TWX

posted in: Communications | 0

Almost in parallel with Germany’s telex system, Bell Labs in the 1930s decided to go telex one better, and began developing a similar service (with pulse dialing and all!) called “Teletype Wide-area eXchange” (TWX).

TWX originally ran 75 bits per second, sending Baudot code and dial selection. However, Bell developed a second generation of “four row” modems called the “Bell 101 dataset,” which is the direct ancestor of the Bell 103 that launched computer time-sharing. The 101 was revolutionary because it ran on ordinary subscriber lines that could (at the office) be routed to special exchanges called “wide-area data service.” Because it was using the public switched telephone network, TWX had special area codes: 510, 610, 710, 810 and 910, some of which remain in use.

The “four row” TWX service had “control characters” that let the machine behave like office typewriters. These provided paragraph indentation, form feeds, and other services that were never available with Baudot codes. However, the TWX code only used 93 of 128 characters.

The Teletype corporation was founded by a Dr. Kleinschmidt. It had the cheapest teletypewriters that could be adapted to the TWX code. Bell purchased the corporation to assure its supply of “model 33” TWX teletypewriters.

The model 33 was the cheapest teletypewriter available for use with computers. Computer people of course wanted a full set of characters. Teletype provided them.

ASCII was born from TWX code. It was formalized as CCITT international alphabet 5. Careful study will show that ASCII traces many character codes back to Baudot, which in turn traces some characters back to manual telegraphy.

Bell’s original consent agreement limited it to international dial telephony. WUTCo (Western Union Telegraph Company) had given up its international telegraphic operation in a 1939 bid to monopolize U.S. telegraphy by taking over ITT’s PTT business. The result was deemphasis on telex in the U.S. and a cat’s cradle of small U.S. international telex and telegraphy companies. These were known by regulatory agencies as “International Record Carriers”

  • Western Union Telegraph Company developed a spinoff called “Cable System.” Cable system later became Western Union International.
  • ITT’s “World Communications” was amalgamated from many smaller companies: “Federal Telegraph,” “All American Cables and Radio,” “Globe Wireless,” and a common carrier division of Mackay Marine.
  • RCA communications had specialised in crossing the Pacific. It later joined with Western Union International to become MCI.
  • Before World War I, Tropical Radiotelegraph put radio telegraphs on ships for its owner, The United Fruit Company, in order to deliver bannanas to the best-paying markets. Communications expanded to UFC’s plantations, and were eventually provided to local governments. TRT Telecommunications (as it is now known) eventually became the national PTT of many small Central American nations.
  • The French Telegraph Cable Company (owned by French investors) had always been in the U.S. It laid cable from the U.S. to France. It was formed by “Monsieur Puyer-Quartier.” This is how it got its telegraphic routing ID “PQ.”
  • Firestone Rubber developed its own IRC, the “Trans-Liberia Radiotelegraph Company.” It operated shortwave from Akron OH to the rubber plantations in Liberia. TL is still based in Akron.

Bell telex users had to select which IRC to use, and then append the necessary routing digits. The IRCs converted between TWX and Western Union Telegraph Co. standards.

Around 1965, in a near-psychotic break with existing standards, DARPA commissioned a study of decentralized switching systems, hoping to find something more advanced than TOR that could still hope to survive a nuclear war. The contractors developed the internet.

The internet was a radical break in three ways. First, it was designed to operate over any media. Second, routing was decentralized. Third, large messages were broken into fixed size packets, and then reassembled at the destination. All previous networks had used controlled media, centralized routers and dedicated connections.

The internet was designed with nearly grotesque economies. It is commonplace for internet packets to use less than 1% of their bits for overhead. This cheapness combines synergistically with the internet’s ability to live on other media. A typical cycle occurs when the internet encounters another network, like telex, fidonet, ATM, or (as we are seeing with cable-modem based internet phones) the public switched telephone network:

  • First, internet protocols are tunneled through the other network, as a convenience, usually for some specialized or office application.
  • Second, users come to expect the reliable global interconnectivity of the Internet, often for e-mail, or nowadays, for web access. Just because it’s old and well debugged, the internet can seduce a user with a young, poorly behaved proprietary network.
  • Third, native applications of the competing network are deprecated, often because “nonproprietary” internet versions of similar services become available.
  • Fourth, an alternative cheaper or higher-speed internet-compatible medium becomes available, and the organization begins to install it.
  • Fifth, the proprietary network is rationalized out of existence as a cost-cutting maneuver, often because the internet protocols have such low percentages of overhead (i.e. wasted) data.

Around this time, T-1 “synchronous” networks became commonplace in the U.S. A T-1 line has a “frame” of 24 bits that repeats 64000 times per second. The first bit, calle the “sync” bit, was used to find the start of the frame. It alternates between 1 and 0. Customarily, a T-1 link is sent over a balanced twisted pair, isolated with transformers to prevent current flow. Each bit of a frame is usually used to send a single voice or data channel. The Europeans began to use a similar system (E-1) that sent bits as “octets” of eight related bits.

In 1982, the U.S. Congress deregulated the IRCs. They began combining to get economies of scale. All of their descendants offer voice, video and data services.

In 1992, computer access via modem combined with cheap computers, and graphic point & click interfaces to give a radical alternative to conventional telex systems: personal e-mail.

E-mail was first invented for Multics in the late 1960s. However it was limited to a single computer until the internet connected them around 1968. Various private networks (UUNET, the Well, GENIE, DECNET) had e-mail from the 1970s, but subscriptions were quite expensive for an individual- $25 to $50 a month, just for e-mail. Internet use was then pretty much limited to government, academia and other government contractors until the net was opened to commercial use around 1989[?]. Individual e-mail accounts were not widely available until local ISPs were in place, funded by people’s desire for web access. This was about 1992.

By using the time-shared systems almost end-to-end, the cost of data communications plummeted to less than 10 cents a message.

International Telex remains available via e-mail ports. It is one’s e-mail address with numeric or alpha prefixes specifying one’s IRC and account.

Telex has always had a feature called “answerback”, that asks a remote machine to send its address. If using telex via e-mail, this address is what a remote telex user will want in order to contact an e-mail user.

This is how smoke-signals became modern digital telecommunications.

This guide is licensed under the GNU Free Documentation License. It uses material from the Wikipedia.

Leave a Reply

Your email address will not be published. Required fields are marked *