Talk:Baudot code

From Wikipedia, the free encyclopedia
Jump to: navigation, search
WikiProject Amateur radio (Rated Start-class, Low-importance)
WikiProject icon This article is within the scope of WikiProject Amateur radio, which collaborates on articles related to amateur radio technology, organizations, and activities. If you would like to participate, you can edit the article attached to this page, or visit the project page, where you can join the project and/or contribute to the discussion.
Start-Class article Start  This article has been rated as Start-Class on the project's quality scale.
 Low  This article has been rated as Low-importance on the project's importance scale.

Baudot keyboard/keyset?[edit]

The keyboard is barely mentioned. It should be described more fully someplace (here or in another article with a linke from here?) and there syould be a photo or diagram. See also chording keyboard and keyer. --Treekids (talk) 19:03, 31 January 2008 (UTC)

Done. I photographed a Model 15 Teletype keyboard, the classic Baudot machine. --John Nagle (talk) 06:44, 11 March 2009 (UTC)

F G H ?[edit]

From the article:

"Code points 0D, 14 and 1A are not used in telex communication."

So how did they represent F, G and H? -- The Anome 17:42, Jun 5, 2005 (UTC)

Hard copy of code missing.[edit]

I think that the paper tape storage using Baudot should be metioned also. Messages/programs can be stored in Baudot on paper/mylar tape. This is done with a machine that punches and reads punched tape. See for an example. Also see "Punched tape" in wikipedia.

Wrong Donald Murray?[edit]

From the Article:

"Around 1901 Baudot's code was modified by Donald Murray by re-ordering the characters..."

When I followed the link on the name, I was presented with a article about someone who was born January 24, 1923. Either someone has one (or more) of the dates mixed up, or the link is to a different person with a similar (same?) name.

On the gripping hand, Time travel might be in play here, but if so, please site sources. :-)

This shows the problem of creating a red link and not immediately filling in with a stub. Someone comes later and creates an article with an unintended subject for the original link. I've removed the linking and added dates. The birth year is given various as 1865 or 1866. JonHarder 00:31, 3 February 2006 (UTC)

Tranmission of Data[edit]

How did the transmitter and receiver work? Were the 5 bits transmitted using 2 different voltages or 2 different frequencies? How did the receiver know when a sequence started and how did it differentiate successive bits that are the same (timing perhaps)?

Baudot code on a DC line was simply transmitted as an on-off signal. On is called MARK and off is called SPACE. On radio links, two different frequencies were used; this is called frequency shift keying. Contrast this with the early stock ticker machines, many of which used DC polarity reversals for signals. Baudot signals were transmitted using Asynchronous communication, with a "start bit", five data bits, no parity bit, and 1.5 "stop bits". The asynchronous communication article ought to have waveforms for this, but doesn't. --John Nagle (talk) 06:50, 11 March 2009 (UTC)

Coldplay album cover[edit]

Of course we have a picture of it here: [[1]]. But how does it decode into Baudot? It looks like either 5 characters, each two rows of colored squares high, but with only 4 bits in each row! Someone has changed it from decoding to "X&Y" to "X96", which is only 3 characters. --Wtshymanski 18:04, 21 June 2007 (UTC)

Start from the bottom left corner, go up - 5 bits. 11101 = "X". The next columns give the FIG shift and then the numbers 9 and 6. Muad 03:56, 2 July 2007 (UTC)

Need the Original Code[edit]

This article starts out by explaining that what we now know as "baudot code" is not the same as Baudot's original code, then goes on to give the code points for ITA2. Seems to me Baudot's original code should be given here (in addition to ITA2). Anyone have it? Rees11 04:18, 29 September 2007 (UTC)

The reference I added today ( provides some of the details asked for above along with a dozen references, which should be very helpful to someone.
Something that should be mentioned is the smell of a room full of these clinking, clanking, mechanical beasties and the quarts of light oil it took to keep all those parts from imploding. (On the teleprinter page anyway; I'd guess the ozone level from Baudot's gear was impressive) Twang (talk) 01:47, 13 March 2008 (UTC)
The article says "US implementations of Baudot code may differ in the addition of a few characters, such as #, & on the FIGS layer. The above table represents the US TTY code." The text is out of sync with the figure. The table shown is for CCITT/ITA No. 2, not the "US TTY code", which has a "$" in the FIGS-D position. I have a WWII-vintage Model 15 TTY, with its original documentation from 1944, and it has both a $ (at FIGS-D) and a £ symbol (at FIGS-#). It matches this table, except for having '£' at FIGS-H instead of '#'. --John Nagle (talk) 17:48, 31 January 2009 (UTC)

