Washing machine drives

Will wrote: Don't know if sequencing was the norm for CDC machines of this vintage or not. At Tek, there was no automation that powered things up in the correct order. The motor-generator in the basement was left running all the time, but during non-operation times, everything else was powered off. I know it made a big difference in the electricity bill to power down during off hours versus leaving the system running. The room also had two Modcomp communications processors that served as the serial front-ends for the Cyber, and a VAX 11/780 that ran VMS during the day, and Berkeley Unix (4.x) during the night. The Modcomps and the VAX ran all the time. The Modcomp and VAX serial lines hooked to some kind of port contender that had a massive panel of serial connections patched into it for all of the terminals scattered around the campus. Also in the room was a huge Calcomp flat-bed plotter that was an offline device. It had a 9-track tape drive attached to it, so any plotter runs required that a tape be cut, put on the Calcomp, and then run. The plotter used pressurized ink (red, green, blue & black), which was a bit of a mess to top off. It made beautiful plots on Mylar, though. I still have a "spirograph" plot that I made with a Fortran program on the Cyber. I think that there were a number of PDP 11 systems in there too...I think that there was an 11/70 or two, and maybe an 11/45. I didn't do anything with the DEC stuff, my job centered around the Cyber. IIRC, the sequence for bringing up the Cyber was: First the CPU chiller was powered up, then the disk drives, then the printer & controller, then the magtape drives (one 7-track and three 9-tracks) and controller, then the punched card reader/punch and controller, then the CPU. The boot magtape was mounted, the deadstart panel checked to make sure that the bootstrap routine was correct, then the coldstart button (located in the center of the console (with the two round vector tubes), beneath the CRT's) was pressed. If all went well, the tape drive ran for a bit, then the boot pack would hammer for a while, and then KRONOS would be running. There were a few operator commands necessary to put the machine into multiuser mode, and that was it. All told, the process took about 45 minutes or so. Funny how people complain today when their computers take a minute to boot up. Rick