new algorithm for sequencing integers that is asymptotically exponent…

…ial but does not repeat values.

This was motivated by realizing when debugging the overflow issues (and adding an assertion) that r2 was a NaN for the very_slow test. This was because while a few samples were taken, they all had n=1, so there was no slope to be computed. I wanted a sequence such that even slow computations could find some sort of a slope, which really should just take two: one with n=1 and another with n=2.

The SlowFib is somewhat complex, but I wanted something that would eventually increase exponentially (and not too fast, not doubling each time), but that did not repeat and started with just 1. And this is what I came up with. I don't know a name for the SlowFib sequence.

Interesting. This seems like a good idea.

Sorry for the radio silence on this. The reason I haven’t merged this yet is that the very_slow test seems to be failing on this branch, and I haven’t had a chance to investigate yet.

I'm afraid I can't see how to help, as the very_slow test does pass on my machine. And no worries about being slow to respond!