used depending on what the flip-flop is designed to do.
S =Set
R =Reset
D =Data
J = ?
K = ?
C =Clock
T =Toggle
Q = ?
So why use J, K, and Q labels?

This is a good question.

In the case of Q, I can think of one reason.

O = Output has one problem... it can be confused with 0 = Zero. Q is similar to O, but
it has a distinguishing feature.

J is associated with I, which looks similar to 1, so it makes sense for an input that
sets the flip-flop to a 1 state but is not the same as the S input. K follows J, _and_
is associated with C (for Clear instead of Clock).

The Wikipedia article

https://en.wikipedia.org/wiki/Flip-flop_(electronics)

hints at some other possibililties.

John Savard

[Carlos E.R.]

Post by Quadibloc

Post by Freddy1X
Since like forever, any tine I have worked with flip-flops, I will encounter
inputs and outputs such as S, R, D, J, K, C, T, and Q / !Q. For most of the
connections the letter to name designation has been given but I've never
seen a reason given for why they carry the labels J, K, and Q. There might
be other letters used depending on what the flip-flop is designed to do.
S =Set
R =Reset
D =Data
J = ?
K = ?
C =Clock
T =Toggle
Q = ?
So why use J, K, and Q labels?

This is a good question.
In the case of Q, I can think of one reason.
O = Output has one problem... it can be confused with 0 = Zero. Q is similar to O, but
it has a distinguishing feature.
J is associated with I, which looks similar to 1, so it makes sense for an input that
sets the flip-flop to a 1 state but is not the same as the S input. K follows J, _and_
is associated with C (for Clear instead of Clock).

I had a memotechnical rule to remember which was which. J for jump (goes
up), K for KO (goes down).

Post by Quadibloc
The Wikipedia article
https://en.wikipedia.org/wiki/Flip-flop_(electronics)
hints at some other possibililties.
John Savard

[Charlie Gibbs]

Post by Carlos E.R.

Post by Quadibloc

Post by Freddy1X
Since like forever, any tine I have worked with flip-flops,
I will encounter inputs and outputs such as S, R, D, J, K, C, T,
and Q / !Q. For most of the connections the letter to name
designation has been given but I've never seen a reason given
for why they carry the labels J, K, and Q. There might be other
letters used depending on what the flip-flop is designed to do.
S =Set
R =Reset
D =Data
J = ?
K = ?
C =Clock
T =Toggle
Q = ?
So why use J, K, and Q labels?

This is a good question.
In the case of Q, I can think of one reason.
O = Output has one problem... it can be confused with 0 = Zero.
Q is similar to O, but it has a distinguishing feature.
J is associated with I, which looks similar to 1, so it makes sense
for an input that sets the flip-flop to a 1 state but is not the same
as the S input. K follows J, _and_ is associated with C (for Clear
instead of Clock).

I had a memotechnical rule to remember which was which. J for jump
(goes up), K for KO (goes down).

Unfortunately that conflicts with the mnemonic I learned for
cursor movement keys in vi: J for jump down, K for klimb up.

[Ahem A Rivet's Shot]

On Mon, 11 Apr 2022 19:46:23 GMT

Post by Charlie Gibbs
Unfortunately that conflicts with the mnemonic I learned for
cursor movement keys in vi: J for jump down, K for klimb up.

I never knew a mnemonic but then I saw it as just a stripped down
WordStar diamond on the wrong hand which didn't take as long to adjust to
as the modal behaviour which now seems totally natural.

[David Mills]

Post by Ahem A Rivet's Shot
I never knew a mnemonic but then I saw it as just a stripped down
WordStar diamond on the wrong hand which didn't take as long to adjust to
as the modal behaviour which now seems totally natural.

It was the arrow keys on the ADM-3A terminal - https://en.wikipedia.org/wiki/ADM-3A

[Douglas Miller]

It's been an interesting voyage into vacuum tube computers and patents from the dawn of electronic computers, but I've not found documentation that describes the origin of "J", "K", and "Q" ("/Q"). The earliest use of those labels, that I've found, is Nelson's patent US2850566 filed Sep 8, 1953. An earlier patent of his US2823855 Nov 26, 1952 describes the same flip-flop with inputs "1" and "0" and output "F" ("/F"). I see a few later patents (from others at Hughes Aircraft) that start using "Q" for outputs, but not finding "J" and "K" inputs. I've not seen anything matching this convention in various vacuum tube computers of the era, either, although they don't really describe a flip-flop exactly like this and probably wouldn't use anything this complex if not absolutely necessary.