The Vansize ref has a very nice table with Baudot's original code. It's out of copyright if anyone wants to copy it in here. I'm not sure what the rules are about copying a Google book scan. Rees11 (talk) 03:02, 27 March 2009 (UTC)

Gauss and Weber[edit]

I don't think it was appropriate to remove the reference to Gauss and Weber. Early histories are not all that hard to come by, and the ones I've checked all credit Gauss and Weber with developing the code, usually citing 1834 as the year. See for example Wire & Radio Communications, vol 34 (May 1, 1916) p. 209 "Printer Systems." At the very least the article should state that the origin is in dispute. I'm not willing to engage in an edit war but I don't think it's right to re-write history like this. Rees11 (talk) 02:37, 27 March 2009 (UTC)

Actual code[edit]

Why doesn't this article give the actual code? This was it:

A 11000 
B 10011
C 01110
D 10010
E 10000
F 10110
G 01011
H 00101
I 01100
J 11010
K 11110
L 01001
M 00111
N 00110
O 00011
P 01101
Q 11101
R 01010
S 10100
T 00001
U 11100
V 01111
W 11001
X 10111
Y 10101
Z 10001
1 11101
2 11001
3 10000
4 01010
5 00001
6 10101
7 11100
8 01100
9 00011
0 01101
& 01011
% 10110
? 10011
= 01111
+ 10001
- 11000
( 11110
) 01001
/ 10111
: 01110
, 00110
. 00111
' 10100
space            00100
shift to letters 11111
shift to figures 11011
new line 00010
line feed 01000

Source: Paul Gannon, Colossus, 454 Intelligent Mr Toad (talk) 09:44, 5 October 2009 (UTC)

See my comments under "Need the original code" above. Your table looks about right, but Baudot's code wouldn't have had a "new line" or "line feed" because the printer didn't have lines. Rees11 (talk) 13:37, 5 October 2009 (UTC)

That's what the book says. They must have performed some function analogous to carriage return and line feed on a typewriter. Intelligent Mr Toad (talk) 08:46, 6 October 2009 (UTC)

I suspect the newline codes were added later when the teleprinter was invented. I still think it would be cool to put in the Vansize table, but I would also be in favor of putting in your table. Do you know how to format it multi-column? (I don't). Rees11 (talk) 12:00, 6 October 2009 (UTC)

Yes the Colossus book is about WW2, so they are teleprinter commands. No, I don't know how to format it. (talk) 05:20, 12 October 2009 (UTC)

The above in table form (in both endiannesses; I see the one on the right more often):
Code Letters Figures Code Letters Figures
0 0 0 0 0 Null 1 0 0 0 0 E 3
0 0 0 0 1 T 5 1 0 0 0 1 Z +
0 0 0 1 0 Carriage return 1 0 0 1 0 D ENQ
0 0 0 1 1 O 9 1 0 0 1 1 B  ?
0 0 1 0 0 Space 1 0 1 0 0 S '
0 0 1 0 1 H #/£ 1 0 1 0 1 Y 6
0 0 1 1 0 N , 1 0 1 1 0 F  !
0 0 1 1 1 M . 1 0 1 1 1 X /
0 1 0 0 0 Line feed 1 1 0 0 0 A -
0 1 0 0 1 L ) 1 1 0 0 1 W 2
0 1 0 1 0 R 4 1 1 0 1 0 J BELL
0 1 0 1 1 G & 1 1 0 1 1 Shift to figures
0 1 1 0 0 I 8 1 1 1 0 0 U 7
0 1 1 0 1 P 0 1 1 1 0 1 Q 1
0 1 1 1 0 C  : 1 1 1 1 0 K (
0 1 1 1 1 V  ; 1 1 1 1 1 Shift to letters
Code Letters Figures Code Letters Figures
0 0 0 0 0 Null 1 0 0 0 0 T 5
0 0 0 0 1 E 3 1 0 0 0 1 Z +
0 0 0 1 0 Line feed 1 0 0 1 0 L )
0 0 0 1 1 A - 1 0 0 1 1 W 2
0 0 1 0 0 Space 1 0 1 0 0 H #/£
0 0 1 0 1 S ' 1 0 1 0 1 Y 6
0 0 1 1 0 I 8 1 0 1 1 0 P 0
0 0 1 1 1 U 7 1 0 1 1 1 Q 1
0 1 0 0 0 Carriage return 0 1 0 0 1 D ENQ
1 1 0 0 0 O 9 1 1 0 0 1 B  ?
0 1 0 1 0 R 4 1 1 0 1 0 G &
0 1 0 1 1 J BELL 1 1 0 1 1 Shift to figures
0 1 1 0 0 N , 1 1 1 0 0 M .
0 1 1 0 1 F  ! 1 1 1 0 1 X /
0 1 1 1 0 C  : 1 1 1 1 0 V  ;
0 1 1 1 1 K ( 1 1 1 1 1 Shift to letters
The letters, it can be argued, are coded to minimize the number of holes punched in paper tape, similar to IBM punched cards. In English, the most common 12 letters are approximately ETAOIN SHRDLU, which match the assigned hamming weights fairly well:
  • Hamming weight 1 (5): E LF SP CR T
    • Figures: 3 LF SP CR 5
  • Hamming weight 2 (10): A S I D R N Z L H O
    • Figures: - ' 8 ENQ 4 , + ) # 9
  • Hamming weight 3 (10): U J F C W Y P B G M
    • Figures: 7 BEL ! : 2 6 0 ? & .
  • Hamming weight 4 (5): K Q FIGS X V
    • Figures: ( 1 FIGS / ;
The figures don't make a lot of sense from a character-frequency point of view. Obviously ( and ) will appear with roughly equal frequency... (talk) 04:55, 3 February 2010 (UTC)
I like the way this table is organized, since it exposes the minimization of paper tape punches, though showing both bit orderings threw me; did the Baudot code not specify bit order? could a paper tape be fed in flipped backwards? if so, how would that be detected? -- Waveguy (talk) 01:25, 4 February 2010 (UTC)
The figures aren't allocated from a character-frequency point of view; they are allocated to make the design of the mechanism easy. Note from the picture in the main article that the keyboard only has three rows. The top row does double duty for letters - QWE...P and numbers - 123...0. So 1 is encoded as the same pattern as Q, 2 as W, etc and the sending mechanism doesn't have to do anything different for letter or figure shift. (Similarly the other characters in figure shift appear on the same key as the letter that shares their encoding.)
I wonder if it is worth making the point in the article that, unlike a modern keyboard where a shift key is held down whilst you type the associated keys, on a 5 level keyboard you press and release the shift keys and subsequent characters typed are interpreted according to the most recent shift key pressed? GrahamN-UK (talk) 13:42, 5 April 2013 (UTC)
There were no objections so I've added a note on shift key use to the Details section. GrahamN-UK (talk) 16:28, 9 April 2013 (UTC)

