Talk:Teleprinter/Archive 1

TTY
Does TTY stands for TeleTYpe? 16@r 22:55, 8 October 2006 (UTC)


 * Yes. Guy Harris 18:10, 9 October 2006 (UTC)
 * Thank you, I am going to modify the article in consequence to make it more understanble at first sight. And as for TDD? *-) 16@r 18:19, 9 October 2006 (UTC)


 * Telecommunications device for the deaf. Guy Harris 18:25, 9 October 2006 (UTC)

Untitled
My parents were both teleprinter operators for the RAF during WW2. One day my dad, who worked at Bomber Command HQ, Northolt, chatted via teleprinter to my mum, working near Brighton, saying he was on leave one week-end. My mum was also was on leave that same weekend. I'm here because of that teleprinter communication in 1945. I say to everyone that my parents met on-line before the Internet was invented! Acb58 20:34, 16 January 2007 (UTC)

Types of teleprinter networks
The section on "teleprinter operation" currently speaks of 5-bit codes and paper tapes, but doesn't mention how the teleprinter was used by humans: How long was the typical message, how much did the customer pay, what was life as an operator like, could small companies have their own teleprinter or were they forced to use the cable/phone company, etc. When reading other literature, the phrase "teleprinter conversation" struck me. Compared to later technologies, it seems the teleprinter had more in common with Internet chat than with Internet e-mail. The teleprinter was of course a direct descendant to the morse key. Operators (live people) were at each end, having a conversation, exchanging personal greetings to each other, and occasionally transmitting (business / official) messages for others. One important difference must have been that typing at the teleprinter keyboard required less special skills than using a morse key, so presumably more people could (theoretically) enter into such direct conversations (but did they?). It's hard to imagine a business manager or army general taking the place of a morse key operator. Instead they would write down a message and have someone transmit it for them (much like e-mail). But perhaps they occasionally engaged in teleprinter conversations (chat) of their own at the keyboard? Have such cases been documented in literature (fiction, history, auto/biographies)? --LA2 12:39, 31 July 2007 (UTC)


 * Good point. There were several ways teleprinters were used.  They were used to send and receive Telegrams for most of the 20th century.  Telegrams were handled in a store-and-forward way, like e-mail.  Users usually submitted telegrams on paper blanks to be typed by operators, but some offices had teleprinters for sending and receiving telegrams.
 * There were also dial up point-to-point teleprinter networks. In the US, there was TWX (the Bell System product) and Telex, the Western Union system.  These systems had international links.  Germany had a very extensive switched teleprinter system.  With those systems, you could conduct a conversation, typing back and forth. But you paid for connect time, not characters transmitted, so this was seldom done.  Usually, you pre-punched a paper tape and sent it at full speed.
 * One sender could drive any number of printers, so news and stock quotes went out over one-way teleprinter networks. Those teleprinters were "receive only" (no keyboard).  --John Nagle (talk) 05:42, 25 January 2009 (UTC)


 * We need more on this. There were at least four types of teleprinter networks:
 * Dial-up, like Telex and TWX. These created a real-time circuit between two machines. US and UK systems had actual telephone dials; German systems did "dialing" via the keyboard.
 * Store and forward systems. This was E-mail, done with electromechanical gear.  System types included "torn tape" centers, where operators tore sections of paper tape off from tape punches ("reperforators"), read the destination, and manually put the section of tape into a reader that would send towards the ultimate destination.  An automated replacement for this was "Western Electric Plan 55-A", which needs references. ("Gilbert S. Vernam, "Automatic Telegraph Switching System Plan 55-A", AIEE Transactions on Communication and Electronics, May 1958, p. 239.") Created an article on Plan 55-A, which worked much like Sendmail, but was implemented with tape punches, readers, and telephone switchgear. Buffering consisted of a paper tape punch feeding into a reader, with a big bin in the middle.  Late 1940s technology.
 * There seem to have been at least four Western Union generations of this:
 * "Torn tape" - operator tears off punched tape for one message, reads header, puts tape in proper outgoing reader.
 * "Plan 21-A" - operator reads printing on incoming tape, pushes button to select destination for message.
 * "Plan 55-A" - system reads incoming punched tape twice, once for header, once for data, routes automatically. Huge installations. Needs a big operating staff to keep everything going.
 * "InfoMaster" - a mainframe computer in Middletown, VA does all the switching, without paper tape.
 * Broadcast systems, or "news wires", like the Associated Press and United Press (later UPI).
 * "Loop" systems, where anything typed on any machine on the net printed on all the machines. Police departments used such systems to interconnect precincts. Need a reference for this.