What I can't tell is whether Nelson coined this nomenclature and it just eventually caught on, or whether there actually were some standards being proposed in the early 1950s and Nelson was an early adopter.

Note, Nelson's flip-flop does differ from a modern 74LS76 in a few details. Nelson's flip-flops (and, it appears, vacuum tube flip-flops in general) operate on pulses being applied to inputs (although there is some hint at things like a "D flip-flop" where an "input level" determines the resulting state after a pulse to another input), so there is no separate "CLK". But, it seems more than coincidence that Nelson describes a flip-flop that is similar to a modern JK flip-flop while using input labels of "J" and "K".

I could not find one earlier patent of Nelson's, ""High-Speed Flip-Flop Counter" filed September 10, 1951, although it seems unlikely it would use "J" and "K" (and "Q"). I also don't have access to the IEEE publications that might shed light on proposed standards.

[Ahem A Rivet's Shot]

On Thu, 07 Apr 2022 13:38:44 -0400

Post by Freddy1X
Since like forever, any tine I have worked with flip-flops, I will
encounter inputs and outputs such as S, R, D, J, K, C, T, and Q / !Q.
For most of the connections the letter to name designation has been given
but I've never seen a reason given for why they carry the labels J, K,
and Q. There might be other letters used depending on what the flip-flop
is designed to do.

All you ever wanted to know about flip flops and then some.

https://www2.ph.ed.ac.uk/~eisenhar/teaching/EM/2012_EM_LectureNotes_02.pdf

There's more in _03.pdf.

Post by Freddy1X
So why use J, K, and Q labels?

'cos it's a JK flip-flop - last two pages of the notes which start:

------
The circuit we need to build is not trivial – it goes by the name JK
Flip-Flop. People rarely go completely mad trying to understand how it
works, usually just a little bit
------

Enjoy!

[Douglas Miller]

Reading Nelson's patent, he mentions J and K inputs very casually, as if everyone already knows what they are. Also, that patent is for a high-speed printing system and not the invention of a type of flip-flop. I see no mention of prior-art there referencing the JK flip-flop, either. It is entirely possible that the JK flip-flop never received a patent, for reasons such as it was widely known/used by the time anyone thought to patent it or it originated in an environment (e.g. academia) where patents were not sought. I have not found a patent for the implementation of a JK flip-flop prior to 1953 (i.e. using vacuum tubes), although various such patents exist for IC (transistor) implementations in the 60's and 70's. I'm inclined to think that the JK flip-flop was an academic concept and that a meaning for "J" and "K" would have to come from the original documentation for that concept.

[Douglas Miller]

Post by Douglas Miller
Reading Nelson's patent, he mentions J and K inputs very casually, as if everyone already knows what they are. Also, that patent is for a high-speed printing system and not the invention of a type of flip-flop. I see no mention of prior-art there referencing the JK flip-flop, either. It is entirely possible that the JK flip-flop never received a patent, for reasons such as it was widely known/used by the time anyone thought to patent it or it originated in an environment (e.g. academia) where patents were not sought. I have not found a patent for the implementation of a JK flip-flop prior to 1953 (i.e. using vacuum tubes), although various such patents exist for IC (transistor) implementations in the 60's and 70's. I'm inclined to think that the JK flip-flop was an academic concept and that a meaning for "J" and "K" would have to come from the original documentation for that concept.

The other aspect is that we have no reason to believe that "J", "K", or even "Q", have any correlation to the English language. There was a lot of computer science coming out for Germany, for example, and so the letters might correspond to German words/terms - or some other language entirely.

