Communications Computers Data communications over telegraph circuits

I'd like to fork off a topic on this discussion, that of computers designed for communication purposes. There was a fairly brief period when these machines were being built. I'm thinking of the G.E. Datanet 30, the IBM 7740, and other computer companies had their offerings. One of the early applications of the Datanet 30 was to replace the Bell 83B1 system controller (a polling device) while keeping the Teletypes that were the out stations. The Datanet 30 has a curious history (see book "King of the Seven Dwarfs" by Homer R. Oldfield). It could be connected to a big (for the time, and physically) disc system to store messages. And then the disc system could be dual-ported to communicate with a more business-oriented computer. This was the basis of the original Dartmouth time sharing system. The Datanet 30 continued as a front-end for G.E. computers into the 635 line. By then it was quite long in the tooth. (We joked that we would have to quit making them because we couldn't buy transistors that bad anymore.) And then one of G.E. corporate activities decided that the Datanet 30 was a communication product and should be transferred from the computer department to the communication products department in Lynchburg, VA. There was a newer Datanet 355 that was awfully high priced and probably didn't sell very well - I know Dartmouth didn't use it, and instead used the Honeywell 316 as the communication front end for their later time sharing system, which continued to use the big G.E. mainframe.

IBM had the 7740 for a while, I guess aiming at the same market as the Datanet 30. But with the introduction of System-360 they went to a wired- program communication front end. The concept seemed to be that by then the communications system was an integral part of the computer system - applications were online - and that there was no further need for a free-standing communication controller for message handling. There were several third-party programmable front-ends for the 360 line; and then IBM went to a programmable device later on.

Western Union seems to have worked with a variety of computer makers. There was RCA making the specialized computers for AUTODIN, and there were two generations of those. Univac had several models that were used in communication switching, including some military models we usually don't hear about. There are pictures in WUTR of various computers, one with a Western Union nameplate on it but almost surely built by somebody else, but I haven't been able to identify it with a particular maker. And they built some of their own communication hardware to attach data lines to computers.

I'm not aware that Burroughs did much in the field. They had a couple of wired-program communication machines used with their business computer systems, and later a programmable communication processor. Control Data had something, because I have a piece of junk in my Debt Collection that is a serial-line adapter for a CDC machine, but I don't know for what.

Then the original ARPANET was built using Honeywell 516 machines, if I remember correctly. The 516 was a result of Honeywell acquiring Computer Control Corp. --

jhhaynes at earthlink dot net