Slashdot Mirror
Historic Microcomputer Restoration?
Pojodojo asks: "I am doing an independent study next semester with my computer science professor which we decided to call Historic Microcomputer Repair and Restoration. I will be working with such classics as the Altair 8080 and the Apple II. After I have repaired and or restored these machines, I will put them in a display for others to see. I have the opportunity for a modest budget to get equipment to put in the display, and would like to know is, what sort of things would you as fellow comp sci geeks like to see in a Historic Computer exhibit?"
.. I don't know the extent to which it fits your definition, but if I was to think of a bitchin' computer (insomuch as it could do some level of computing). It would be an Amiga 500, god I loved that... if you want something a little more in the line of "computer" I would say collosus, the original bletchly park beast... it could still out perform a P4....
*''I can't believe it's not a hyperlink.''
A DEC PDP-11/73, my personal favorite.
Probably the easiest computer to rebuild from the classic era as there is only one bus (Unibus), and nothing but traces and some very simple electronics on the backplane. Well that and you could hit them with a hammer.
The PDP-11 series, along with the PDP-8's were some of the first nodes on the ARPANET and you can still get working Ethernet adapters for them.
Hell, I still miss mine (Viper tape drive, RSX/11, RSTS/E 10, BASIC Plus2, 512MB EDSI drive).
(You can still find these things running if you look hard enough... (Try asking old medical/dental offices, most of them ran PDP/11's))
The Geek in Black
I know my BCD's (when I'm Sober)
I'm not sure whether this can be easily retrofitted into other computer designs, but one of the coolest things on the LINC--sometimes billed as "the first personal computer"--was the adjustable speed.
The LINC had a pair of dials on it: one was a (continuous) potentiometer, like a volume control, the other was a four-position "decade" switch. The pair of dials joint produced a signal that could be used to make any of a number of front-panel functions auto-repeat at a variable rate. In particular, you could make the "single-step" function auto-repeat. The pot adjusted the repeat rate continously over about a ten-to-one range. Each switch position was a factor of ten faster than the last. The slowest speed was about two per second.
This means that you could make the LINC single-step through its programs at an rate from about 2 to 200,000 steps per second... the later being about half of its full speed.
So, you could take a program... run it at 2 steps per second and watch the lights flash... then gradually speed it up over a five decades to 200,000 steps per second. At 2 steps per second if you watched closely from time to time you'd see one dot on the screen flash momentarily. As you sped it up the, the flashes would occur more rapidly... then you could see it was forming characters... then lines of text appearing at about the speed of a dot matrix printer... then finally a whole screen of flicker-free text.
Meanwhile, the LINC's speaker, attached to bit 6 of the accumulator, would gradually change from ticks to a buzzes to beeps.
I never saw anything that gave you such a feeling for just how incredibly goddam fast a computer was. Even one running at about 0.5 megahertz clock rate.
You actually could build a LINC from scratch, I suppose, since it was discrete components and the design was public domain. But it would be equally interesting to take a "stock" computer of almost any vintage and give it a continuously variable clock, a la the LINC.
"How to Do Nothing," kids activities, back in print!