[Don Poitras]

Post by Douglas Miller

Post by Douglas Miller
Reading Nelson's patent, he mentions J and K inputs very casually, as if everyone already knows what they are. Also, that patent is for a high-speed printing system and not the invention of a type of flip-flop. I see no mention of prior-art there referencing the JK flip-flop, either. It is entirely possible that the JK flip-flop never received a patent, for reasons such as it was widely known/used by the time anyone thought to patent it or it originated in an environment (e.g. academia) where patents were not sought. I have not found a patent for the implementation of a JK flip-flop prior to 1953 (i.e. using vacuum tubes), although various such patents exist for IC (transistor) implementations in the 60's and 70's. I'm inclined to think that the JK flip-flop was an academic concept and that a meaning for "J" and "K" would have to come from the original documentation for that concept.

The other aspect is that we have no reason to believe that "J", "K", or even "Q", have any correlation to the English language. There was a lot of computer science coming out for Germany, for example, and so the letters might correspond to German words/terms - or some other language entirely.

Ben Eater describes the JK flip-flop here:

At 6:30 he says the letters don't have any special meaning that he knows of.

[Douglas Miller]

Post by Don Poitras
http://youtu.be/F1OC5e7Tn_o
At 6:30 he says the letters don't have any special meaning that he knows of.
--
Don Poitras

I agree, there may be no special meaning. But, this video is not compelling evidence of that since it could be 80-100 years after the invention of the JK flip-flop. Like most of the rest of us that have learned electronics after the dawn of integrated circuits, he simple has not heard of an explanation. It sounds to me as though the term was in use at least 20 years before I first saw a JK flip-flop.

There are a lot of sources on the internet claiming the Jack Kilby is the "JK", but Nelson's 1953 patent calling out the J and K inputs to a flip-flop seems to predate any possible invention by Jack Kilby - unless it was from academic work he did prior to working for TI.

[Charlie Gibbs]

Post by Douglas Miller

Post by Don Poitras
http://youtu.be/F1OC5e7Tn_o
At 6:30 he says the letters don't have any special meaning that he knows of.

I agree, there may be no special meaning. But, this video is not
compelling evidence of that since it could be 80-100 years after
the invention of the JK flip-flop. Like most of the rest of us that
have learned electronics after the dawn of integrated circuits, he
simple has not heard of an explanation. It sounds to me as though
the term was in use at least 20 years before I first saw a JK
flip-flop.
There are a lot of sources on the internet claiming the Jack Kilby
is the "JK", but Nelson's 1953 patent calling out the J and K inputs
to a flip-flop seems to predate any possible invention by Jack Kilby -
unless it was from academic work he did prior to working for TI.

Now that we've sorted out JK flip-flops, we can move on to the J and K
valves on scuba tanks. Rumour has it that these were the designating
letters in an illustrated catalog of dive equipment.

[Gordon Henderson]

Post by Charlie Gibbs
Now that we've sorted out JK flip-flops, we can move on to the J and K
valves on scuba tanks. Rumour has it that these were the designating
letters in an illustrated catalog of dive equipment.

Quite possibly, but the "J" valve I've head call the "Jesus" Valve because
some people may well say something like "Oh Jesus" when they run out of air
and have to pull it to get the reserve..

Gordon

[Charlie Gibbs]

Post by Gordon Henderson

Post by Charlie Gibbs
Now that we've sorted out JK flip-flops, we can move on to the J and K
valves on scuba tanks. Rumour has it that these were the designating
letters in an illustrated catalog of dive equipment.

Quite possibly, but the "J" valve I've head call the "Jesus" Valve because
some people may well say something like "Oh Jesus" when they run out of air
and have to pull it to get the reserve..

Good one. Just my luck that I had a K valve when I ran out of air at 60 feet.
I found that you can drag one or two more breaths out of the tank as you ascend
and the ambient pressure decreases.

Thank goodness someone soon came up with those newfangled pressure gauges...

[Peter Flass]

Post by Douglas Miller

Post by Don Poitras
http://youtu.be/F1OC5e7Tn_o
At 6:30 he says the letters don't have any special meaning that he knows of.
--
Don Poitras

I agree, there may be no special meaning. But, this video is not
compelling evidence of that since it could be 80-100 years after the
invention of the JK flip-flop. Like most of the rest of us that have
learned electronics after the dawn of integrated circuits, he simple has
not heard of an explanation. It sounds to me as though the term was in
use at least 20 years before I first saw a JK flip-flop.
There are a lot of sources on the internet claiming the Jack Kilby is the
"JK", but Nelson's 1953 patent calling out the J and K inputs to a
flip-flop seems to predate any possible invention by Jack Kilby - unless
it was from academic work he did prior to working for TI.

