[bitc-dev] C interfacing vs. portability

[Jonathan S. Shapiro]

On Thu, Feb 26, 2009 at 10:20 PM, Geoffrey Irving <irving at naml.us> wrote:
>> Though in practice O(N log N) *is* O(N). N in representable program
>> states is bounded by the total number of electrons in the universe,
>> and logN of that is surprisingly small. I have annoyed several
>> algorithms faculty with this observation over the years.
>
> log N is about 10 or 20, and I don't think assuming 10 = O(1) is a good idea.

I really need to introduce you to one or two algorithms people who
don't adequately appreciate the relationship between naive order
statistics and stand-up comedy.

shap