Re: About those lights
"The original bulbs appeared to fail more frequently at that altitude than at sea level."
Although the tiny pea bulb failures were a fact of life on the ICL System 4-72 mainframe - I never noticed any particular difference at 1500m in Pretoria.
The second generation technology EELM KDF9 prototype had had no engineers' panel because "these mainframes will be so reliable". They also said that about not needing parity on the ferrite bead memory. The production design had to add a small engineer's panel.
When the engineers moved on to design the third generation technology System 4-70 series they made sure that there was a glorious cpu panel including a 36x10*** matrix of green, red, and yellow illuminated squares.
There were also 109 illuminated push buttons - and a total of 62 lever switches in four rows to cater for setting 32 bit address/data words. People who used the panel regularly developed a callous on the top edge of their fingers from clearing all the switches in one long sweep. Oh - and there was a rotating knob for some forgotten purpose.
The whole display was about 2x3 feet (60x100cm) on the end of a standard 6' (1.8m) tall cabinet.
IIRC the incandescent bulbs of the matrix were multiplexed fast enough that the eye couldn't see the flicker.
***The matrix was definitely 36 columns - but probably bigger than 10 rows. I've just counted as best I could. The home movie image is a vertical pan shot lasting only a few seconds - and unlit rows are subjective in the blank space. The (by then rather dusty) Standard 8 film was eventually transcribed to VHS tape - which in turn was eventually transcribed to DVD.
Biting the hand that feeds IT