Actual historical Baudot code?[edit]

I'm no expert, but the above tables and references still don't look like the Baudot Code described and pictured in Alan G. Hobbs, "5 Unit Codes" (itself not quite the original, but the British Post Office version of the Baudot code). See also [2],[3], and James Gleick's The Information: A History, a Theory, a Flood (where I originally saw the pic). I guess we can't use the nice picture without permission, but we could add a table that contains the same information and stays faithful to its peculiar arrangement. I'll try to do one later and post it here first. --Kai Carver 12:11, 31 May 2011 (UTC)

Here's a (more or less finished) wiki version of the above-mentioned picture. Should we add it to the article, in the History section, before "Murray code"? :
A 1 P  %
B 8 Q /
C 9 R -
D 0 S  ;
E 2 T  !
E' & U 4
F f V '
G 7 W  ?
H h X ,
I o Y 3
J 6 Z  :
K ( t .
L = * * Erasure
M ) Figure Blank
N Letter Blank
O 5
Let. Fig. V IV I II III Let. Fig. V IV I II III
A 1 - .
E 2 X 9/
Y 3 S 7/
/ 1/ Z  :
I 3/ W  ?
U 4 T ²
O 5 V ¹
Let. Bl.
J 6 K (
G 7 M )
B 8 R -
H ¹ L =
F 5/ N £
C 9 Q /
D 0 P +
Fig. Bl. * *
"Figure 1 shows the allocation of the Baudot code which was employed in the British Post Office for continental and inland services. It will be observed that a number of characters in the continental code are replaced by fractionals in the inland code. Code elements 1, 2 and 3 are transmitted by keys 1, 2 and 3, and these are operated by the first three fingers of the right hand. Code elements 4 and 5 are transmitted by keys 4 and 5, and these are operated by the first two fingers of the left hand." Alan G. Hobbs, "5 Unit Codes", "Baudot Multiplex System"
--Kai Carver 17:43, 2 June 2011 (UTC)

An other original source/reference:

Figuier, Louis (1819-1894) (1867–1891), "chapitre IV: le télégraphe Baudot, à transmission multiple", in Furne, Jouvet et Cie (Paris), Les merveilles de la science, ou Description populaire des inventions modernes. 5-6, Suppléments. 5 / par Louis Figuier... (in French), Bibliothèque nationale de France, pp. 531 – 539, ark:/12148/bpt6k24678x  Unknown parameter |online presentation= ignored (help); Unknown parameter |rights= ignored (help)

Baudot-Murray Code Invalid As Shown[edit]

