Talk:Gray code

Can Baudot's use of reflected binary codes be explained, or even verified? What I find in sources don't show any Gray-like code, nor how we might have used them. Dicklyon (talk) 01:33, 19 December 2020 (UTC)

I mean, I can see that if you sort his codewords in Gray-code order, the vowels come out first, in alphabetical order, then a few things and the consonants in order. Presumably there's a reason for that. But what? I can't see a relationship between how Baudot codes were used and the properties of Gray codes. And I can't find a source that mentions it. Anyone? Dicklyon (talk) 03:16, 21 December 2020 (UTC)

: (edit-conflict) Hi Dick, I have meanwhile added many references describing this in better details. I'm still trying to dig deeper in the history to find answers to a few of my own open questions, but regarding your question, what I found in the sources so far is that this particular arrangement was chosen to make it easier for the operator to memorize the patterns, and possibly also to make it easy to enter the chords. The synchronous Baudot telegraph used a chorded keyboard and the operator's input had to be manually kept in sync with the machine while keying in the chords, so this was very timing-sensitive.

: One source also mentioned that the codes were arranged in order of frequency, but this is wrong (the later Murray code was arranged this way, but not the 5-level Baudot code).

: --Matthiaspaul (talk) 23:46, 21 December 2020 (UTC)

::What sources say such things? My impression was that it was more about the scanning machine's teeth, nothing to do with the operator. Dicklyon (talk) 04:25, 22 December 2020 (UTC)

I trimmed that section down. I couldn't find anything in sources to suggest the Mimault's telegraph used a Gray-like code and there was a bunch of text and excess refs unrelated to Gray code. Baudot and his code have their own articles. Dicklyon (talk) 23:36, 21 December 2020 (UTC)

{{ping|Matthiaspaul}} Could you be so kind as to explain how the material you added back helps understand how the Gray code was used? Can you explain how it was used and why it mattered? Is there a sourced explanation we can incorporate, or just the somewhat cryptic French one? Dicklyon (talk) 20:18, 22 December 2020 (UTC)

{{ping||Matthiaspaul}} That section on telegraphy remains cryptic and uninterpretable, and has now been bloated up with fancy tables that don't help at all. Why/how do the properties of a Gray code become relevant in that context? I'd delete the section if no relevance can be shown. Dicklyon (talk) 05:33, 29 December 2020 (UTC)

:Huh? The section is top relevant here per the sources because it documents the usage of what we now call Gray code or reflected binary code long before Stibitz and Gray, and even by two people independent of each other, Schäffler and Baudot (both in the telegraphy business). Schäffler's usage can be traced down to a telegraph he produced in 1874 (and Lambert claimed to have shown him this code in 1872). Baudot's usage can be traced down to a telegraph he built in 1875/1876 (many sources attribute this to his 1874 patent, but the prototype documented there still used a 6-level rather than a 5-level code). Baudot's research on this started in 1872 as well. (Mimault - at his time unsuccessfully - claimed priority on some aspects of Baudot's telegraph, including the code, so this would deserve at least being mentioned for NPOV.)

: The 1872 date is relevant because it coincidentally matches the date when Gros described his "baguenodier".

: The Schäffler and Baudot code tables clearly show that they actually used codes very similar to Gray's. Some of the modern sources (even top RS ones) actually call them "Gray codes".

: Like you I want to find answers as to why they used these codes because this is historically interesting, but the question if this belongs here or not is already answered by the fact that they used these codes, not why.

: It is relevant to describe the codes as fixed-length 5-level codes - ideally, we would avoid the term 5-bit, as some authors do, because this is long before Shannon's introduction of bits, and even the idea of binary codes was new and terminology non-established (that's why some of the historical descriptions are what you call "cryptic" - they weren't in the context of their times).

: As I mentioned already, the Baudot multiplexing telegraph was still a synchronous telegraph and it used a chorded keyboard. The operator's input had to be manually kept in sync with the machine while keying in the chords, which was very timing-sensitive.

: These timing constraints could have been one of the reason(s) for why the code was arranged the way it was.

: What I also found in the sources is that the code was arranged to be easy to type and remember for the operator. I don't know if this was the primary goal or a by-product of the timing constraints. Either case, it certainly contributed to reducing the error rate and increasing the speed an operator was able to key in the chords while keeping in sync with the machine.

: --Matthiaspaul (talk) 14:40, 30 December 2020 (UTC)

::I think it is wild speculation to associate the choice of a Gray-like code with the telegraph being multiplexed, or synchronous, or having a chorded keyboard. If anything, sources suggest maybe some internal scanning order of matching the inputs, which is itself unrelated to the code-letter ordering. Different tables use different codes, sometimes Gray-like and sometimes not. So I think it best to say that some sources have recognized Gray-like codes in some of Baudots machines, rather than to put all the stuff that is only speculatively related. Dicklyon (talk) 05:37, 6 January 2021 (UTC)

::And what is the point of the "Plan of 5-level signals" table? And what is the point of the Schäffler table that doesn't even associate the codes with letters or anything meaningful? Are they just there as pretty pictures, or is there something we can learn from them relevant to Gray codes? Dicklyon (talk) 05:42, 6 January 2021 (UTC)

::From the [https://www.computer.org/csdl/pds/api/csdl/proceedings/download-article/12OmNzZEAtH/pdf Zemanek ref], it seems clear that, in Schäffler's case at least, the reflected binary code was part of the printer's internal scanning order; there's no necessary connection from there to the ordering of characters in the printer or the assignment of codes to characters, except that they have to be consistent. This makes good sense. Saying that Baudot used reflected binary in his code makes much less sense; did he have a printer with characters in that order, and so decided to assign the codes that way? And who first observed that Baudot used a reflected binary code? I've ordered a copy of the Knuth volume that has this, but that seems to be 2005, and it came into our article in 2002, so I still wonder from where. Dicklyon (talk) 05:44, 24 January 2021 (UTC)

::And the [https://gallica.bnf.fr/ark:/12148/bpt6k5574340x/f11.item Moncel ref] goes into detail on Baudot's "combinateur" which lays out the symbols with Gray code, to control the scanning recognition for printing, as in Schäffler's machine. Too bad it's in French; can anyone translate the juicy bits for us? I'm pretty sure it's still just an internal detail, not related in any significant way to what codes go with what letters. More about printing than about telegraphy, really. Dicklyon (talk) 06:48, 24 January 2021 (UTC)

I OCR'd, corrected, and had translated the Moncel ref. Lots of interesting details in there, including a section on the alphabet, but no clue about any Gray-code-like properties. The key bit where it could have been mentioned is here:

The characters of the type wheel do not follow each other on this

wheel in alphabetical order, but in the following order:

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

A É E I O U Y B C D F G H J ...