Some of these unices are specialized for real-time or embedded systems. QNX
looks like an OS for embedded, but I think there are several others around.

[Freddy1X]

Post by Peter Flass

Post by Douglas Miller

Post by Don Poitras
http://youtu.be/F1OC5e7Tn_o
At 6:30 he says the letters don't have any special meaning that he knows of.
--
Don Poitras

I agree, there may be no special meaning. But, this video is not
compelling evidence of that since it could be 80-100 years after the
invention of the JK flip-flop. Like most of the rest of us that have
learned electronics after the dawn of integrated circuits, he simple has
not heard of an explanation. It sounds to me as though the term was in
use at least 20 years before I first saw a JK flip-flop.
There are a lot of sources on the internet claiming the Jack Kilby is the
"JK", but Nelson's 1953 patent calling out the J and K inputs to a
flip-flop seems to predate any possible invention by Jack Kilby - unless
it was from academic work he did prior to working for TI.

Some of these unices are specialized for real-time or embedded systems.
QNX looks like an OS for embedded, but I think there are several others
around.

I appreciate everyone's reply. It's a question that I never knew I had
until a few days ago. Like many things, J and K are accepted as-is without
explanation. Like measuring temperature in Fahrenheit, or length in inches.
If you look deep enough, you can usually find out why it is done that way.

I wouldn't say that the '53 patent is the definitive source. Perhaps
instead this is what nailed the definition down for fun and profit.

As for Q, after looking over discreet component circuits for flip-flops, I
suspect that it derived from the common use of Q( number) to designate
transistors in a circuit. But then where did transistors get the Q? Also
flip-flops existed long before transistors were invented. Was anything
similar standardized in the vacuum tube or relay days? I know that relays
had their 'K' designation for quite a while.

Freddy,
turning over rocks.

[Ray Hughes]

Post by Freddy1X

Post by Douglas Miller

Post by Don Poitras
http://youtu.be/F1OC5e7Tn_o
At 6:30 he says the letters don't have any special meaning that he knows of.
--
Don Poitras

<<<< some snipping below>>>

Post by Freddy1X

Post by Douglas Miller
I agree, there may be no special meaning. But, this video is not
compelling evidence of that since it could be 80-100 years after the
invention of the JK flip-flop. Like most of the rest of us that have
learned electronics after the dawn of integrated circuits, he simple has
not heard of an explanation. It sounds to me as though the term was in
use at least 20 years before I first saw a JK flip-flop.

I appreciate everyone's reply. It's a question that I never knew I had

ponent circuits for flip-flops, I

Post by Freddy1X
suspect that it derived from the common use of Q( number) to designate
transistors in a circuit. But then where did transistors get the Q? Also
flip-flops existed long before transistors were invented. Was anything
similar standardized in the vacuum tube or relay days? I know that relays
had their 'K' designation for quite a while.
Freddy,
turning over rocks.

Alan Turing used the letter q to denote states in what came to be known as Turing machines. Presumably the q stood for quanta, emphasizing a state's discrete rather than continuous nature. This happened in the 30s when quantum theory was permeating the scientific æther.

Cite

answered Dec 27, 2012 at 19:26
user avatar
I. J. Kennedy
37711 silver badge9
######################################

RayH(the other one)

[Douglas Miller]

I'm not sure why I missed this before, but the wikipedia article https://en.wikipedia.org/wiki/Flip-flop_(electronics) states in the History section that the flip-flop was invented in 1918, and points to Hughes Aircraft using a type of flip-flop that we today call JK, and cites Nelson's use of the J and K inputs in his 1953 patent. However, this doesn't tell us whether Hughes "invented" the JK flip-flop or whether it was a concept that they simply used extensively, nor whether it had a name before. I would like to read more on Jordan and Eccles, and other inventions between then and 1943 Colossus, to see if there are any papers on other types of flip-flops.

