grunwald@uiuccsb.UUCP (10/20/83)
#R:ihuxq:-35000:uiuccsb:3200013:000:1217 uiuccsb!grunwald Oct 19 16:29:00 1983 Hmm. I may be mistaken about the numbers. PLATO communications was designed around the concept of "clusters" (sites) of terminals. You have a large CPU -> site controller data channel but a drastically reduced site controller to CPU channel. It may be that the CPU->site controller is 1200 baud per station, and the site controller->CPU is also 1200 baud split between the 32 stations at a site. Thus, your actual keys/second can vary depending on site load. I do remember that it's very difficult to get more than 10 keys per second of reliable reverse channel traffic. This is not much of a problem when you're doing what PLATO was designed for -- CAI work. However, once you move into the more modern uses (local processing with data being sent to the CPU) then you really become limited by the reverse bandwidth. But that's neither here nor there. The point was to illustrate that it's possible to use commercial cable networks for data communication under certain conditions and that people are in fact looking into it. You'll still be tying up a phone line, so it's not much of a solution. Spoken : Dirk Grunwald University of Illinois USENET : ihnp4 ! uiucdcs ! grunwald CSNET : grunwald.uiuc@Rand-Relay