If you look at the codes corresponding to this order, they are the reflected-binary codes. The reason for choosing such a code is that they are laid out consecutively in that order on a wheel with 5 contacts, and they don't want more than one transition, potentially causing a glitch, in going from one position to the next as the wheel turns. Some of the other sources imply that, but this one really doesn't. I'm going to remove it.

My point is, the Gray code is not about the assignment of codes to letters. It's about the internal ordering of the characters on the printing wheel. More to do with printing than with telegraphy, and nothing to do with multiplex or with the keyboard; synchronous, yes, if one character is printed per wheel revolution and it turns at a constant speed.

Feel free to hat/hide this if you know how: Dicklyon (talk) 00:03, 25 January 2021 (UTC)

{{pre |

Multiple transmission printing telegraph systems and

elementary signal combinations

by M. Th. du MONCEL.

(Continuation and end.)

M. Baudot's system.

We have seen by what series of combinations and

reasoning Mr. Mimault had been led to his system

telegraphic printer, who was first (in 1874) electro-chemical

and 5-wire, then electromagnetic and one-wire, by its application to

the Hughes device and its combination with the Meyer system. Point

The start of Mr. Baudot's system was different. At the time he has

was designed, there was much concern about the telegraph at

multiple transmissions from Mr. Meyer who had given excellent

results, and M. Baudot sought whether there would be no way of applying

the principle of this system at Hughes' printing telegraph, in

used for some time on the main lines of Europe.

But this idea was difficult to realize, precisely because of

the unequal spacing of the prints which could vary from 1

time up to 28 times, without there being any way to regularize it,

since the type wheel in this telegraph runs at a

perfectly uniform way. It is certain that if by some means

mechanical we had been able to carry out

impressions after the same space of time and to ensure that

the letter Z, for example, which is the last of the alphabet, could happen

in front of the printing mechanism as fast as the letter B, we would have

been able to devote a determined time to this function which would have been

still the same, and the signal preparation time could have

be used for other transmissions made by other devices,

as in the Meyer system. But this problem that had

preoccupied as early as 1848 Mr. Highton, though difficult to resolve, had

did not frighten M. Baudot, because in 1872 he had combined a system

telegraph in which functions of this kind were

obtained. He had in fact succeeded by means of such a printer

four-wheeled types, progressively advancing relative to each other

to the other, to obtain the printing in Roman characters of the different

letters transmitted according to the Morse vocabulary. In this system,

current emissions corresponding to lines moved

longitudinally the wheels on their axis, so as to bring either

one or the other of these wheels above paper, and the emissions

which corresponded to points, rotated this axis,

different quantities depending on whether one or more

the other of the wheels was above the paper.

By modifying this system a little, M. Baudot was quick to make it

better able to meet the demands of the problem it had posed, in

reducing to one the wheels of the types and reacting on its

motor axis three electromagnets which, by means of three wheels

ratchet of different diameters, could turn it one

greater or lesser quantity.

The action of these three electromagnets depended on a kind of

rheotome governed by two electromagnets interposed in the

line circuit and reacting, one to place this or that of the three

first electromagnets in the circuit of a fairly strong local battery

to determine the rotation of the corresponding ratchet wheel, the other

to close this circuit under the influence of an inversion of the current of

line succeeding the first programs produced. However for

obtain the stop of the type wheel when passing the

character designated before the printing mechanism, it was necessary that

movements of the three ratchet wheels which controlled the

walk were exact multiples of each other, and that these

multiple were such that the combined and repeated movements of

these wheels could make the wheel of the types take the 28 positions

necessary for printing the different characters of the alphabet.

However, this result could be obtained in a fairly simple way by

arranging these wheels so that, for a single action produced

by the 3 electromagnets, their movement was in the ratio of

numbers 1, 3, 9; because by producing two successive emissions of the

line current, each wheel could increase the stroke of the

wheel of types from single to double, and one could obtain by

combination of these 6 movements 26 different positions of this

wheel, which could suffice for the immediate printing of

alphabetic characters. Nevertheless like movements too

extended from the wheel of current types and reversals

in unequal numbers were to cause some inconvenience, Mr.

Baudot preferred to increase the number of original broadcasts of the

current as well as that of the electromagnets called to react on the

different ratchet wheels, and by bringing this number to 6, it was found

leads to arranging them in such a way as to provide movements

proportional to the numbers 1, 2, 4, 8, 16, 32, which allowed him

to obtain 63 combinations without using each time more than one

current reversal. Still he thought to delete this one by

subjecting the rheotomes, at both stations, to a movement

synchronic. He then reduced the number of electromagnets to 5,

rightly thinking that the 31 combinations they could provide

were quite sufficient.

At the time when M. Baudot dealt with the provision we have just come

to exhibit, that is to say in 1873, his apparatus did not yet resolve

the problem we talked about at the start. It was, like

those of MM. Highton, Mimault, Whitehouse, a telegraph at

independent impression which could not theoretically present

advantages that because a character, to be printed, had

no need to wait for all those placed before him in order

alphabetical had passed. There was still a long way to go to the application

from the multiple system to the Hughes, and moreover these movements

progressive wheel types could result in large

implementation difficulties. It is By seeking an intermediary less

delicate in its functions and especially less complicated between the wheel

types and electromagnets called to designate the signals that

M. Baudot was led to the ingenious device to which he gave the name

of combiner , and which enabled it, by making it a device

waiting for transmitted signals, to make use of the systems

telegraphic synchronous motion printers, and to use

to other transmissions the time intervals which could

exist between the formation of signals on this waiting device and

their impression. It was in 1875 that this important invention was

patented, and it constitutes, by its very object, a very

marked between the system of M. Baudot. and those of MM. Highton,

Whitehouse and Mimault who had preceded.

As for the use of electro-mechanical functions on the rise

geometric in the combiner in question, we have seen

how M. Baudot had been successively taken there; But

regardless of construction considerations that may have

put on the way, and without having recourse either to Pascal's triangle or to

the theory of algebraic combinations, it was enough for him to relate

to the well-known 5-needle Wheatstone telegraph, to find out

than with 5 signal elements combined two to two, three to three,

four to four, etc., he could get 31 likely combinations

to represent the letters of the alphabet and the most used signals

in telegraphy, and as by means of distributing apparatus placed

at both ends of the line he could make react

the currents transmitted on the electro-magnetic organs called to

providing the elementary signals, the problem of transmission

direct all alphabetic signals to the dialer are

was thus solved in a fairly simple manner, without requiring

like the 5-wire Wheatstone telegraph.