[Douglas Miller]

some interesting information on the EDVAC (https://babel.hathitrust.org/cgi/pt?id=mdp.39015002095639&view=1up&seq=561) talks about use of the "Eccles-Jordan type trigger circuit" flip-flop (S-R), but also a toggling flip-flop ("Binary Counter") where a pulse on one input causes the outputs to reverse state (i.e. J=K=1), plus a "clear" input that is hard-wired to either set or reset the FF. This is clearly not a JK flip-flop, but likely represents a predecessor (a JK could be probably achieved by adding logic to the inputs). I didn't find schematics, but not being a tube person they probably wouldn't help me understand it better. This may help fill in the timeline, showing a predecessor to the JK being used in the mid 1940s. By 1953 it would seem that Hughes Aircraft was using a flip-flop that functionally resembles the JK (although I don't think we have a good description of what Nelson's "J" and "K" inputs actually do).

[Douglas Miller]

After reading Nelson's 1953 patent more, I have doubts as to whether his J and K inputs refer to a JK flip-flop. Based on the block diagram, logic expressions, and description, it sounds more like the J and K inputs are essentially S and R inputs. External logic is combining a clock pulse based on other bits (stages) and control logic to produce a pulse on either his J or K input, and the conditions seem to indicate that these inputs are performing S and R functions. This patent may not represent the existence of JK flip-flops in 1953.

[Douglas Miller]

Post by Douglas Miller
After reading Nelson's 1953 patent more, I have doubts as to whether his J and K inputs refer to a JK flip-flop. Based on the block diagram, logic expressions, and description, it sounds more like the J and K inputs are essentially S and R inputs. External logic is combining a clock pulse based on other bits (stages) and control logic to produce a pulse on either his J or K input, and the conditions seem to indicate that these inputs are performing S and R functions. This patent may not represent the existence of JK flip-flops in 1953.

Correction: I found this description of the flip-flops in the patent:

"Each flip-flop or bistable multivibrator includes two
input terminals, hereinafter termed the j-input and the k-
input terminals, respectively, and two output terminals
for producing complementary bivalued electrical output
signals hereinafter termed Q and /Q, respectively. Sig-
nals applied separately to the j-input and k-input ter-
minals set the flip-flop to conduction states corresponding
to the binary values one and zero, respectively, while sig-
nals applied simultaneously to both input terminals trigger
or change the conduction state of the flip-flop."

So, this describes something very close to the modern JK flip-flop, the difference being that the J and K inputs are strictly pulses (there is no clock input), rather than control signals that determine how the clock input functions in modern JK flip-flops.

[Dennis Boone]

Post by Douglas Miller
the wikipedia article https://en.wikipedia.org/wiki/Flip-flop_(electronics

It also has the following gloss for the types:

Flip-flops can be divided into common types: the SR ("set-reset"), D
("data" or "delay"[13]), T ("toggle"), and JK.

Still doesn't explain what J and K stand for, if anything.

De

[Douglas Miller]

Post by Dennis Boone
Still doesn't explain what J and K stand for, if anything.
De

Another clue is in U.S. Patent 2,828,855 from Nov 1952, also by Nelson. This describes the flip-flop used in the later patent, but describes the two inputs as "1" and "0". There's even a schematic of the flip-flop (using vacuum tubes, of course) and a detailed description of operation. So, at least as far as Nelson and Hughes Aircraft were concerned, somewhere between Nov 1952 and Sep 1953 the terminology for the inputs to this flip-flop changed from ("1" and "0") to ("J" and "K"). We still don't have an explanation for that change, whether it was internal to Hughes Aircraft or from some external source (i.e. the flip-flop was used elsewhere and already labeled "J" and "K"). I'm certainly no expert at interpreting patent lingo, but the level of detail describing the operation of the flip-flop might imply that this is the origin of what we now call a JK flip-flop. There are some other patents referenced, so it might be worth tracking those down to see if they shed any light on this flip-flop.

