Data communications over telegraph circuits 1896
Data communications over telegraph circuits 1902
Floyd L. Davidson Fax is very useful for things e-mail cannot do, such as to transmit documents. Sophisticated users might have ways to encode a document as a...
I think the fax image was transmitted with gray levels, even if the resolution was not very good. The transmitter used a chopper wheel to generate 2500 Hz. AC from the scanner signal. They do say though that the signal is substantially black-and-white just because of the nature of the thing being scanned. (I'm cheating by reading from a copy of a Desk-Fax specification right now.) Says it takes about two minutes to transmit the message. And that the circuit must have a loss no greater than 25db at 2500 Hz. Also points out that for Desk-Fax the customer's transceiver has to be on the same power system as the telegraph office; they used synchronous motors and relied on them to stay in phase once they initially got into phase. Notes that the image from the scanner is negative - light reflected from the white background and no light reflected from dark areas - and has to be inverted when applied to the recorder so that there is no voltage for the background and a voltage to make marks on the recording paper.
Data communications over telegraph circuits 1903
In our company, everyone (sophisticated or otherwise) has convenient access to Lexmark multifunction machines (I forget the model # at the moment, and I am typing this...
Typically 12 to 16 channel frequency division multiplex over a voice grade line. They got 4 or 6 channels using time-division multiplex over a grounded telegraph line. Coe's book says that Postal Telegraph built voice grade lines early on, as they wanted to use a "harmonic telegraph" which was an early frequency division multiplex. A. G. Bell was working on a harmonic telegraph when he made the discovery that led to the telephone; his compebreastor Elisha Gray went on to perfect the harmonic telegraph.
Data communications over telegraph circuits 1899
Jim Haynes Interesting point. I sensed my time-sharing systems were locally based in cities. In 1970 there was a...
Data communications over telegraph circuits 1900
He was being *kind*! It was actually a lot worse that that. All true. However... none of that was expected to be shared or to interact...
Data communications over telegraph circuits 1901
The Internet didn't exist yet. Low cost voice calls were what eliminated the Telegram, and started the decline of TWX-Telex, before TCP...
Early on there were no modems to work on the switched network; if you wanted to use a modem you had to order a voice grade private line and supply your own modem, or order telegraph lines and the telephone company supplied the modems. AT&T had a very strict policy of "no foreign attachments" to the switched network.
Then in the early 1960s they started marketing "data sets" which were modems you could lease along with telephone service and connect your own data equipment. The RS-232 standard came out of this development, the need to have a standard interface between the telephone company modems and the customer's equipment. They also at the time of the TWX cutover to dial started using modems and the switched network to carry the TWX traffic. That's pretty much the way business computing was done in that time frame.
They used private lines and a multiplexer that could run a bunch of SABRE terminals over a single voice-grade line. More examples of the government trying to keep W.U. with exactly one foot in the grave.
They introduced a "hot line" voice service, where the subscriber would pick up a telephone at one location and it would automatically ring the telephone at the distant end. I don't imagine there was a large demand for that kind of service. Another goofy idea was a switched data service where the customer could choose the bandwidth per connection and pay accordingly. I call that goofy because a lot of data equipment could only operate at a fixed speed anyway; and it's hard to expect the customer to do a cost-benefit analysis to choose the bandwidth each time they want to do a data transmission. And then the infrastructure had to support the highest possible data rate even if the customer rarely used it, and the switching was further complicated by the variable bandwidth.
IBM was famous for its "muscular" sales force, but it was also notable for the quality of its manufacturing engineering. E.g. magnetic core memory was a cash cow for IBM because they were stringing core planes by machine while their compebreastors were doing it by hand or sending it offshore to be done by hand. And when you look at some of the old IBM products you see how finely they were designed for low-cost quanbreasty production, while competing products were obviously more costly to make.
Data communications over telegraph circuits 1898
Remember that through the 1960s and into the 1970s there was an awful lot of computer time sharing being done with the Teletype Model 33 as the terminal. Of course time sharing is not...
Of course there are some artificialities built into pricing on both sides. The telephone business used profits from long distance to subsidize local service in places where it was costly to provide. And the telegraph business needed profits to subsidize their unprofitable offices. Yeah, I wish there was some place you could go to read a good analysis of what sent W.U. down the drain. The quick answer is that they spent more money than they took in, with no reversal of that situation in sight. But then we technical people tend to think more in terms of particular product and service offerings that seem unwise.
jhhaynes at earthlink dot net
Data communications over telegraph circuits 1897
Jim Haynes I found some posts you made to the Telecom archives with Business Week articles from the...