Communications Computers Data communications over telegraph circuits 1945
Communications Computers Data communications over telegraph circuits 1947
Jim Haynes I'm just guessing, but I don't think the Teletype Corp had that much electronics capability...
Data communications over telegraph circuits 1948
That was all pretty much worked out by the mid-1930's or so! Some of the test equipment developed post...
I think the definition is easily blurred. Aren't all routers today basically computers designed for communications?
I suspect the trend has been that instead of gear with internal logic designed specifically to do certain things the gear today uses microprocessors or its own general purpose processor and is then programmed to do communications tasks. Obviously it will have some hardware and instructions to facilitate that type of work.
Communications Computers Data communications over telegraph circuits 1946
Well I meant to limit it to the circa 1960 time frame, when there were computers being built specifically for data switching...
In other words, you probably could take a modern ESS box and reprogram it to do general purpose computing work. It might be rather limited and a lot of work, but it could still take some input, slosh it around and calculate, and put out results. In contrast, I don't think you could do that with an early ESS box even though it was programmable--it was too dedicated to a narrow range of tasks.
As you noted, in the history of computing, there was always the blend of doing comm work as part of the main system or farming it off to a separate box. You noted the IBM S-360 and later devices like the 3705. Further, the end terminal or terminal controller could have some intelligence of its own.
Another issue is offering simply a connection between two points or storing and forwarding information. Western Union's computers had to store and forward information in some applications. (Their original 1948 modernization simply used punched paper tape to store a message until it could be routed and forwarded). I wonder if modern switches temporarily store--even for only a fraction of a second--a stream of bits before sending them onward as a result of the sharing of multiplexed facilities.
Generally we today take a huge amount of behind-the-scenes work for granted when we use computers. Even on my old BBS reader, this message would have all sorts of ASCII control characters to direct Procomm-Telix to paint the screen. Now on a fancier setup, there are all sorts of stuff going on to relay my typing to all who are reading it.