Here is now how M. Baudot realized the

advantages of this telegraphic arrangement to the point of

view of the speed of transmissions. If in a time t, we can

transmit a single signal, we can, in a double time 2t, and

by the intervention of the distributor who will have enabled a

new signal, transmit three different signals, two of which will be

isolated and one resulting from the combination. In a triple time 3t and

with a new signal element in addition supplied by the distributor, we

will be able to transmit 3 singly and 4 in combination, in all 7.

In a quadruple time 4t, the number of these different signals

can thus rise to 15, and in a fivefold time 5t, we can

choose between 31 different signals corresponding to the different

letters of the alphabet. During this time 5t, the most complicated signal

can therefore be reproduced. Now, assuming that each permutation

line wire on the distributors is done at the same time as

that from one letter to another on the printer, we could prepare on

this one the printing of such letter that one would like while the wheel

of the types would have carried out only the 5/28 of its revolution.

However, since it takes some time to prepare a signal, it

must admit that part of the revolution of this wheel is

used in this preparation, and M. Baudot paid him a quarter of

its circumference. The other three quarters therefore correspond to the 28

alphabetical signals, and if we assume that this wheel of types

does as in the Hughes two revolutions per second, each

distributor contacts corresponding to a type of the wheel in

question will have a duration represented by (0 ", 5) / (28 + 9) or 0", 0135

(0.0135 s), and this duration is more than sufficient, since, according to

experiments with the Hughes apparatus, it was recognized that the

t necessary for the transmission of a signal on a line of 500

kilometers does not exceed 0 ", 003. However, starting from this duration

0 ", 0135, we find that the distributor, running synchronously

with the wheels of the types, could perform 7 multiple transmissions

during each revolution of these wheels1), which are transmitted

multiple could therefore cause the impression of 7

letters, on 7 receivers, in half a second, i.e. 840 letters per

minute or 504 dispatches of 20 words per hour. If we increased the

speed of distributors and receivers to the point of not attributing

transmissions that a duration of 0 ", 003, the output could be

increased to over a thousand dispatches per hour. These calculations, however, do

should be considered as purely theoretical, and, in the

practice, it is hardly necessary to count on a yield

nel to the number of multiple transmissions that can be established.

Now in M. Baudot's apparatus this number does not exceed 5, and in

admitting that with the Hughes one can transmit a letter and

half a turn, with the Baudot device only one

yield increase in the ratio of 5 to 1.5, i.e. a little

more than three times. Experience has shown, moreover, that

send 300 dispatches per hour in this way on a

800 kil.

1) Each letter requiring 5 successive contacts of 0 ", 0135 or one

total duration of 0 ", 0675, each turn of the distributor carried out in 0", 5

can only activate a number of receivers represented by

the ratio of 0 ", 5 to 0", 0675. Now this number is 7.407. By separating the

series of 5 contacts by an interval equivalent to one contact, or not

could only have 6 multiple transmissions.

From this preamble, we see that the Baudot system, like

remain those of MM. Highton, Whitehouse, Mimault, features four

different kinds of devices: manipulators, receivers,

intermediate waiting devices, or combiners, and a distributor

general whose function is not only to put

successively the line in relation to each of the systems

telegraphs, but also to have a single line wire produced

effects that would determine a line of 5 threads. We are going to study

successively these various Organs; but, first, we must

say that these devices are arranged for five transmissions

multiple and that, like those of the Meyer telegraph, the different

systems which compose them are established on the same table, '

which is arranged to allow 5 employees to be

conveniently installed on its sides. For this purpose, this table carries

on each of these sides three advanced parts on which are

fixed the devices specific to each transmission, and the employees

are placed in the re-entrant parts. The middle of the table is

occupied by the driving devices, the distributor and the shaft intended for

provide movement to all receivers; we can see some, figure 9,

the layout for one of the receivers.

Manipulator . —Each of the manipulators who is represented seen by

above, figure 8, consists of a five-key keypad or for

better to say of a vertical board AB behind which are

articulated five Morse keys, three on the right, two on the left, which

are arranged one above the other so that

fingers of both hands can easily react to the levers that

finish them. These keys press, on the side opposite to the lever and by

via a spring, on a common metal rod K which

keeps them in a fixed position; they are from elsewhere

trimmed on both sides of the lever, below the lever

itself, of four spring forks F, F which each rub

on two blades, one of which is continuous and the other cut in two, this

which constitutes, for each key, a quadruple switch. This

complicated arrangement was adopted to ensure, as

in Mr. Wheatstone's rapid telegraph, that the broadcasts of

currents can be positive and negative, and that those following

to programs already produced in the same direction, can be

find performed under an electrical influence of less energy

than those produced for the first time or those which

follow reverse emissions. Figure 11 shows the

electrical arrangements of these switches and their connection mode

with the distributor, which is shown in part developed on a

flat surface to the left of the figure. But before talking about these

connections, it is important that we say a few words in the way

how the various devices are connected and how is

arranged * the distributor itself; we will therefore have to

refer to figure 9.

We have already seen that in this system all the receivers are put in

movement by the same motor shaft. This tree is in hh , and its

movement is provided by a fairly powerful clockwork mechanism

which does not need great precision, because it is, as it is

will see later, electrically adjusted with each revolution of the motor shaft.

The same is not true of a second M

placed at the other end of the table and which sets the

Machine set in D . Not only must it be regularized by means

of a vibrating blade, as in the Hughes apparatus, but the

distributor itself must still be provided with a double mechanism

corrector in order to make it work completely synchronously with

that of the corresponding station, and subject to this synchronism the

operation of the receivers it governs. To achieve this double effect,

the distributor's mobile system G carries a sort of box

of gearing V which we will discuss at the moment and by means of

which he can have his movement suspended for a time

more or less short when it is ahead of its correspondent.

On the other hand, the motor shaft hh which turns the receivers E ,

crosses the axis m 'of the distributor's mobile system, so as to

rotate concentrically with it while maintaining movement

completely independent. With this arrangement, we understand that

it suffices to adapt to this tree hh a ZZ ebonite disc fitted with a

metal contact, so that a particular wiper is carried by the

mobile system G of the distributor, can react electrically to a

brake adapted to the motor mechanism of the shaft, and slow down its

movement at each turn of it, if it happens, like that by the way

must take place since this mechanism has no moderator, that this

movement tends to take more and more speed.

We will study this device later, but to finish with the

links of different. devices between them, we must add that

each receiver B is accompanied by a combiner C , and that these

combiners are both connected to the distributors D of the two

corresponding stations and receiver mechanisms

to which they correspond. This connection is purely electrical

in the first case; but it is both mechanical and electrical

in the second, because if the electromagnets of these combiners

perform the circuit combinations that must provide the

different signals, a working mechanical system must be

agree with the wheel of the corresponding receiver types, may,

by meeting the contacts related to these combinations,

determine a local electrical action capable of operating

the printing mechanism of the receiver. We are now going

study in detail these different organs and we will start

naturally by distributors.