[Douglas Miller]

Patent 2,828,855 references an article by A.D. Booth: An Electronic Digital Computer, Electronic Engineering, December 1950; pages 492 to 498. I've not been able to locate an online copy of that article (publication), but I wonder if it describes this type of flip-flop. Other patent references seem to be about the components/ideas used to construct this type of flip-flop, but not the actual (JK) flip-flop, so it may be the case that this represents the origin of the JK flip-flop, at least as Intellectual Property (an implementation).

[Scott Lurndal]

Patent 2,828,855 references an article by A.D. Booth: An Electronic Digital=
Computer, Electronic Engineering, December 1950; pages 492 to 498. I've no=
t been able to locate an online copy of that article (publication), but I w=
onder if it describes this type of flip-flop. Other patent references seem =
to be about the components/ideas used to construct this type of flip-flop, =
but not the actual (JK) flip-flop, so it may be the case that this represen=
ts the origin of the JK flip-flop, at least as Intellectual Property (an im=
plementation).

I have this in box A023 in storage[*], I'll see about digging it out.

Booth, Andrew D. and Booth, Kathleen, H.V. Automatic Digital Calculators Hard A023

[*] Picked it up when they closed the Burroughs Library in the Pasadena
plant (which was originally the Electrodata plant).

[Scott Lurndal]

Post by Scott Lurndal

Patent 2,828,855 references an article by A.D. Booth: An Electronic Digital=
Computer, Electronic Engineering, December 1950; pages 492 to 498. I've no=
t been able to locate an online copy of that article (publication), but I w=
onder if it describes this type of flip-flop. Other patent references seem =
to be about the components/ideas used to construct this type of flip-flop, =
but not the actual (JK) flip-flop, so it may be the case that this represen=
ts the origin of the JK flip-flop, at least as Intellectual Property (an im=
plementation).

I have this in box A023 in storage[*], I'll see about digging it out.
Booth, Andrew D. and Booth, Kathleen, H.V. Automatic Digital Calculators Hard A023
[*] Picked it up when they closed the Burroughs Library in the Pasadena
plant (which was originally the Electrodata plant).

Copyright date: 1956

Chapter 2 is an interesting overview of mechanical computers
- Automatic Sequence Controlled Computer (ASCC)
- Harvard Mk. 2 (Add: 200ms, Mul: 700ms)
- Bell Labs machine (bi-quinary coded) (Add: 333ms, Mul: 1s, SQR: 4.3s)
- IBM Pluggable Sequence Relay Calculator (add: 25ms, Mul: 900ms, Div: 1.2s)
- BARK, Zuse
- Automatic Relay Calculator (built by the authors).

Chapter 3 highlights:
- Key electronic discovery (after that of the triode) was made as long
ago as 1919, when Eccles and Jordan showed how a pair of triode valves
could be connected to form a circuit having two stable states.
- ENIAC consumed 100Kw, 18000 tubes.
- EDSAC, projected in 1941, completed 1949, 512 34-bit words Hg delay line storage
Can operate on half-words (17-bit), (17-bit add: 34us, 34-bit add: 70us, mul: 8.5ms
A particular advantage of EDSAC lies in the carefully constructed library
of subroutines, by means of which complex problems can be solved by assembling
together methods of solution previously used on simpler calculations
First machine operating on the principle of a large store containing both
numbers and instructions to be put into actual computing service.
- ACE (1950). 512 32-bit words. (Add: 32us, Mul: 1ms), less than 1000 valves
- Plate 1: Nice B/W photo of Institute for Advanced Study computer, Princeton
40-bit words, CRT storage tube (Add: 10us, Mul: 300us)
(notes it takes 20 minutes to fill storage using punched paper tape)
- In construction at time of writing:
- SEAC (Bureau of Standards), replaces thermionic valves with semi-conductors
- UNIVAC (similar but working on decimal scale). Constructed by E&M for Census Bureau.
- Whirlwind (MIT), CRT storage, over 1000 16-bit words, 6000 valves, (add: 5us, Mul: 40us)
- Williams (Ferranti) 256 40-bit words (Add: 1.2ms, Mul: >3ms)
- TRE (Malvern, England), SWAC (inst. num.analysis, California), ORDVAC (U of Illinois)
all used CRT storage.
- Booth & Booth machine storage using "the magnetic remanence of ferromagnetic
materials" in 1947. Nickel-plated cylinder rotating at high speeds with reading
and recording head to enable binary data to be recorded in the form of
magnetized elements on its surface.
- SEC (Birkbeck college, london). 256 21-bit words. (add 1.6ms) , followed by
APE(X)C, APE(N)C - 1024 32-bit words (add: 600us, Mul: up to 20ms)
British Tabulating Machine Company HEC 2M, HEC 4.
- NORC (IBM for Navy) 2000 13-digit (and sign) words. (add: 15us, mul: 31us)
Includes elaborate terminal organs (I/O devices?). Cost $2.5million
- Experiments are now in progress, at Birkbeck College and MIT directed
at using very small toroids of ferromagnetic material, arranged at the lattice
points of a net of single conductors as a storage device. This technique was
suggested by one of the authors (booth or booth) at a lecture in Cambridge
in 1947 under the name 'iron nail store' seems very promising; access times
on the order of a few microseconds are easily attained and, despite its use
of macroscopic elements for stgorage, the device is compact and relatively
inexpensive.
- Discusses elimnation of thermionic valves with semi-conducting devices, or
replacing the gating element (valve) by a magnetic structure of equivalent
speed but effectively infinite life.

