Data communications over telegraph circuits 1894

That was all they intended it for. A way to get a telegram to-from the nearest WU office. They were looking for a cheaper, smaller subsbreastute for teleprinters that they were then putting in customers' offices.

Well maybe I'm wrong on this, but they didn't have any nationwide network of voice-grade lines to connect fax customers to one another.

They didn't to my knowledge digitize images at all; they keyboarded the text of the message received by DeskFax, turning it into Baudot to go over the network. If you try to forward them as analog faxes you have the problem of picking up noise and phase distortion along the way. It can be solved - they did put in some long distance private fax networks such as the weather map network for the Air Force.

It seems goofy to me, because you have to have a location where the machine won't be robbed and vandalized, and you have to buttume a clientele who don't have telephones or they would just phone in their telegrams. Or for that matter, couldn't you go to a pay phone and phone in a telegram and pay for it through the phone?

I suppose that never got any farther than the Philadelphia field trial.

I doubt that had anything to do with it. Remember that until Carterfone if IBM supplied the modem it would have to go on a private line. There were the IBM card transceivers with builtin modems. IBM was not exactly a leader in online computing. There was SAGE, but I believe AT&T supplied the modems for that system. The book is fairly recent, but of course anything dealing with telegraph history is going to have a lot of 19th century stuff in it. Coe does cover things up through much of the Postal Telegraph era and up to WW II. They did get into that; both W.U. and AT&T had switching systems for private corporate and government networks. And WU put in a transcontinental microwave system; but that gave them a lot more bandwidth than they could use. For some reason they never got the television network business. Maybe they didn't run microwaves to where the television network stations were located. It's interesting that in the mid 1950s the three military services went in three different directions. The Air Force leased Plan 55 from W.U. and made some use of G.I. maintenance. The Navy leased the 82B1 system from AT&T. And the Army went with a purchased military-nomenclature (AN-FGC-30) switching system made by Automatic Electric and Kleinschmidt. These all had to inter-operate. Ultimately the Defense Communications Agency took all that over and made all the services use AUTODIN, which had started out as an Air Force network and which was supplied by W.U.

Now if W.U. had used their microwave network to do what MCI did they might still be with us. But MCI had to go through a bruising legal battle with AT&T before it prevailed. (Of course today's MCI is nothing like the original concept of selling microwave bandwidth only on high-traffic routes and using Bell lines to connect to their subscribers.)

I've argued that Telex was probably a mistake for W.U. It put them into head-to-head compebreastion with AT&T in a field where AT&T owned all the local loops needed to extend service to their customers. It required a big investment in new teleprinters and new switching equipment. The switching equipment was electromechanical and thus nearing obsolescence. And AT&T could carry TWX traffic on their switched voice network for only the marginal cost of the extra traffic; whereas W.U. had to build an entire network dedicated to Telex. And they couldn't profit from the worldwide nature of Telex because they had been forced to divest their cable operation.

--

jhhaynes at earthlink dot net