Distributor . - The distributor, in M. Baudot's system, is a

slightly more complicated than in the Meyer system, because it has five

parallel rows of contacts distributed around the circumference of

two ebonite drums D , d (fig. 9) of different diameters, and one

of these rows q4, arranged on a particular disc whose

circumference follows the surface of the drum, is likely to be

moved circularly to adjust the devices according to the

length of telegraph lines. The contacts of the first three

rows g4, q5, q6, arranged on the largest drum and the

following disc, are distributed for each of these rows in

six series having six contacts each, except the last one which has only

four; this is reserved for the correction which we will see later

mode of action, and the other five correspond to the five systems

telegraphs intended to provide multiple transmission. Their

contacts are consequently connected, for one of the rows, to the

manipulators, and for the other rows to combiners and

receptors of each of these systems; however, one of these

contacts, the last in each series, is connected directly to the pole

negative of the line stack and only plays a passive role, as

will see right away. The last two rows of contacts q1, q3, which

are fixed on the small drum d and which are nothing more than two

rings divided into six equal parts, are intended to connect to the

line through the distributor's trotters the contacts of the

first row and second row, depending on whether a switch is set

the disposition of each employee arranges the line for the

transmission or reception. Figure 11 shows the

development of these contacts and their mode of liaison with

different parts of the device.

The contacts of the first row of the large drum match

in series to the five manipulators and are individually connected to

each of the keys of the corresponding manipulator; so these are

transmission contacts. Those in the second row are the

receiving contacts and correspond like the first ones, by

series, to the five combiners, while being individually connected

to the five electromagnets that are part of each of these

combinators. Finally the contacts of the third row

still communicate in series, both with the electromagnets

of local combiners through the receiving contacts to which they

are connected by a U-shaped slider, which presses on both rows, and

with the manipulators, by one of the switches of the keys

we will call local switch . It is through the

contacts of this third row that dispatches are printed

at the start and that the combiners are brought back to their position

normal before they are brought into play again (see Figure 11).

Above the distributor which is fixed, except the part corresponding to

the first row of contacts, support the changeover springs,

which are seven in number. Five correspond to the five

rows of contacts we talked about, and the sixth, which is

precisely the U-spring mentioned at the moment, precedes the others,

in their walk, a distance equal to the length of one of the

contacts. These springs are attached to a VG rotating arm , fig. 9, put

moving by a hollow axis m ' depending on the mechanism

clockwork regularized M and through which passes as we have seen

the end of the horizontal shaft hh which controls the movement of

receivers. However, this movement is only communicated to this arm.

via the gearbox already discussed

and which is none other than a ratchet wheel V to which it is connected by a

strong ratchet with several teeth. This ratchet, represented in large, in 0,

figure 10, with its accessories, reacts on the opposite side on a

rocker fitted with an ankle c which, at each revolution of the arm G

carrying the springs, passes over an articulated lever Ip

the end of which is terminated by an inclined plane p. This lever is engaged

on an electromagnetic trigger i (figure 9), adapted to the armature

a of a particular electromagnet e, and this electromagnet is related

with distributor correction contacts. Now it follows from this

mechanism, a constantly renewed correction that maintains the

movements of the movable arms of the two distributors in

correspondence in a state of perfect synchronism. Indeed the

ankle position c, fig. 10, of the clutch pawl in the

two distributors is such that when the movements are

perfectly synchronized, this peg, on both devices,

at the same time arrives at the beginning of the inclined plane p of the lever

engaged; but precisely at this moment the trotters of the

distributors arrived at both stations on the contacts of

correction we talked about, and since these contacts are related

to the corrective solenoid and, they can transmit the current to

through it and release the engaged lever lp . Therefore the ratchet

clutch can pass over the inclined plane p of this lever

without disengaging the motor mechanism of the walkers. If at

on the contrary, one of the movements is faster than the other, the contact

which causes the corrective electromagnet to react is not carried out on the device

walking faster than after the wise step of the pawl c on the

lever engaged I p , and this then disengages this pawl which does not

can re-engage only after having crossed the inclined plane p ; it

naturally results in a small delay in the walking of the supporting arm

trotters, and this delay may be enough to understand the greatest

speed with which it was animated. This action is also ensured by a

second articulated lever r , which presses on the inclined plane i? and below

which engages the ankle c . Since the electromagnet e is a

Hughes electromagnet, its armature a must be put back into position and

this function is carried out at the same time as the reconnection of the

mechanism, by the action of two eccentrics b and f, fig. 9, who

react on it by means of two levers l and g, the action of

one ahead of the other a little.

As the contacts related to the second row of the

distributor cannot correspond, in position, to those of the

first row, since the effect produced cannot be achieved at the same

time of arrival and departure, and that this lack of correspondence

is more or less accentuated depending on the length of the line, it is

necessary, in order to bring these contacts to an agreement between them, to settle the

reciprocal position of the two discs which carry them, and it is for

that the first, q4 is likely to move on its axis.

This movement is carried out using a pinion wrench K, figure

9, engaged in a window adapted to the movable disc and one of which

edges, parallel to its circumference, is provided with a small

rack.

By means of this system, the trotters of the two distributors in

connections therefore pass at the same time at both stations

on the corresponding contacts of each series and can, therefore

way, successively establish the junction by the line,

different keys of each manipulator with the electromagnets

of the corresponding combiner. Only, as the action is

successive, it is necessary that these electromagnets

maintain their frame in the position made by the

current that has passed through them, so that this action by combining with

one or more others in the combiner, can provide the signal

desired. It is for this reason that we had to use

polarized armature electromagnets.

Combiner . - The combiner is composed, like the distributor,

a fixed part and a moving part and in addition to a system

electromagnetic composed of the five electromagnets of which we

have spoken, and which acts like a multiple relay system

double contacts. Figure 11 gives a representation

theoretical.

The fixed part consists of five double metal discs with

notched circular rim, arranged in such a way that the void

practiced in one of the ledges is almost filled with a

protruding part cut in the rim of the juxtaposed disc. These

two parts of each disc are isolated from each other,

such that the circumference which they form externally is

composed of parts which may be unequal in length, but which

are isolated from each other and which alternately belong to

two different discs, capable of being electrically connected

with different circuits. All of these double discs, however, are

provided at a point of their circumference, which is the same for all,

and over an arc of about 80 degrees, with a very large notch

filled with an insulating material, which leaves the device inactive for

about a quarter of a revolution of its moving part, and it is

precisely during this time that employees prepare their

signal to manipulators.