 * One possible source of Plan 55 information is http://www.westernunionalumni.com/mplsrpf.htm. Also note that Vernam has a special place in the history of cryptography based on his work with one-time tape systems. Wa3frp (talk) 14:24, 11 March 2009 (UTC)


 * I just found a good source on Plan 55-A: Leonard Kleinrock's PhD thesis at MIT.. This is a great find. Kleinrock is one of the inventors of the ARPANET, one of the people who developed theory for network congestion, and it turns out he developed the theory to analyze Plan 55-A. --John Nagle (talk) 05:04, 12 March 2009 (UTC)

Minor fact checking

 * Marked ham radio free teletype with removed parts story as "cite needed".

This "story" is close to the truth but has a whisper-down-the-lane quality. First, the teleprinters released to hams after WW II were mainly model 26 machines, not model 15 machines. Second, I've seen and heard where the keyboard of one machine was smashed and the page printer of another teleprinter was smashed requiring a ham to acquire two such machines that could be reassembled into one. It does not seem credible to me, nor have I ever seen - save this article update - that teleprinter parts were removed by the military by request of AT&T. The next line, when modified to read "...Radio amateurs were permitted to buy complete machines from commercial operators after signing a contract which stated that the machines would never be used in commercial service and would not be sold to anyone who had not also signed the contract..." is fact based. The October 1969 issue RTTY Journal has a copy of this type of contract. Wa3frp (talk) 15:15, 3 March 2009 (UTC)
 * That's a "cite needed" situation. See WP:V. --John Nagle (talk) 06:33, 11 March 2009 (UTC)


 * The Model 15 definitely has a cast metal frame, but it's probably mild steel, not iron, because it has small-diameter drilled and tapped holes. Fine threads usually don't work well in cast iron.  Minor point.  Don't have a source either way; even the Teletype manuals don't say. --John Nagle (talk) 07:18, 26 February 2009 (UTC)

"In popular culture" needs citations
Nothing in the "popular culture" section has citations. I'm going to pull most of that unless some citations appear. --John Nagle (talk) 02:07, 13 March 2009 (UTC)
 * Deleted "popular culture" section, per WP:TRIVIA. --John Nagle (talk) 04:51, 24 April 2009 (UTC)

"Two stop bits"
The article says that Teletype machines required two stop bits. Actually, the mechanical machines needed a stop period of about 1.5 bit times. I need a solid source for this. --John Nagle (talk) 16:53, 7 May 2009 (UTC)


 * Teleprinter required at least 1.5 stop bits on received signals but some could also transmit with 2 stop bit. UART ICs could be configured to transmit 1, 1.5 and 2 stop bits, but this did not affect reception. Sv1xv (talk) 18:54, 7 May 2009 (UTC)


 * The stop bits can be 1.0, 1.42, 1.5 and 2.0. It depends on the service.  50 Baud Telex running at 67 words per minute was a common user of 1.42 stop bit time. I have a document showing the gear sets available for the Teletype Model 28.  The unit codes vary from 7.0 to 7.42 to 7.5 to 8.5. Wa3frp (talk) 23:46, 7 May 2009 (UTC)


 * The original Teletype Model 15 manual (1944 edition) says that the typing unit runs one-seventh faster than the keyboard unit, which reflects the stop time. When the keyboard is running continuously, the typing unit is supposed to be declutched one seventh of the time.  But there may be some additional stop time in there. Unfortunately, there's no cam timing chart in the adjustments manual. Incidentally, there's an advantage to using a non-integral stop time.  With a stop time of one or two bits, it's possible for some repeated character sequences to stay in sync incorrectly.  With an odd-length stop time, the receiver will slip until it's synched correctly. --John Nagle (talk) 00:52, 8 May 2009 (UTC)


 * That's not to mention how it's talking about ASCII codes, parity etc... I think someone's making up their own history to conveniently suit other facts even though there may not be any other connection than co-incidence (or imagination). A "100 WPM", typing pool era teletype (with the "word" being a poorly defined figure somewhere around 5 symbol characters plus one space and a small average amount of punctuation), using 5-bit Baudot code, one start bit and 1.42 stop bits (necessary for the reliable mechanical operation) at ~75 baud, somehow translates to an early 110 baud electronic computer modem standard by way of 7-bit ascii, one parity, one start and two stop bits? Erm, ok then... There's nothing else involved, then?
 * (I could believe the even earlier 75 baud standard, and its use with 75 up/1200 down split services such as Prestel which may have incorporated some leftover teletype kit as a receiving terminal for the page selection/transmitting equipment, but dunno where the 110 came from!) 193.63.174.10 (talk) 12:23, 12 June 2009 (UTC)