The "Pattern of Impulses" used in the table under the section ITA2 do not appear to be correct. The table caption cites an external page and the values shown there and those in the article do not match. The cited page's values are correct and so I believe that those shown in the article should be changed. JRSofty (talk) 11:11, 17 August 2011 (UTC)

I think the table in the article has the bit order reversed, with least-significant bit to the left, rather than the more usual order with the most-significant bit to the left, as in the two different references cited in the article. I agree it should be changed, but at least I understand why the table is so strangely different! (see also the discussion about endianness above) --Kai Carver (talk) 16:27, 21 August 2011 (UTC)

Undid big insertion with too much OR.[edit]

Undid this:

This is commonly implemented as a carriage-shift in mechanical teletypewriters; there is no way to differentiate upper-case and lower-case letters in ITA2 and most machines using it will print entirely in lower case, so that the well-known upper/lower case carriage-shift mechanism of the mechanical typewriter can be rededicated to the LTRS/FIGS distinction. Older machines required the sending user to switch manually between LTRS and FIGS mode using two dedicated keys, later machines commonly had separate keys for letters and figures and automatically inserted LTRS and FIGS codes whenever needed.
In ITA2, characters are expressed using five bits. ITA2 uses two code sub-sets, the "letter shift" (LTRS), and the "figure shift" (FIGS). The FIGS character (11011) signals that the following code is to be interpreted as being in the FIGS set, until this is reset by the LTRS (11111) character. "ENQuiry" will trigger the other machine's answerback. It means "Who are you?"
Sending "ENQuiry" will trigger the receiving machine's answerback, a mechanism that will automatically "type" a short pre-set identification string. "ENQ" thus means "Who are you?". CR is carriage return, LF is line feed; the separation of these two functions, which are almost always used in direct succession, happened because mechanical constraints limit the speed of the returning carriage, so that it cannot go all the way from right to left in the time allotted to the transmission of a single character. Thus teletypewriters generally require a sequence CR-LF or even CR-CR-LF to reliably start a new line when receiving pre-recorded messages at full speed, and later machines often had a separate key to create such a sequence with a single key-press.
CR is carriage return, LF is line feed, BEL is the bell character which rang a small bell (often used to alert operators to an incoming message), SP is space, and NUL is the null character (blank tape).
BEL is the bell character which rang a small bell at the receiving machine, often used to alert operators to an incoming message, or to mark a high-priority message that must be acted upon immediately. SP is space, and NUL is the null character (blank tape) which causes the receiving machine to just sit idle and wait for more characters. A couple of NULs could be used to optically separate messages on a continous tape while still allowing them to be sent in one session.


  1. No citations.
  2. The 5-bit machines were upper case only, not lower case only.
  3. End of line sequences are a difficult issue. CR LF is sufficient on a well-adjusted Teletype. Ham radio operators often sent CR CR LF LTRS LTRS to make sure the new line would start properly.
  4. The relationship between ENQ, WRU, answer-back, and ITA2 is more complex than that. See the CCITTT Telex standard from 1992.[4]. (I think the USTTY and ITA2 tables may be reversed in Wikipedia. USTTY does not have WRU, while ITA2 does.)
  5. NUL is an ASCII term. The all zeroes code is called the "blank" in the 5-bit world.
  6. The "small bell" isn't that small. In a Teletype Model 15, it's a 3 inch gong.

--John Nagle (talk) 03:46, 21 February 2012 (UTC)

  1. Granted, most of this is from personal experience. I had sort of hoped for others to provide citations...
  2. All machines I have worked on were lower-case only. Maybe a German specialty? Or a military thing? Admittedly, I don't think I ever used a non-German, or non-military, Baudot machine, but used lots of different models in the German army. I kinda just assumed that it was standard practice...
  3. German military standard was CR CR LF and many of the newer machines had a dedicated key for this exact sequence. Explanation concerning carriage speed comes from older mechanics who had used them and done repairs on them for decades (that was in late 1990 or early 1991). Again I have no written source, maybe it was just a legend.
  4. The terms "NUL" and "small bell" were in the article as I found it (and still are in there after your reversion).
Since you obviously know more about the topic than I, a mere anonymous stumbler-upon, do, I'd love you to correct my changes and re-insert that part of its content which it makes sense to keep in your opinion. Thank you, -- (talk) 04:37, 21 February 2012 (UTC)
There's a German 4-row layout with a 5-bit encoding used on the Siemens T100 [5] and some later machines, but I'm not sure what the encoding is. --John Nagle (talk) 06:11, 21 February 2012 (UTC)

Table is hard to read[edit]

The table shown for the Baudot code is very hard to read. There seem to be two tables, but not labeled (Continental & Inland?). And it's hard to pick out the actual characters encoded from all the other columns, especially as they are not in the standard leftmost spot. My suggestions (in order of least change):

  1. Add titles to the 2 tables.
  2. Add background colors to the actual character columns (possibly slightly different colors for the letters & figures columns).
  3. Move these columns to the leftmost spots.
  4. Fix the column labels for the fingers -- they use roman numerals in the table, but the paragraph below refers to them with arabic numerals.
  5. Convert the 5 columns for fingers into 1 column with 0's & 1's (like the Murray code table later in the article).
  6. Combine the 2 tables into 1, with the bit pattern shown only once, and 2 sets of characters, 1 for continental and one for inland (again, similar to the Murray table below, which does that for left-endian and right-ending bit codes).