Chapter 9 (gates)
- "Electronic Flip-Flops". No use of the term 'J-K'
- Discusses non-binary storage (Dekatron, S.T.C-tron, Trochotron).

Chapter 17 (some applications of computing machines)
- Physics
- Mechanical Translation
"It the first place it must be stated, quite clearly, that
what is envisioned, by the present authors at least, is a
dictionary translation with possible grammatical notes. It
is in no way expected that a translation of literary quality
will be produced"
- Games
"A somewhat surprising application of automatic digital calculators
is that of an opponent in various games." Example: Whist, Naughts
and crosses.
- Machine Learning and Intelligence (Aside: I hope the author didn't have dogs).
"The problem with producing a conditioned reflex in a digital calculator
is much more complicated".
Discusses "training" a computer, where the disapproval stimulus is to reboot.

Sorry, I could find no reference specifically to J-K style flip-flops,
although one might consider that the 'J' is the important letter and
may have been assigned in honor of Jordan, and the K just logically followed.

[Dennis Boone]

I've ordered a copy of Phister's 1958 book. Will be interested to
see what sort of terminology it presents.

I tried searching the QST indexes with no joy, but since the web tool
only searches the title, that's probably not very definitive.

De

[Douglas Miller]

I've ordered a copy of Phister's 1958 book. Will be interested to
see what sort of terminology it presents.
I tried searching the QST indexes with no joy, but since the web tool
only searches the title, that's probably not very definitive.
De

I've not been able to find the exact article referenced in the patent. Booth seemed to write a lot. I'm guessing that the article is referenced because it is pertinent to the patent, but as to whether it mentions a flip-flop that operates like the ones Nelson describes is another matter. Note, it may not mention "J" and "K", but rather a FF that toggles which both inputs are "activated" together (pulsed in the case of vacuum tubes, it seems).

[Douglas Miller]

While looking at information about various early-50s vacuum tube computers, found this in the Univac section: http://www.bitsavers.org/pdf/univac/1101/ERA-1101-f41-StandardsElectronic_Mar51.pdf An IEEE publication defining computer terms. This particular issue defines only a generic flip-flop, and does not seem to differentiate (or acknowledge) types. It would be interesting to see other issues of this publication, however I don't have an IEEE membership anymore and can't find any other public copies. What I'm wondering is if there was any establishment of terms between Nelson's 1952 patent and 1953 patent that might explain him changing from inputs labeled "1" and "0" with outputs labeled "F" and "/F" to the later inputs labeled "J" and "K" with outputs labeled "Q" and "/Q". Of course, the source of that change might have been solely from Nelson, or from fellow Hughes colleagues, or from some other interaction with (outside) computer engineers.

The Univac drawings seem to only use SR flip-flops, although there is some talk of binary counters but I don't see anything about how those are implemented. I did find a schematic of a simple toggle flip-flop, but nothing sounding like a JK version. There is also some strange definition of a flip-flop as a mono-stable device (using an RC time-constant) rather than bi-stable.