In figure 11, we assume the rings formed by these different

double disc systems, developed in a straight line, and for

distinguish from each other the parts belonging to each

disc coupled, one reached them, the ones in black, the others in

White. As these black and white parts are, by the fact, only

isolated contacts connected to the contacts of the armatures of the five

electromagnets of the electromagnetic system, we

distinguish from each other by calling black the contacts

indicated in black, and white contacts not tinted. That put us

will examine how these different series of

contacts with respect to each other.

The bottom AAA ring , etc. (fig. 11), which we will designate under the

No.5. Door, as seen, 8 black contacts and 8 white contacts

of the same length, except the last of white which is only half

others. If we assume the metal part of these disks divided

into 31 equal parts, each of the black and white contacts of this

fifth ring would correspond to 2 divisions, except the last of the

whites who would only understand one. The fourth ring does not carry

4 black contacts and 5 white contacts which each correspond to

4 divisions, except the last two which are white and do not include

that one and two divisions. They are placed in relation to the

contacts of the fifth ring, so that the contacts

black start and end in the middle of each of the black contacts

of this fifth ring. The third ring carries only two

black contacts and three white contacts, and these black contacts, like

previously, are arranged to start and end at

middle of two consecutive black contacts of the fourth ring, this

which causes that the two white contacts which are at the ends

include only 3 and 4 divisions, while the others in

include 8. The second ring has only one black contact left

and two white contacts which include the first 16 divisions, the

seconds 8 and 7 divisions, and always commits the black contact

and ends in the middle of the two black contacts of the third ring. Finally

the first ring has only a black contact and a white contact, the

the first comprising 16 visions, the last 15. The black contact

then begins at one end of the indentation and ends at

middle of the contact of the same type of the second ring.

If we carefully consider the reciprocal arrangement of these various

contacts, it is immediately recognized that, thanks to this arrangement,

five springs R1 R2 R3 R4 R5 placed in a straight line and which

would revolve around these 5 rings, can never meet

at the same time two separations of black and white contacts, and by

therefore the functions of each of them are clearly

determined to complement the closures of the local circuit at

through the printing mechanism. The various black and white contacts

of these rings are also connected by wires to the double

contacts A, B, C, D, E of the 5 electromagnets of the combiner which it has

previously discussed and which constitute what we

call the electro-magnetic rheotome. This binding is made of

such that the white contacts correspond to the contacts

under which the reinforcement rests in normal times, and

that the black contacts correspond to the upper contacts on

which support these frames when they are deflected. In

examining the position of this or that of the reinforcements a, &, c, d, e on

can easily, according to this explanation, find the open ways

through the combiner.

The electro-magnetic system is moreover nothing more than five

Siemens polarized electromagnets, whose armature oscillates between

two stops forming the previous contacts A, B, C, D, E, and

found maintained in the last position it occupied, by

result of its polarity and the remanent magnetism of the electromagnet.

These reinforcements being the switching members intended to put in

action the printing mechanism, are naturally related to this

mechanism and the local battery P, the circuit of which must be completed by

the combiner; but as they can act more or less

large number, they must, with the different rings of the

combiner, be an integral part of a continuous circuit closed by the

mobile system of the combiner and, therefore, be linked between

both of them, except the one that communicates directly to the

pile P. It is for this reason that reinforcements b and c , d and e are

metallically united as seen in the figure.

The mobile part of the combiner is composed, like that of the

distributor, of a series of 5 spring trotters R1, R2, R3, R4, R5,

suitable for an arm mounted on the axle of the type wheel and which turns

with it, and like the 31 characters on this wheel

correspond exactly to the 31 divisions according to which were

established the contacts of the combiner, these springs pass

successively before these different divisions at the same time as

the different characters of the type wheel pass in front of the

printing mechanism. Consequently, if the type wheel is

suitably placed in relation to this trotter system, we can

ensure that by the time this system reaches the tenth or the

fifteenth division of the combiner, for example, the tenth or the

fifteenth letter is placed in position to be printed.

The mobile system of the combiner being the counterpart of the system

electro-magnetic and having to complete the circuit whose path is

prepared by this last system, must have its trotters connected two to

two, like the armatures of electromagnets; only this

connection must be made in an opposite way, so that the current

transmitted circulates meandering through the five rings of the

combiner. Also it is the springs R4 and R3, R2 and R1 which are

connected together, and it is the fifth Rs that communicates with the

battery P via the printing electromagnet I.

With this arrangement, it is easy to see how the current of the

pile P is closed at each turn of the trotters and according to the action

determined on one or another of the electromagnets. Indeed, suppose

that the lower keys of the corresponding manipulator have

deflects, through the distributors, the reinforcements e and c

of the combiner: the current leaving the battery P will be directed by

the armature which has not moved on the white contacts of the fifth

ring of the combiner, and as to get out it must pass through a

white contact of the fourth ring, a black contact of the third, a

white contact of the second and a black contact of the first, it cannot be

find in these conditions that when the trotters will have arrived at

the twenty-fourth division; then the circuit crossed will be as follows:

armature a , 6th white contact of the 5th ring, res out R1, spring R2,

4D white contact of 4th ring, armature b, armature c deflected, 2nd

black contact of 3rd ring, spring R3, spring R4, 2nd white contact

of the second ring, armature of the armature deviated, black contact of the

first ring, R5 spring, printing electromagnet, battery. The

printing mechanism then being brought into play, prints the letter in this

moment at hand, and this letter is the twenty-fourth of the wheel of

types. We will see later that this letter is the S.

We now understand, from the functions that we come from

to analyze, which will be possible by the different combination of

positions of the electromagnetic rheostome armatures,

combination carried out under the influence of the manipulators and by

through the distributors, not to obtain the closure of the current

local printer that at the very moment when the letter of the

types, designated by this combination, arrives in front of the mechanism

printer.