I think any or all of these changes would make the table much more readable. T-bonham (talk) 04:21, 6 July 2012 (UTC)

They might well do so. Would you care to make them? Guy Harris (talk) 06:33, 6 July 2012 (UTC)

Inland & Continental[edit]

The article doesn't explain WHY there are 2 separate codes, or what "inland" and "continental" mean. Ace Frahm (talk) 10:06, 8 November 2013 (UTC)

Britain is an island and the British are sometimes insular. Given that the tables came from the British Post Office, it would seem that "Inland" would relate to the British Isles where the extended character set was obtained by the letters/numbers shift device. "Continental" would refer to the rest of Europe. There is an old joke about a British newspaper headline "Fog in Channel, Continent cut off". --TedColes (talk) 10:44, 8 November 2013 (UTC)
Yes, the E acute show the contiental code is the french one from Émile Baudot, and british telegraph probably needed to go through France to address rest of continent.
What is not clear is why does inland (british) use an extended character set? Did they have any additional character? — Preceding unsigned comment added by (talk) 21:29, 23 June 2014 (UTC)
At least one additional character - £. They also had some fraction parts. Guy Harris (talk) 21:39, 23 June 2014 (UTC)
This fraction concept is not so clearly described. Is it of a real use? Because with a metric system, the use of fraction does not appear necessary. For instance, is it limited to fraction with odd numerator?
Another question is why inland use mutliplex and continental simplex? — Preceding unsigned comment added by (talk) 00:25, 24 June 2014 (UTC)
Metric system? At the time, the British weren't using SI units, they were using Imperial units, the units of the British Empire, on which the sun never sets.
I don't know why they only had odd numerators, though. Guy Harris (talk) 02:05, 24 June 2014 (UTC)
Because, if the denominator is even, you only need odd numerators. The usual denominators are 8 and 10. USTTY (fractions variation, used for stockbroker purposes (not a stock ticker, that's completely different) has ⅛, ¼, ⅜, ½, ⅝, ¾ and ⅞. Inland code could do all those, plus tenths, (because they had "9/" in two characters.) Even Unicode doesn't go that far; it has fractions only up to ⅞. John Nagle (talk) 04:07, 24 June 2014 (UTC)

international telegraph alphabet[edit]

For clarity, might be baudot code and International Telegraph Alphabet should be on two separate wikipedia article?

  • Baudot Code: the invention; the transition from morse to digital text; pionner of digital communication
  • International Telegraph Alphabet n°2: worldwide standardisation

Then the question is which article does the International Telegraph Alphabet n°1 belongs to...

Tables are in pages 14 and 17.

Page 17 introduces


editor: P U B L I S H E D BY T H E G E N E R A L S E C R E T A R I A T OF T H E I N T E R N A T I O N A L T E L E C O M M U N I C A T I O N U N I O N



pages=23 and 24 — Preceding unsigned comment added by (talk) 21:58, 26 July 2012 (UTC)

And then there's the International Telegraph Alphabet n°5, also called just International Alphabet n°5, which has nothing to do with Baudot code (and everything to do with the non-8-bit bytes of the representation of every page on Wikipedia). "International Telegraph Alphabet" should probably be a disambiguation page for all the different ITAs.
This page covers some, but possibly not all, 5-level telegraph/teleprinter codes. There's Emile Baudot's code, which the article claims became ITA1. Then there's Donald Murray's code, which became the Western Union code, which, in turn, became ITA2. The article claims that no teleprinters used Baudot/ITA1, but they did use Murray/ITA2.
My inclination would be to keep Baudot/ITA1 and Murray/ITA2 together on this page, to give the full history, and have "International Telegraph Alphabet No1" (and other spellings, such as "International Telegraph Alphabet 1" and "International Telegraph Alphabet #1" and "International Telegraph Alphabet No. 1") and "International Telegraph Alphabet No2" (and other spellings), but not "International Telegraph Alphabet" with no number, redirect to the appropriate sections of this page. "International Telegraph Alphabet" should be a disambiguation page. Guy Harris (talk) 22:33, 26 July 2012 (UTC)

Gray code logic[edit]

