Wednesday, December 27, 2017
Why don't computer languages handle proportions?
In handwritten algebra, and on a slide rule, you can deal with proportions like
x/3 = 5/y
This situation is extremely common in everything from electronics to cooking. You can vary EITHER X OR Y and see how the other number changes. Aside from proportions, the essential idea of an equation is a balance with both sides able to vary. Change one side and the other changes appropriately to maintain the balance.
You can't do this in any computer language that I'm aware of. Way back when I first learned Fortran in 1966, I had to adapt to the idea that the equal sign is always a one-way street. You have to rerig all proportions and equations so one variable is the output and everything else is the input. I adapted but never bothered to ask WHY.
In 1966 processors and memory were extremely simple, but those limits are long gone. Now it would be easy to write a parser or compiler that treats an equation as a balance. Why not?
= = = = =
Analog computers were based deeply and totally on the concept of balance. Start with a differential amplifier, which is the electronic equivalent of a classic balancing scale. Apply an input to one side, then either manually or automatically adjust the other side to reach zero output. The result you read is not the output; it's the adjustments of the other side. The output is just the equal sign or the fulcrum. 
Labels: Asked and unanswered
¶ 3:36 AM