[Bob Eager]

Post by Douglas Miller

Post by Douglas Miller
Reading Nelson's patent, he mentions J and K inputs very casually, as
if everyone already knows what they are. Also, that patent is for a
high-speed printing system and not the invention of a type of
flip-flop. I see no mention of prior-art there referencing the JK
flip-flop, either. It is entirely possible that the JK flip-flop never
received a patent, for reasons such as it was widely known/used by the
time anyone thought to patent it or it originated in an environment
(e.g. academia) where patents were not sought. I have not found a
patent for the implementation of a JK flip-flop prior to 1953 (i.e.
using vacuum tubes), although various such patents exist for IC
(transistor) implementations in the 60's and 70's. I'm inclined to
think that the JK flip-flop was an academic concept and that a meaning
for "J" and "K" would have to come from the original documentation for
that concept.

The other aspect is that we have no reason to believe that "J", "K", or
even "Q", have any correlation to the English language. There was a lot
of computer science coming out for Germany, for example, and so the
letters might correspond to German words/terms - or some other language
entirely.

See also the original terminology for 'wait' and 'signal' in the context
of Dijkstra semaphores.

'wait' was 'P', and 'signal was 'V'. Possibly these were for the Dutch
words 'verhogen' (increase), and 'prolaag' (try to reduce) although
there is some disagreement about these.

[Douglas Miller]

Post by Bob Eager
See also the original terminology for 'wait' and 'signal' in the context
of Dijkstra semaphores.
'wait' was 'P', and 'signal was 'V'. Possibly these were for the Dutch
words 'verhogen' (increase), and 'prolaag' (try to reduce) although
there is some disagreement about these.
--
Using UNIX since v6 (1975)...
http://www.mirrorservice.org

And don't forget "daemon" (German spelling for the English word "demon") which was coined by German computer scientist(s). At least in that case the "d" in "crond" still makes sense.

[John Levine]

Post by Douglas Miller
And don't forget "daemon" (German spelling for the English word "demon") which was coined by German computer scientist(s). At least in that case the
"d" in "crond" still makes sense.

Nope. It's a British spelling and became popular because CTSS had six
letter filenames, so daemon was more fun than demon and reminded them
of Maxwell's demon.

https://ei.cs.vt.edu/~history/Daemon.html

[Douglas Miller]

Post by John Levine
Nope. It's a British spelling and became popular because CTSS had six
letter filenames, so daemon was more fun than demon and reminded them
of Maxwell's demon.
https://ei.cs.vt.edu/~history/Daemon.html
--
Regards,
Please consider the environment before reading this e-mail. https://jl.ly

That'll teach me to trust a comp-sci prof.

[Rich Alderson]

Post by John Levine

Post by Douglas Miller
And don't forget "daemon" (German spelling for the English word "demon")
which was coined by German computer scientist(s). At least in that case the
"d" in "crond" still makes sense.

Nope. It's a British spelling and became popular because CTSS had six
letter filenames, so daemon was more fun than demon and reminded them
of Maxwell's demon.
https://ei.cs.vt.edu/~history/Daemon.html

The British spelling is of course taken from the Latin _daemon_, which in turn
is a nativized borrowing from Greek _daimo:n_ "(minor) god". (The "minor" is
in parens because even the major gods can be referred to as "daimones" in e.g.
Homer.)

[Carlos E.R.]

Post by Rich Alderson

Post by John Levine

Post by Douglas Miller
And don't forget "daemon" (German spelling for the English word "demon")
which was coined by German computer scientist(s). At least in that case the
"d" in "crond" still makes sense.

Nope. It's a British spelling and became popular because CTSS had six
letter filenames, so daemon was more fun than demon and reminded them
of Maxwell's demon.
https://ei.cs.vt.edu/~history/Daemon.html

The British spelling is of course taken from the Latin _daemon_, which in turn
is a nativized borrowing from Greek _daimo:n_ "(minor) god". (The "minor" is
in parens because even the major gods can be referred to as "daimones" in e.g.
Homer.)

And a daemon is not the same as a demon (has a negative context).