After sorting voyels and consons in lexicographic order: Baudot code looks like Gray_code, base on indicated column order in wikipedia page:

Let ·Fig. · V · IV·   · I · II·III·
 A  · 1   ·   ·   ·   · ● ·   ·   ·    
 /  · 1/  ·   ·   ·   · ● · ● ·   ·    
 E  · 2   ·   ·   ·   ·   · ● ·   ·    
 I  · 3/  ·   ·   ·   ·   · ● · ● ·    
 O  · 5   ·   ·   ·   · ● · ● · ● ·    
 U  · 4   ·   ·   ·   · ● ·   · ● ·    
 Y  · 3   ·   ·   ·   ·   ·   · ● ·    
 B  · 8   ·   · ● ·   ·   ·   · ● ·    
 C  · 9   ·   · ● ·   · ● ·   · ● ·    
 D  · 0   ·   · ● ·   · ● · ● · ● ·    
 F  · 5/  ·   · ● ·   ·   · ● · ● ·    
 G  · 7   ·   · ● ·   ·   · ● ·   ·    
 H  · ¹   ·   · ● ·   · ● · ● ·   ·    
 J  · 6   ·   · ● ·   · ● ·   ·   ·    
 Fig. Bl. ·   · ● ·   ·   ·   ·   ·    
 *  · *   · ● · ● ·   ·   ·   · 
 K  · (   · ● · ● ·   · ● ·   · 
 L  · =   · ● · ● ·   · ● · ● · 
 M  · )   · ● · ● ·   ·   · ● · 
 N  · £   · ● · ● ·   ·   · ● · ●
 P  · +   · ● · ● ·   · ● · ● · ●
 Q  · /   · ● · ● ·   · ● ·   · ●
 R  · –   · ● · ● ·   ·   ·   · ●
 S  · 7/  · ● ·   ·   ·   ·   · ●
 T  · ²   · ● ·   ·   · ● ·   · ●
 V  · ¹   · ● ·   ·   · ● · ● · ●
 W  ·  ?  · ● ·   ·   ·   · ● · ●
 X  · 9/  · ● ·   ·   ·   · ● · 
 Z  ·  :  · ● ·   ·   · ● · ● · 
 –  · .   · ● ·   ·   · ● ·   · 
 Let. Bl. · ● ·   ·   ·   ·   ·   ·  — Preceding unsigned comment added by (talk) 20:30, 3 August 2012 (UTC) 

Physical order of letter (voyels first+ consons in alphabetical order is illustrated in figure 6, page 249

Titre : Journal télégraphique
Éditeur : [s.n.?] (Berne)
Date d'édition : 1884
Langue : Français
Identifiant : ark:/12148/cb32802376k/date
Source : Bibliothèque nationale de France, département Sciences et techniques, 4-V-1216
Relation : — Preceding unsigned comment added by (talk) 11:49, 6 August 2012 (UTC)


I also found some Gray Code related page:

And a nearly complete reference not on gray code, but on baudot history.

I understand that Gray Code, Baudot code, and Qwerty are from the 1870 time

The name "Gray code" is from 1953 or so; as the article says:
Bell Labs researcher Frank Gray introduced the term reflected binary code in his 1947 patent application, remarking that the code had "as yet no recognized name". ...
The code was later named after Gray by others who used it. Two different 1953 patent applications give "Gray code" as an alternative name for the "reflected binary code" ...
However, the concept does date back to (at least) the 1870's; as the article later says:
Reflected binary codes were applied to mathematical puzzles before they became known to engineers. The French engineer Émile Baudot used Gray codes in telegraphy in 1878. ...
so, yes, Baudot code is, apparently, what would today be called a Gray code. Guy Harris (talk) 00:32, 4 August 2012 (UTC)
A Gray code is only meaningful for a parallel implementation of something that has inherent sequence, like an encoder. Baudot transmission is serial, and the properties of a Gray code don't add any value. --John Nagle (talk) 06:10, 13 October 2012 (UTC)
I do not know if before being serial, Baudot transmission was parallel. But keyboard and reception are parallel oriented. (For my personal opinion, due to the complexity of the mechanism to translate serial code in parallel code. I cannot imagine they didnot use parallel mode in their research before industrialising the serial one.)
The decoder (of figure 9) behavior is described in details although it might apear rebarbative to the XXI° century reader.
In short, once the five bits are received and stored within a bobine/relais, the five aiguilles/goujons are set in the corresponding position, all this while they are in the neutral section.
Each received character is synchronous with one turn. Only one letter out of 31 matches the five aiguilles/goujons position.
If Gray code was not used, this would not work as between two letters some intermediate states might occur “some spurious position”, due to the nature of the technologies, “it is very unlikely that switches will change states exactly in synchrony”.
Figure 9 is commented: The original explanation (in page 250) does not name the Gray Code, but through the example, it cannot be another way; it is obvious the Gray property is used: « C'est seulement lorsqu'ils arrivent sur la case「 W ? 」que tous les goujons se trouvent sur des cavités et, par conséquent, peuvent y entrer. L'axe tourne, le cadre ce s'infléchit par l'action du ressort R et le bras descend avec sa partie libre. Dans la case suivante 「 X 」, les goujons remontent de nouveau, le cadre ce se redresse parce que le goujon 3 ne peut as descendre dans une cavité, et cet état continue jusqu'à la fin du tour qui les ramène dans le cercle neutre. Donc, à chaque combinaison produite sur le manipulateur à la station de départ, correspond à la station d'arrivée une case spéciale et une seule des 31 cases du combinateur où tous les goujons pourront entrer dans des cavités et où le levier descendra dans le tube de l'axe.» (Journal télégraphique, 1884) (talk) 20:55, 25 October 2012 (UTC)

