1974 Sinclair Scientific Calculator Repair and Restoration
retrorepairsandrefurbs.com
retrorepairsandrefurbs.com
http://files.righto.com/calculator/sinclair_scientific_simul...
(which was discussed previously https://news.ycombinator.com/item?id=6302364)
The ability of the designers to do more with less is reminiscent of Steve Wozniak's work on the Apple I/II and other projects
I would argue against this statement which I’ve seen many times. RPN is not more difficult to use, it is simply different from what people have been using for the last few decades. Not being used to an human-machine interface and having to learn it should not be confused with difficulty of use.
Someone used to vi/vim has all the ease in the world to use it, but it might take a newcomer more time to adapt.
RPN has a very natural data waterfall flow, especially when stacks can grow to multiple levels. Of course you don’t have a direct view of it, but it becomes second nature. You simply know what that the higher value in the stack is your previous result.
I think it would be easier for a child to parse Lisp with rainbow parens than learn all those rules of order of operation most students find painful.
Newer calculators have "natural textbook entry", where they show the equation in mathematical notation as you enter it, so no stack required in your head. The entry method (shown linearly) for division is: <divide> numerator <down> denominator <right>. The keystroke count is lower than it would appear because you can leave out all of the right-most cursor keystrokes.
The method could be further generalized to eliminate the need to know about precedence on entry. For example ab(c+d+e)*f could be: <mul> a <next> b <add> c <next> d <next> e <done> <next> f. Visually it could show the order of operations either with precedence, or in some new way (perhaps text size changes to indicate order).
It's interesting to think how programming languages could look in the post-teletype era.
Also, RPN ftw!
They initially had problems with battery life, but took advantage of the fact that they could strobe power to the LEDs and chips without completely compromising functionality to reduce the power consumption by a factor of 100. That allowed them to use hearing-aid batteries.
My intuition is they were drafted by hand and reduced repographically based on the age and simplicity…which I mention because a pencil and a scanner are still an option today.
My father wouldn't buy anything he didn't think could be reasonably repaired. Hell--it was a selling point most people took for granted.
From tv sets to most calculators, and even early computers, most had schematics available from the company, and the boards were laid out with repair in mind.
My old HP 454a Oscilloscope was designed to be repaired.
I look at these old circuits and they are art.
I once heard an old Electrician say, "The art of our trade died when we stopped sweating wires."
This is like the Casio version of HP-41- alphanumeric keystroke programmable with cassette interface.
No plastic heatstakes (unlike HP), but you're hosed if the flex cable to the LCD is broken.
One of the most impressive things recently is the guy who made replacement LCDs for the classic Radio Shack / Sharp pocket computers (they don't use flex cables). That a hobbyist can do this at reasonable price is a sign that we are living in amazing times.
https://www.tindie.com/products/halfbakedmaker/lcd-replaceme...
However the LED display was also the main reason why this calculator needed 4 AAA batteries (which I imagine didn't last that long), while later calculators with "boring" LCD display could run for years on fewer batteries. My father had an early electronic watch with an LED display, which had a button you had to press in order to read the time - kinda like a smartwatch...