Scales for building your own slide rule (1999)
sliderulemuseum.com
sliderulemuseum.com
I like the footer:
>In the beginning, at the time of the great flood, Noah went thru his ark after it landed, and found two small snakes huddled in a corner. Noah looked at these poor specimens - and said "I told you to go forth and multiply - why haven't you?"
>The poor snakes looked up at Noah and replied "We can't because we are adders....."
>Noah looked a bit perplexed, and then proceeded to tear bits of planking from his ark. He went on to build a beautiful wooden platform. He gathered up the snakes and placed them on the platform, and joyfully told the snakes - "Now go forth and multiply, because even adders can multiply on a log table"
I wonder how much error is introduced by storing these as pixel-aligned GIFs?
Also, the error for a pure monochrome version (that doesn't do subpixel resolution) would be at most 1/1400 pixel. Assuming a 1:10 scale for that, the error would be at most 10^(1/1400), or a factor of about 1.001646.
(Of course, that would increase once you start using the slide rule)
I was never practiced enough to be good at an abacus, but learning to use a slide rule is pretty easy. First, learn how logarithms can be used to do computation. Back then every engineer or even science student, as I was, learned to use tables of logarithms to perform multiplication, division, and exponentiation rapidly. Today this is a bit of a lost art, but if you understand it, the operation of a slide rule becomes obvious. The most basic scales are just marked off logarithmically. The scales are a compact version of the log table values. Appending logarithmic scales corresponds to multiplication and so forth. It's like learning vim, learn a few basics and then tinker around with it to get better.
The great multiplicity of scales on slide rules from that period also provided access to trig values, natural logs, hyperbolics, square roots (same log scales but scaled by half), etc.
For many problems, one could approach three significant digits of accuracy with a slide rule. The scales didn't have to be straight they could be circular. This led to a circular slide rule, less convenient to hang from ones belt in a holster, but with some larger (and some smaller) resolution scales. I have even seen cylindrical slide rules with helical scales wrapped around the surface of a cylinder. These could obtain an extra digit or more of accuracy.