M. Baudot imagined still other simpler combinators in

their construction which have the advantage of being able to operate

mechanically the printing mechanism, and consequently without

local current. In these combiners, the fixed part of the device is

mobile, and reciprocally the mobile part constituted by the springs

walkers is fixed. These springs are in fact replaced by

species of articulated rockers which carry fixed normally close

of their axis of the arms pressing on a lever depending on the system

first impression. Five Hughes electromagnets, in connection with the

distributor, are placed in front of one of the ends of these rockers

so that their frame, when detached, can tilt them and

consequently release their arm from the printing lever. Above

the opposite end of these rockers, is the mechanism

combinator proper which is arranged much like the one

that we have studied previously, but which, instead of contacts

different in nature, has alternately hollow parts and

protruding arranged, moreover, like these contacts. This cylinder,

as we said at the beginning, turns with the wheel of

types of the printer, and, in this movement, provokes

naturally the lowering of the rockers that the protruding parts

meet; so that when the turn of this cylinder has been

accomplished, all these rockers had to be lowered, either

mechanically by the combiner, or electrically by the

electromagnets. Then the printing mechanism is released and can

produce the impression, but this impression can be done more or

sooner depending on the position and number of lowered scales

electrically, because the combiner only completes the action thus

produced, and this complement is only carried out when the

position of this combiner corresponds to the arrival of the letter

transmitted in front of the printing mechanism. This one is loaded

then, after printing the letter, take care to re-enter all

the scales and put all the reinforcements back at the same time

deviated from the electromagnets in contact with them.

This system, as is easily understood, could still be

electrically combined. It would suffice for this to keep at 5

combinator electromagnets the arrangement we have

studied in the first place, and to consider the rockers from which it comes

to be a question of scull switches oscillating between two

contacts and with an idle contact. By connecting these double contacts

to those of electromagnets, and by metallic

flip-flops two by two in an inverse manner to that of the reinforcements

of these, one of the switch systems can serve as

complement to the other, and the combinator cylinder by carrying out

required this complement, determines the impression by launching the

local current through the printing electromagnet. We win at this

system the elimination of the 5 trotter springs, and the construction of

Combiner cylinder is much simpler, since there is no longer any

isolated contacts or double discs. M. Baudot now gives

preference for these two systems; but as it is the first who

has been executed so far, we had to stop there longer.

Receiver . - The receptors in this system look like

much to the part of the Hughes Telegraph which constitutes the

printing mechanism; a type T wheel, figures 9 and 12, whose

characters occupy only three quarters of the circumference; a

printing wheel 0 provided with 32 pointed teeth in the part of its

circumference corresponding to the types of the preceding wheel and which

is mounted on the same axle of this wheel; a mechanism for

permutation of numbers and letters; a printing system I, J, x, x

put into action under the influence of a trigger

electromagnetic; such are the various parts which compose it.

This printing system, however, does not work as in

the Hughes apparatus; the axis with the four cams not being there,

printing is done under the influence of the motor which sets in motion

the types wheel and combiner, and through the wheel

of 32 teeth 0 which was discussed previously. This indeed has

for function, when the J armature of the electromagnet is

detached, to lead an arm Ha; fixed on the articulation axis of this

frame, which is currently within reach of its teeth; and

as this one is equipped with a system of rollers NH ## on which

the paper strip is rolled up, this strip can be pressed

against the T-type wheel. This roller system consists of the

remainder of two small guide cylinders xx around which the

strip of paper, and an NH rolling mill system, one of the

cylinders, mounted on the axis of the frame itself, carries the snap

PP 'intended to advance the paper. It's easy to understand,

moreover, that the wheel of 32 teeth O which thus governs the impression

can, being provided with a permutation mechanism similar to that which

is suitable for the correcting wheel of Hughes devices, determine

printing letters or numbers when the arm H x ; wearing the

rollers meet, between the teeth of this wheel, the appendix

system which activates this mechanism.

To obtain that after each printing the reinforcement of

the electromagnet is mechanically replaced in contact with its

poles, Mr. Baudot establishes on the support of the mechanism a rocker with

spring L which, being met by a peg I adapted to the wheel

32 teeth O, can be tilted far enough back when passing through

the indented part of this wheel, to make the arm travel

impression H #, in the opposite direction of its first movement, the arc

circle he had described under the influence of the trigger

electro-magnetic and the drive produced by the O wheel.

The result is that the armature J is again brought into contact with

the electromagnet I, and therefore able to provide

new action.

Linking devices to each other. - Now that we have described

the way in which the organs of the manipulators of the

distributors, combiners and receivers, we will

to be able to study more easily their mode of connection, and we

let's start with the manipulators first.

We have seen that these devices were each equipped with four

switches having the form shown, fig. 13, where only two

are figured. These switches each consist of a spring

inverted U rubbing on three contacts, one which is long and which

corresponds more or less directly to the distributor's contacts,

the other two which are short and which also correspond more

or less directly to both poles of the line stack. In fig.

11, which represents all the connections of the devices, these four

switches are indicated only by their contacts, and it is necessary

to admit consequently that there exist above them the trotters in

U of which we have just spoken, which bring together in long contact,

depending on whether the button is raised or lowered, the upper contact

or the bottom contact. In this figure, only the

switches related to three of the keys of a manipulator,

the connections being always the same for the other keys. Over there

same reason, only part of the distributor's contacts have been shown,

and these contacts are shown on the left at the top of the

figure. The trotter springs of this distributor are indicated in r, r1

r2, r3, r4, and the direction of their movement as well as that of the springs

R1, R2, R3, R4, R5 of the combiner is indicated by arrows. The

+ and - signs indicate that the contacts to which they belong

are placed in direct contact with the two poles of the line stack, and

these same signs surmounted by the letter R indicate that a

resistance has been introduced through the communication wires of this

battery to reduce the voltage.

From the inspection of the figure, we first see that the first

switches of each key are set by their long contacts in

report with the plates of the local distributor, and only receive the

current, except that of the first key, only through the

second and third switches of the preceding key, which

communicate to it, depending on whether the transmissions are made with

currents succeeding each other in the same direction or in opposite directions,

more or less strong electric charges. It is precisely these

variable loads that keep the line at the same

potential and realize the benefits Mr. Wheatstone has achieved

in his fast telegraph with the compensating currents.

Let us indeed follow the course of the currents in a transmission made

using keys 3 and 4 down. Unweakened positive current will be

first directed to the third distributor contact; because the line is

already under the influence of a negative charge that it still has

in normal times, and this current transmitted by the third key

comes through the third switch of the second

key not lowered. Immediately afterwards, a new positive current is

sent to the distributor's fourth contact by the fourth key

lowered but it is weakened, because it does not reach the first

switch of this key only through the third

switch of the third key which is then lowered and whose

second contact is related to the weakened pole of the battery. As

by the time this current crosses the line, it is already loaded

positively, it therefore only needs a weak positive charge to

take back the potential it must have to function

regularly. If instead of lowering keys 3 and 4, we had lowered

keys 2 and 4, it would not have been the same: a first current

positive non-weakened would have been transmitted by the second touch to the

second distributor contact in the same way as that

previously transmitted by the third key, but the one that would have

transmitted the fourth touch would not have been weakened, because the third

key not having been lowered, the current would have arrived at the first

switch of the fourth key by the first contact of the

third switch of the third key, and this current not

weakened would have been essential to reverse the load

weakened negative that the line would have acquired under the influence of this

third key not lowered.

