Some corrections to the "worst input" msgs

[Tim Peters]

A "0" and a "1" got interchanged in the 4th columns of the last two
lines of:

> If we shift our view of these troublesome cases "to the left" by one
> bit, it promises to make life a lot easier:
>
> xxx1 | 0000000...0000001yyy  
> xxx0 | 1111111...1111110yyy
> xxx1 | 0000000...0000001yyy
> xxx0 | 1111111...1111110yyy

The ASCII art should read

xxx1 | 0000000...0000001yyy  
xxx0 | 1111111...1111110yyy
xxx0 | 0000000...0000001yyy
xxx1 | 1111111...1111110yyy

instead.  Blame it on Emacs's flawed "M-x copy-4-lines-but-shift-them-
left-a-character-position-across-the-first-vertical-bar" command <ahem>.

Similarly, the "M-x adjust-asymptotics-appropriately" command left out a
power of two in:

> ... the run-time in practice appears to be O(log(c)), and I suspect it
> would be pretty easy to prove that O(log(c)*log(log(c))) is a bound on
> the worst-case time (I'm just counting bignum operations there ...

While the worst-case time may indeed be O(log(c)*log(log(c))), I don't
in fact suspect it would be easy to prove that; I did & do suspect it
would be easy to prove that O(log(c)^2*log(log(c))) is a worst-case
bound.

hoping-we-can-all-sleep-better-now<grin>-ly y'rs  - tim

Tim Peters   Kendall Square Research Corp
timaksr.com,  ksr!timaharvard.harvard.edu