Sub-poll!

• 0E0
• -1
• 2147483647
• -2147483647
• π or e or sqrt(2)
• i
• -Inf
• NaN

Apparently, the totalOrder predicate from IEEE 754-2008 specifies that -NaN < -Inf < negative finite numbers < -0 < +0 < positive finite numbers < +Inf < +NaN.
So if floating point values are allowed, then I think that "-Nan" would be the obvious starting point ... coz you can't not get no lower than that ;-)

(Hopefully, indexes have to be real integer values.)

Cheers,
Rob

"Hopefully, indexes have to be real integer values."

I don't see that they have to be "integer" values, but it gets tricky for current hardware if they aren't limited precision rational numbers. In fact even "integer values" is a deal breaker if they aren't limited precision.

Optimising for fewest key strokes only makes sense transmitting to Pluto or beyond

• π or e or sqrt(2)

Close: my answer would have been the first index is Φ ≈ 1.618, second is e ≈ 2.718, third is π ≈ 3.14, and so on, picking irrational (and preferably transcendental) numbers whose integer portion is the next counting number. (I had forgotten that Φ, like √2, is irrational but algebraic, and had started to write that as "picking transcendental numbers" before having a long think about it; still, I prefer Φ since it has a predefined symbol.)

Ooh, another fun set: ln(2) ≈ 0.693, √2 ≈ 1.414, 2**(√2) ≈ 2.665, ... I'm sure someone could come up with a whole series of those whose integer portions are increasing, but all the terms only combine operators and 2. ;-)