It remains for us to discuss the functions of the fourth switch of

each key, functions that are double, because this switch is used

both for local impressions and recall to their position

normal of the electromagnetic armatures of the combiner. As

for the others, the two contacts of this switch are connected to the

two poles of a battery; but this stack is a local stack, and each

long contacts of these switches is connected to a contact of the

third row of distributor. Normally, this long contact

being connected to that of the switch contacts related to the

negative pole of the local battery, it happens that when the manipulator does not

not working, all the contacts of the third row of the

distributor are negatively charged, and therefore when the

small spring, in U of the distributor (the one that precedes the others) comes to

pass over these contacts, it successively transmits this load to the

receiving contacts who, transmitting it in turn to

electromagnets of the local combiner, through them determine the

closure of five negative currents. Now these negative currents

then recall to their normal position those of the reinforcements of these

electromagnets which would have been deflected in the previous turn of the

distributor, and as the action of the U-shaped trotter precedes that of the

other wipers, the combiner is placed in position to provide

new combinations before the passage of these. Of a

on the other hand, and for the same reason, when the manipulator is put in

game, the distributor contacts in relation to the keys

lowered are positively charged and operate the

electromagnets of the same local combiner, which therefore determines

printing of the dispatch at the outgoing post and before it is

transmitted to the receiving station. Under these conditions, only one

switch may be sufficient, because the circuit, being local, is not subjected to

effects of load variations which influence both transmissions

across the lines.

Alphabetical system . - Figure 14 below shows the

alphabetical system adopted by M. Baudot. The different signals

that can be done with the right and left keys are

indicated by small circles placed in squares, and these signals

being arranged like the numbers to be combined in a table of

multiplication, we can see, by following the leagues horizontally and

vertically, what is the letter designated by each combination

signals. This table is double to match the two

wheel positions of types which provide printing of letters

and that of numbers. This is how we see that the letter

corresponding to a simple lowering of key N ° 1 of the

right manipulator is' A, that lowering the first

right key and the first key on the left give the

J, that the three keys on the right and the two keys on the left

lowered give the P, that the isolated lowering of the two keys of

left gives the white letters or the numbers, etc. We

must however note that the order of the keys on these

tables must be interpreted, in relation to that indicated on the

fig. 11, as if the two keys on the left represented the

keys 1 and 2, and as if the three keys on the right represent

keys 5, 4 and 3, keys 2 and 5 being indicated plots

which correspond to indexes. If we follow on the combiner the

numbers of the divisions to which the various

combinations indicated in these tables, it is recognized that the letters

that they designate do not correspond to their rank in the order

alphabetical. This is due to what M. Baudot wanted, as in

the Morse alphabet, apply the simplest combinations to

most frequently repeated letters in dispatches. The

characters of the type wheel do not follow each other on this

wheel in alphabetical order, but in the following order:

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

A É EIOUYBCDFGHJ White numbers

17 18 19 20 21 22 23 24 25 26 27 28 29 30 31

KLMNPQRSTVWXZ t Blac letters

Operation of devices . - Now it's time to see

how all these devices work, and we'll assume that

it is the third manipulator of station A which is put in

action to transmit the letter H to station B. The employee of

station A will then lower the first two buttons of the

right manipulator and the second from the left manipulator.

Depending on the arrangement of the devices, these lowered keys will be

those we have designated in fig. 11 under the numbers 2,

4, 5. This reduction will be carried out during the passage of the resorts

walkers in front of the insulating part of the distributors that we

suppose to walk synchronously. When these springs

will reach the third series of contacts of these distributors, the

line will be put in contact, by contacts N ° 2 of this one with the

reinforced positive pole stinks a line and that by key N ° 2. The

positive current arriving through the second contact of the distributor of the

station B to the second electromagnet of the third combiner, will

tilt its armature d on the contact in relation to the black contact

of the second ring of the combiner. Almost at the same time the

keys 4 and 5 of the manipulator will transmit through contacts 4 and 5

distributors of equally positive currents that will cross the

electromagnets 4 and 5, and will bear their armatures a and b on the

contacts corresponding to the black contacts of rings 4 and 5 of the

combiner. However, the positive charges thus transmitted do not

will not be the same, because button 3 is not lowered, the

positive charge which will be transmitted to button 4 will be reinforced,

while it will be weakened for key 5 due to

lowering of button 4 which preceded it. So we will find ourselves

in the case of good transmission, and the three reinforcements a , b , d

deviated will open to the local current, when the trotters of the

combinator will come to pass, the following way: deflected armature a ,

fourth black contact of the fifth ring of the combiner, spring

R1, spring R2, second black contact of the fourth ring,

deflected armature b , armature c , second white contact of the third

ring, spring R3, spring R4, black contact of the second ring,

deflected armature d , armature e , white contact of the first ring,

spring R5, battery. But that will only be when the trotters are

arrived in front of the thirteenth division that this current will be completed. Gold

this thirteenth division corresponds precisely to the letter H.

It is easy to understand that the same effects being reproduced on

electromagnets of the local combiner of station A which transmits, the

letter H will be found in the same way printed under the influence of

fourth switch of the three down keys.

Ultimately, we see that, by this system, all letters of

the alphabet and numbers can be printed under the influence of

five keys that are held constantly under the fingers,

and that we lower in such or such order as is appropriate to re

present the 31 letters and signs of the alphabet. Without doubt this

impression is not made instantly at the time of

transmission, but the time separating successive impressions

is regularized, and can be used for other transmissions, which

are carried out successively in the same order and which

quintuple the number of dispatches sent and received.

This device was built with great skill by Mr.

Dumoulin-Froment, the son-in-law and successor of the illustrious builder

M. Froment, and as I have already said, the first tests were very

satisfactory. It is hoped that this system can be

advantageously applied in practice.

Postscript . —As a result of an error by the copyist certain sentences

from the previous article that had been erased in pencil on my

manuscript, have been reproduced and suggest that the first

M. Mimault's system was likely to apply to the

multiple transmission; but as we could see by the last one

paragraph of this article and the preliminary account of the system, it does not

is not so. This system could have no other result than

to print directly and independently of each other the

different alphabetic characters, as did the rest

Mr. Highton's device. Multiple transmission did not have

moreover its raison d'être, under these conditions, since there was then

no time wasted in transmissions.