 * Agreed. I took out the paragraph which equated 100 speed to 110 baud.  It's not. Info about 75/1200 modems might go into the Prestel and Minitel articles.  --John Nagle (talk) 17:01, 12 June 2009 (UTC)

General cleanup
I've done some cleanup and reorganization. I've added some descriptive material on how teleprinters were used; someone had mentioned above that there was no information about what people did with the things. The machine data is now organized by manufacturer and model. --John Nagle (talk) 18:08, 9 May 2009 (UTC)

+++ Is this a good place to discuss the "RUBOUT" character used when an error occurred while operator was typing a message on paper tape? If the operator detected the error before typing another character, he could backspace the paper tape punch, press RUBOUT which punched all channels of the offending chacter. This would be ignored on the receiving end when the tape was transmitted. (I realize the article is already rather large) I always thought that one of the most important differences between the model 33 and model 35 was the duty cycle. The model 33 had a 1 hour on and two hour off duty cycle where the model 35 was continuous duty.

DGerman (talk) 22:29, 15 June 2009 (UTC)


 * RUBOUT and its history are discussed at ASCII. The duty cycle issue is hinted at in the list of Teletype products. The Model 15 and the 28/35 were the heavy duty machines; all the others didn't last very long. (It's easier to find a working Model 15 (1930s technology) today than a working Model 33. The tradeoff, though, is that a Model 15 has about 500 oiling points and several hundred adjustments.) --John Nagle (talk) 01:58, 16 June 2009 (UTC)


 * Deleted "Interest in teleprinters is actually beginning to grow again", which someone else marked as needing a cite. If there's growing interest, it's not growing by much. See Greenkeys Digest to get a sense of the size of the community. --John Nagle (talk) 05:07, 11 May 2010 (UTC)

Obsolescence wasn't even
In some settings the teleprinters stuck around longer than in others. For example settings characterized by a combination of security needs, long formal approval processes for equipment, and limited budgets for replacements. I still used them in the German military during an exercise in 1999 or so and at that time there was still an active automated teleprinter network for internal military use in place, in which you could dial into other machines by typing special codes directly on the TTY itself. Also I think I've read somewhere that our Foreign Office used them to communicate with the smaller embassies around the world well into the 2000s, in combination with tried-and-true encryption devices. -- 77.189.14.38 (talk) 16:17, 15 March 2011 (UTC)

Teletype model 31
I recently found out that there was a Teletype Model 31. This was a tape printer, like the Model 14, but used a typewheel, like a model 35. Released around 1951. The previous tape printer design was the Model 14 from 1924, and Western Union still used them, so Teletype came out with something for WU. A rare machine, and not very interesting. --John Nagle (talk) 18:58, 18 November 2011 (UTC)


 * I merged the list of Teletype models into the list on the Teletype Corporation page and removed the list from here; I've added some information about the Model 31, with your link as a reference, to that list. Guy Harris (talk) 22:52, 2 April 2012 (UTC)

Teletype redirect
Teletype currently redirect here. I know it was often used as a synonym for teleprinter, but it was a trademark of the Teletype Corporation and I think it should redirect there.--agr (talk) 16:43, 3 April 2012 (UTC)


 * If that's done, there should probably be a hatnote in the Teletype Corporation page sending people to this page if they're interested in the general concept of teleprinters rather than specifically interested in the Teletype Corporation or in its teleprinters. Guy Harris (talk) 16:48, 3 April 2012 (UTC)


 * I agree on adding the hatnote. It's my opinion that the redirect was put in place when the Teletype Corporation article was a four sentence stub. I am working, as time permits, to continue to improve the Teletype Corporation article.Wa3frp (talk) 03:42, 4 April 2012 (UTC)

Stop bit at 1.42 bit times
Someone added a note about the 1.42 bit stop time being related to HF radio fading. It's much older than that. The Teletype Model 14 manual (1931 edition, but the machine is from 1924) says (p.15, "Synchronism") "The reason the selector cam sleeve is made to rotate one-seventh faster than the transmitting camshaft is this: it is not possible to maintain several units at exactly the same speed". The stop bit as sent is always longer than the receiving end needs, so the receiver pauses for a few milliseconds on every character. This gets things back in sync on every character, even at full speed. On mechanical teletypes, this involves a mechanical clutch. The 1.42 number comes from a gear ratio within mechanical teletypes.