Another possible layout[edit]

I intend to present a layout clearer and more understandable than th on of the current article. At least, the layout here after presents the baudot code with some kind of logic.

Or at least it is a layout with a logic near of the logic of the first model constructed by Sholes used a piano-like keyboard with two rows of characters arranged alphabetically as follows:

   - 3 5 7 9 N O P Q R S T U V W X Y Z
    2 4 6 8 . A B C D E F G H I J K L M

So the Baudot code with this similar layout logic:

                     ●  ●  ●  ●  Ⅲ
               ●  ●  ●  ●        Ⅱ
            ●  ●        ●  ●     Ⅰ
Ⅳ                                   Ⅴ = " "
·  let   ¬  A  É  E  I  O  U  Y
·  chf   ¬  1  ¬  2  ¬  5  4  3
●  let   ¬  J  H  G  F  D  C  B
●  chf      6  ¬  7  ¬  0  9  8
                     ●  ●  ●  ●  Ⅲ
               ●  ●  ●  ●        Ⅱ
            ●  ●        ●  ●     Ⅰ
Ⅳ                                   Ⅴ = "●"
●  let   *  K  L  M  N  P  Q  R
●  chf   *  (  =  )  ¬  +  /  -  
·  let   ¬  _  Z  X  W  V  T  S
·  chf      .  :  ¬  ?  ¬  ¬  ¬

— Preceding unsigned comment added by (talk) 22:53, 3 August 2012 (UTC)

Previous keyboard[edit]

Hughes telegraph (1866-1914), first telegraph printing text on a paper tape. Manufactured by Siemens und Halske, Germany; range: 300-400 km

When this article deals with the Murray Qwerty Keyboard; it might make sense to indicate that other keyboard were considered before, such as explained in Ticker tape page article (See Illustration). — Preceding unsigned comment added by (talk) 09:23, 5 August 2012 (UTC)

The Hughes printing telegraph shown doesn't have a binary code, although it does have a character sequence. The print wheel turns continuously in sync (if everything is properly adjusted) with a sending rotor at the other end. When you push a key, it raises a contact under the sending wheel, and when the rotor comes around and hits it, a pulse is sent. Unfortunately, the site that had video of one working is no longer available. --John Nagle (talk) 06:39, 13 October 2012 (UTC)

IA1; Murray code, murray keyboard, markrom keyboard and IA2.[edit]

  • This article does not deal with IA1, nor with other variants of BAudot'original code, such as the 1877 article by Émile Baudot.
  • And why the É character in Baudot Code? Beause of Émile?
  • This article does not deal with the logic of the order of Murray code.

It appears t be decribed in Figure 8. Murray printer code, 1889.17:

To maximize the structural stability of the tape, 20 Murray arranged the characters in his code so that the most frequently used letters were represented by the fewest number of holes in the tape.

  • This article does not deal with the Murray keyboard (1901 variant)
  • The explanation that Murray code is based on a Qwerty keybard is half true, as it appears the digit association to Qwerty is only from 1902.

The article does no deal with morkrom keyboard.

