RPN calcs should be a teacher’s recommendation

[Garth Wilson]

I express one of my frustrations in this paragraph of the RPN operations page of my treatise on 6502 stacks ("stacks" plural, not just the page-1 hardware stack):

The argument in school calculators is "Algebraic lets you enter the equation just as you see it on paper!"  Well, for one thing, a lot of programming work falls outside of actual equations.  For another, even in the field of calculators, in real life we don't usually have an equation in front of us.  I think through it, "Let's see—I need A, and then square that; now take B, multiply it by C and add that result to the earlier one.  Now I need D raised to the power of E, and divide the earlier result by that.  Done.  Oops, no, I still need to take the log of that..."  That's the way much of real engineering is.  It's only in school that you get canned problems and have the equation in the book and you're supposed to just drop the numbers in the chute and turn the crank—then we wonder why we get graduates who passed all their classes and yet show a disconnect between that knowledge and any understanding of what the problem is in their circuit on the workbench!  I've hired a lot of electronics technicians and a few engineers, and I gave all the applicants a circuit-analysis test with a dozen simple problems.  Not one of them ever got it all correct.  It was kind of frustrating.