This is getting deeper into the subject than Wikipedia needs. Here's a link to a set of maintenance manuals for older mechanical teletypes. For Wikipedia purposes, it might be worth having an article on "unison devices", a 19th century term for getting two things in sync and still a category of patents. Edison's first big invention was a unison device for a stock ticker, and the history of unison devices continues through teleprinters, TV horizontal hold, and modems, to today's phase-locked loops. There's an article on Synchronization, but it has no references and is mostly a list of pointers to other articles. --John Nagle (talk) 04:37, 8 April 2013 (UTC)

Bain not successful
Noted that while Bain was perhaps the first to try to build a printing telegraph, it was a flop. He couldn't keep both ends in sync, despite, as the reference puts it, "ever larger governors". John Nagle (talk) 08:05, 3 January 2017 (UTC)


 * Forgive me if I'm being slow but where is Bain's telegraph mentioned in either of those references?SQMeaner (talk) 10:27, 3 January 2017 (UTC)


 * Nowhere, but those references weren't part of this section, and thus weren't for this discussion. I've added a  template to the previous section so that the references end up there, rather than in this section. Guy Harris (talk) 18:33, 3 January 2017 (UTC)


 * The reference describes a number of machines. The quote


 * "A much more elaborate pattern of chemical telegraph also introduced in 1843 had identical mechanically-driven sending and receiving instruments. This had a rotary sender using punched tape, on the Jacquard principle, running under two metallic feelers to make and break the electrical circuit; to receive messages the punched paper tape roll was replaced by strips of chemically-sensitive paper, which had to be kept damp, running under the metal points to cause the marks. The need for precise synchronisation defeated this first effort at automatic telegraphy, despite the use of ever larger centrifugal governors on each instrument."


 * is in the description of the "chemical telegraph of 1843". The telegraph mentioned in the article is, I presume, the one described in the section "The Mechanical Telegraph", which says:


 * "The mechanical telegraphs used galvanic electricity to manage power produced by clockwork. In Bain’s first telegraph pulses of electricity created by rotating a dial over contact points were used to release and stop a type-wheel turned by weight-driven clockwork; a second clockwork mechanism rotated a drum covered with a sheet of paper and moved it slowly upwards so that the type-wheel printed its signals in a spiral. The critical issue was to have the sending and receiving elements working synchronously. Bain attempted to achieve this using centrifugal governors to closely regulate the speed of the clockwork."


 * which seems to indicate, from "attempted to achieve", that it didn't work well enough either. I'm not sure why the chemical telegraph was described as "this first effort at automatic telegraphy" given that, in the section before it in the article, it discusses the mechanical telegraph. Guy Harris (talk) 18:46, 3 January 2017 (UTC)
 * I don't know. Bain made several telegraphs operating on different principles and the claim that the first quote listed here is referring to Bain's printing telegraph seems shaky to me. The second quote here doesn't describe whether Bain's printing telegraph was successful or not either, only going so far as to say Bain 'attempted' to make a printing telegraph. For what it's worth I've found a reference of my own which I've listed below which indicates that Bain had constructed a working model of his telegraph by 1840. Seeing as he managed to interest someone in sponsoring him that seems to me to imply that the model worked. I'm not sure whether it was truly practical or not, but writing it off as a failure doesn't seem right to me based on the evidence I've seen so far. I'd also like to point out that the reference below describes 'a version' of his printing telegraph being used in 1844 by the London and South Western Railway.SQMeaner (talk) 19:11, 3 January 2017 (UTC)


 * https://www.slhf.org/sites/default/files/publications/slhf12_alexanderbain.pdf


 * For a detailed description of Bain's printing telegraph, see . It didn't have a full alphabet, just 0-9, "-", and space. Bain later built some non-printing telegraphs that worked, but did not, apparently, make any further progress on printing mechanisms. The "chemical telegraph" didn't have a sync problem; that was just a mechanism for printing dots and dashes using damp paper treated to change color when a current passed through it.  Without amplification, the early telegraph people all struggled with making a sensitive enough receiver. See Syphon recorder for one solution.  Getting enough power through to control a printing mechanism was very tough. The "chemical telegraph" addressed that problem. John Nagle (talk) 01:10, 4 January 2017 (UTC)


