On 7/5/2006 at 5:18 PM michael_short at comcast.net wrote: processing. There conversion and small number of It was pretty surprising what a 1710 could handle; The 1712 Mux could handle about 300 signal sources/destinations. These could be analog, digital (output) or make-break (input). And the 1710 had a real-time clock, courtesy of the 1711 data converter. Interrupts were very primitive. "Terminal" usually meant a 1717 output printer with some sort of input device (usually a few pushbuttons, but a keyboard was an option IIRC). You could have up to 20 of those--and they had the bimodal alpha-or-numeric personalities that the other 1620 I/O devices did. The 1717 was an interesting bird in that you could write both numerical output and alpha output in the same operation--there were embedded "mode change" escape sequences. One significant difference between the 1710 and 1620 was the 1710 had the ability to test for its own hardware faults, instead of halting with a CHECK STOP. Practically, I don't know how much good this was, but one could test for parity errors and other faults. It's pretty amazing what what amounts to a 5-bit computer (I don't count the parity bit) running with a 100 KHz clock rate and doing its math by memory table lookup (well, the Cadet did, anyway) could do. Cheers, Chuck http://www.99-bottles-of-beer.net/language-ibm-1620-sps-916.html