For those issues there are some piece of information just here after:

The Evolution of Character Codes, 1874-1968 Eric Fischer — Preceding unsigned comment added by (talk) 10:06, 5 August 2012 (UTC)

USTTY vs ITA2[edit]

For a while, I was confused by the differences between USTTY and ITA2. (This is a practical problem for me; I restore old Teletype machines.) The Bell System and military machines usually seem to be USTTY, while Western Union Model 14 machines from as early as 1926 are ITA2. The difference is that BELL and apostrophe are reversed. Old manuals for the Teletype Model 14 indicate that you could order either option. (This added some mechanical complexity for BELL, which is the only non-typing function that operates in only one shift.) There are also "fractions" (1/8, 1/4, 3/8, 1/2, 5/8, 3/4 and 7/8 replace most of the punctuation marks), and "weather" (weather symbols replace punctuation) character sets. Also, some machines have "£" instead of "#". All those are just different typebar sets; the mechanism itself is the same. --John Nagle (talk) 06:22, 13 October 2012 (UTC)

Still looking for a definitive source on this. Online sources differ on whether BELL is on FIGS-S or FIGS-J for ITA2. (See [6], which is reversed from the tables in the Wikipedia article.) Teletype Corporation Bulletin 1048, page 11 [7] shows this as a configuration option. It uses the terminology "Bell on shifted-S" and "Bell on shifted-J", and gives the part numbers to order based on customer preference. The manual also says "Check carefully the position of pullbars in segment and the arrangement of the type pallets in the typebars. If these differ in any way from the arrangement shown on this page, the parts should be ordered by name...". Looks like there was an attempt to get machines standardized, back in 1941. But even that table allows four different arrangements (fractions/communications, and "Bell on shifted-S"/"Bell on shifted-J". (I own four Teletype machines, and they have three different character sets.) It's probably worth mentioning that there were some variations. I have two Model 14 machines with a pilcrow (¶) on FIGS-F, replacing "!". Some machines have "£" in place of "#". Telegrams tended not to use much punctuation. One begins to see why. John Nagle (talk) 18:24, 25 October 2012 (UTC)

Original Baudot code.[edit]

I suggest to provide this table for the original Baudot code.

This table can replace the previous one mixing the british and the continental code in a single table, and keeping the several orders to display the table.

Missing items CCIT ITA-3 (and ITA-4) "ITA-2 P"??[edit]

There is mention in more than one place and on Wikipedia here about a six bit ITA-3 code that was evidently put forth by the CCIT in 1970 See this link for an example.And here is more. This article is lacking any information about this, its history and rational for being. After typing this, I just found this Wikipedia article. Also, there is a mention of "ITA-2 P" - evidently a recent variant of ITA-2. The book "Technical Handbook for Radio Monitoring VHF/UHF" 2011 states that "ITA-2 P" is a variant of ITA-2 that is converted to a seven bit code with synchronization and parity bits. I wanted to mention this for a possible (needed) update to this article concerning these codes, whoever can get to it first. Nodekeeper (talk) 15:10, 15 September 2014 (UTC)

See ARQ-M. That's not really a different code; it's a protocol to encapsulate ITA-2 for error checking and radio transmission. It adds acknowledgement and retransmits. John Nagle (talk) 04:44, 16 September 2014 (UTC)

Big changes without cites[edit]

These changes [8] by Verdy p (talk · contribs) raise some concerns. There are no citations, and some of the material inserted may be incorrect. The section "The code position assigned to Null was in fact used only for the idle state of teleprinters. During long periods of idle time, the impulse rate was not synchronized between both devices (which could even be powered off or not permanently interconnected on commuted phone lines). To start a message it was first necessary to calibrate the impulse rate a sequence of regularly timed "mark" pulses (1) by group of 5 pulses, which could also be detected by simple passive electronic devices to turn on the teleprinter; this series of pulse was generating series of Erasure/Delete and also initiliazing the receiver state to the Letters shift mode, however the first pulse could be lost, so this power on procedure could thne be terminated by a single Null immediately followed by an Erasure/Delete character." does not apply to ordinary teleprinters. The whole point of the Baudot systems is that it's asynchronous, and resynchronizes on each character. (That text may refer to some early electromechanical multiplexing system. It's not listed as a known electronic multiplexing system.[9]) Unless some references appear, I'm going to revert this. Thanks. John Nagle (talk) 03:52, 28 May 2016 (UTC)

Async communication endianness, LSB on Left vs Right[edit]

The article should be specific about which code element was sent first in the original serialization of Baudot for teleprinters, i.e. the endianness. Was it code element 1? (Or is that really called code element I?) Was that really set by the index finger of the right hand (and is that thus the "first" finger?). Then moving to the right 2 (II) and 3 (III)? Then over to the left hand moving to the left for 4 (IV) and 5 (V)? So the bits can't be read out across the keyboard either left to right or right to left? I have the same question for the latter codes, and I'm also wondering why the tables show both LSB-left and LSB-right versions for ITA2/Baudot-Murray code. Does that suggest that some of this was really not standard? Or does it reflect people trying to read punched tapes both as they feed in to a reader, and as they come out of a tape punch, which might well be different orientations? And how were tape readers oriented - any standardization there? Thanks. ★NealMcB★ (talk) 21:30, 26 July 2016 (UTC)

  1. ^ Ordre donné par
  2. ^ Code de Gray
  3. ^ Ordre donné par