 * I've done a bit more digging and I'm now fairly certain that Bain's telegraph was up and running by April 1844 at the latest, as you can see in the link below. However, that seems to be beside the point now as I've found even earlier printing telegraphs by Caul August Steinheil in 1836, which used Morse style dots and dashes, and by Alfred Vail in 1837, which used a full alphabet. Does anyone think these guys deserve a mention or should we focus on Bain's and House's printing telegraphs for this article?SQMeaner (talk) 02:41, 4 January 2017 (UTC)


 * https://books.google.co.uk/books?id=IE84AAAAMAAJ&pg=PA268&dq=Bain+1844+Telegraph&hl=en&sa=X&ved=0ahUKEwixl-TisKfRAhXqKcAKHcp8CxwQ6AEIIDAA#v=onepage&q=Bain%201844%20Telegraph&f=false

Washington to Moscow hotline- suggestion
Suggest someone who may have had first hand knowledge add some detail about the former iteration of the hotline that was established to link the leaders of the two nations most likely to use nuclear weapons against each other. — Preceding unsigned comment added by 160.136.159.179 (talk) 22:57, 10 February 2020 (UTC)

Vague sentence in Manufacturers > Teletype Corporation section of article
The sentence "Very few complex machines can match that record." for the Model 15 is vague. There should be a following sentence, that says something like "For example, the whatchamacalit was known to last for about 28 years before complete failure". Otherwise, this should be removed. Scott Daniel (talk) 05:19, 16 August 2018 (UTC)


 * Or, more appropriately, "For example, the whatchamacallit only remained in production for about 28 years"; the preceding two sentences were


 * "The Model 15 stands out as one of a few machines that remained in production for many years. It was introduced in 1930 and remained in production until 1963, a total of 33 years of continuous production."


 * so the record isn't one for how long a machine remained in operation after it was manufactured, it's one for how long the line of machines remained in production.


 * But, no, they don't give any examples of other complex machines that either did, or didn't, match that record. Guy Harris (talk) 09:39, 16 August 2018 (UTC)

"Here Is" vs "Answer Back"
I changed the "here is" section to "answer back mechanism". Here's a video of a Teletype Model 28 answer back mechanism. This is a Baudot machine. So this predates ASCII teletypes. John Nagle (talk) 20:00, 12 February 2017 (UTC)


 * How do you reconcile this with the fact that the associated key is labeled "Here is"? Jeh (talk) 20:16, 12 February 2017 (UTC)
 * The red "Here is" key is mostly for testing. It has no code. The red keys at the top of the keyboard perform various local functions mechanically.  The key to query the other end is was originally marked "WRU", (Who Are You). This goes back to the Teletype Model 15.. In ASCII-1963, there was a "WRU" character. In the final version of ASCII, it's "ENQ". The purpose of this is so a user could dial up an unattended machine and query its identity before sending to it. John Nagle (talk) 22:31, 12 February 2017 (UTC)
 * The key is labeled "Here is", but the Teletype documentation calls it an answer-back mechanism as early as February 1964 . (That's my video linked above.) Kb8ojh (talk) 01:24, 15 February 2017 (UTC)

Baudot vs. ITA2
Someone recently changed "Baudot" to "ITA2". That's OK. For more than you ever wanted to know about five level codes, see. There were a lot of variations. The mapping for letters and numbers was the same on all machines, but special characters varied a lot. There's ITA2, USTTY, "Fractions" (1/8, 1/4, 3/8, ... used for stock market applications (this is different from a stock ticker; those had a completely different code) and "Weather symbols". I have five Teletype machines, and only two of them have identical character sets. John Nagle (talk) 08:00, 15 August 2017 (UTC)


 * According to Baudot code the first variant from original Baudot (the Murray code, which soon became ITA-2) differed quite a lot in the letters. After that, the letters stayed the same. Also according to that article there was never actually a full-alphabet keyboard for the true original Baudot code; the operator had five keys, one for each bit! Jeh (talk) 08:15, 15 August 2017 (UTC)

TeleType with Screen
On Science Channel's Dark Matter they depict a teletype with a CRT in 1958 given to Dr Van Allen. This led me to wonder: When were those ubiquitous green CRTs first used? 76.21.107.221 (talk) 03:12, 15 September 2011 (UTC)