}}

{{cite journal |title=La télégraphie et quelques autres applications de l'électricité à l'Exposition universelle de 1878: Appareils duplex, quadruplex et multiplex |series=La télégraphie et quelques autres applications de l'électricité à l'Exposition universelle de 1878 |language=fr |author-first=Timotheus |author-last=Rothen |date=1878-12-25 |journal=Journal Télégraphique |publisher=Le Bureau International des Administrations Télégraphiques |publication-place=Berne, Switzerland |volume=IV / #10 |number=12 |id=ark:/12148/bpt6k5661719x |issn=2223-1420 |eissn=2725-738X |pages=247–254 [252–254] |url=https://gallica.bnf.fr/ark:/12148/bpt6k5661719x/f12.item |access-date=2020-12-19 |url-status=live |archive-url=https://web.archive.org/web/20201219004551/https://gallica.bnf.fr/ark:/12148/bpt6k5661719x/f12.item |archive-date=2020-12-19 |quote-pages=252–253}}

{{quote |

If we look at the pI disc, we find in divisions 2 to 5, 10 to 15, 22 to 25 and 30, in all four notches, in the pII disc 7, in the pIII disc 9, in the pIV disc 10 and in the pV 8 disc, for a total of 38 notches. […] By notching the discs according to the drawing in figure a, we would have obtained 38 jolts for the disc levers, during a single rotation of the latter. The regular functioning of the apparatus would perhaps not have been hampered by these 38 jolts, but, in any case, they would not have been favorable to its functioning, since the levers of the discs are intended to establish the contacts. of the local current of the printer relay. […] M. Schäffler was therefore led to seek a more advantageous solution in the displacement of the notches, so as to obtain a more suitable series of divisions. […] He solved this problem by empire. […] Figure b was formed using 31 small pieces of wood, which could be moved at will. Wood No 1 was set aside, Mr. Schäffler using only 30 permutations. […] He moved the antlers until he came to figure b. This is how we find wood 8 next to wood 21 and so on. This arrangement was the most favorable and the notches of the discs followed each other in such a way that the lever of the disc pI fell once, pII 1 time, pIII 2 times, pIV 4 times and pV 7 times, in all, all the 5 levers 15 times, in a notch, instead of 38 times as in the first arrangement. […] The only purpose of this arrangement is therefore to free Mr. Schäffler's device from a few drawbacks. [...] If now M. Baudot's model resembles M. Schäffler's permutation disks, we can simply conclude that M. Baudot enjoyed the same advantages as M. Schäffler. […] However, the two systems cannot be absolutely equal because Mr. Baudot uses 31 permutations, while Mr. Schäffler is satisfied with 30. […] In general, the two devices are only alike in idea apply the multiplex system to printing devices. 

}}

The "notch" I presume is between positive and negative contact regions on the disc, corresponding to bit transitions between codes. Minimizing them is good, and is equivalent to having only one bit transition per code transition. But this guy misses the point. It's not about minimizing the number of jolts but about avoiding the possibility of glitching, when the printing wheel scans for a match to the character code. He got this close to being able to say something about the reflected binary code and it's raison d'être, but flubbed it. I already removed this ref from the article, since it has nothing relevant to Gray code.

Let's look at how the article's comments on Baudot got to where they are.

• In [https://en.wikipedia.org/w/index.php?title=Gray_code&diff=360380&oldid=360378 this 2002 edit], we got "The French engineer Émile Baudot used Gray codes in telegraphy in 1878. He received the French Legion of Honor medal for his work.", unsourced, from User:Heron. I presume he got that from a source, but don't know what.
• In June 2014, at Talk:Gray_code/Archive_1#Baudot code, one of the first use of Gray code, an IP proposed saying more about Baudot.
• On July 3, 2014, the IP [https://en.wikipedia.org/w/index.php?title=Gray_code&diff=615367627&oldid=614187185 added a ref] to Pickover's [http://books.google.com/books?id=JrslMKTgSZwC&pg=PA392#v=onepage&q&f=falseThe Math Book], of 2009, which says "The French engineer Émile Baudot used Gray codes in telegraphy in 1878", quoting our article without attribution. And it has a direct copy, plus color, of the patent drawing that I upload in 2006. Seems like clear WP:CITOGENESIS to me.
• On July 4, 2014, User:Glrx pushed back on the talk discussion and asked for a reliable source, but didn't do anything about the article.
• That's where it sat until December 17, 2020, User:Matthiaspaul [https://en.wikipedia.org/w/index.php?title=Gray_code&diff=prev&oldid=994696206 started adding] a whole bunch of refs about Baudot, most not saying anything in support of him using a Gray code, as far as I can find.

What we really need are secondary sources that connect these telegraphy bits to Gray codes. I see Knuth does that, so I'm getting a copy to inspect in depth. Dicklyon (talk) 03:01, 24 January 2021 (UTC)

Based on discussion above, and more studying of sources, I've pared it back again. It's clear that both Baudot and Schäffler had discovered and used the essential properties of Gray codes in their printing mechanisms, so it's best to focus on sources that say something about that. Multiplexing and keyboard differences are irrelevant, and assignment of codes to letters nearly so. Dicklyon (talk) 00:40, 25 January 2021 (UTC)

:I needn't have waited for the Knuth book, as I see I had found it before and linked it above ([https://books.google.com/books?id=IkuEBAAAQBAJ&pg=PT442&dq=stibitz+gray+kode&hl=en&newbks=1&newbks_redir=0&sa=X&ved=2ahUKEwix28X-0s3pAhWkHDQIHacJBfEQ6AEwA3oECAQQAg#v=onepage&q=stibitz%20gray%20kode&f=false here]). It says "More significantly, Γ5 was used in a telegraph machine demonstrated in 1878 by Émile Baudot, after whom the term 'baud' was later named. At about the same time, a similar but less systematic code for telegraphy was independently devised by Otto Schäffler." That about it: "used in a telegraph machine" is supported by the sources, but the Gray code is still not very relevant to the Baudot code, or Schäffler's code, itself. It's an internal detail of the sequential character matching at the print wheel. Not sure why he says "code for telegraphy" in Schäffler's case. Dicklyon (talk) 00:07, 3 February 2021 (UTC)

An IP editor [https://en.wikipedia.org/w/index.php?title=Gray_code&diff=1024554000&oldid=1023725478 claims] the expressions are different from what's given in the ref. ~Kvng (talk) 13:18, 25 May 2021 (UTC)

:The number of transition in each dimension is necessarily even. The IP's claim look therefore plausible, while the current claim in the Wikipedia article must be wrong. --FvdP (talk) 15:42, 17 December 2021 (UTC)

I'm not sure why the table of values at the very top of the article also includes a different coding scheme which is not explained anywhere else nor has an article of its own. It makes the table more difficult to read while not adding anything that's related to the article, I suggest it's best removed Ruse.mp (talk) 07:58, 29 December 2022 (UTC)