Back when I was programming I used it when dump-reading. Say you have a printed stack of paper a foot and a half tall. You need to find a certain address in it, but you have no idea where it will be. So you split the stack roughly half way and examine the address – higher or lower? Then you split either the top half or the lower half (depending on which direction the address was), and repeat the process. It would only take 5-6 splits to find any address in the dump – very quick and efficient.

I moved the Speed slider to the left, to speed it up by shortening the delay between sorts.

I dragged “Sound Sustain” way to the right, near 30, to be able to hear distinct tones. Before that, it was more staccato machine gun bursts.

It’s fun! And highly educational, especially when you can change the shape of the input data to see how sensitive these algorithms all are to the starting conditions.

The QuickSort(ternary, LL ptrs), for example, tore through random input (accompanied by theremin-like sound effects). But when the input data was presorted descending, it turned to molasses, and–I’m not making this up–the sound effect sounds like the rhythmic groaning of Sisyphus pushing that boulder up a hill.

alcaray’s bubble sort is also slow, and makes a distinct moaning sound over sorted-descending input. Radix Sort(LSD), which I hadn’t heard of, turns out to tackle both sorted and random data with alacrity, and suitably enough, with a jazzy ray gun sound track.

Thank you Timo for writing it, and ER for linking to the video.

I looked up the “bogosort.” The example given was sorting a deck of cards by throwing it into the air, piling it up, seeing if it was sorted, and if not, throwing it up in the air again. You’d wait a long time for that one to finish (8 * 10^67 combinations).

When I learned how binary sorts worked, I remember telling myself “Damn, that’s really clever. What will they think up next?”

Sorry about that.