For me it's less intuitive than a tree, but you can use a matrix to express all possible strategies of a two-player sequential game. This can help you visualize credible vs. noncredible threats from the second player. Ultimately I think the tree is more helpful in solving it visually, though.

Great explanation, thanks for posting that!

Technically, yes, but HN has an 80 char limit in the title. ;)

A vacuum sucks a lot more than Microsoft.

Google vs. Wolfram Alpha is something I'm ok with.

I'd rather type 2\pi than \frac{\tau}{2}.

Don't forget AKS, which is deterministic and faster than trial division. Can be a bit of a pain to implement though.

Another vote for AKS. Our cryptography lecturer skimmed over the Fermat and Miller-Rabin tests, then just said "AKS is better" and left it at that.. I believe the best approach is to do a few iterations of a non-deterministic test to quickly rule out some numbers and then start on a deterministic one. Wikipedia's description of the algorithm still looks pretty simple and being deterministic makes me feel all warm and fuzzy inside.

The documentation and download links for Plan at http://plan.dpk.org.uk seem to be broken so i'm not sure if this has bigints.

Having implemented AKS in Python and Sage, it can get pretty slow - obviously, nowhere near a good probabilistic test even to a staggering degree of certainty. That said, most of the algorithm is very quick. It might be smart to run its basic compositeness tests, do a pseudoprime test, and only do the polynomial mod operation if the number passes.

Plan isn't done yet. You can see the preliminary Ruby implementation at http://github.com/dpkendal/Plan, but I'm now working on a "final" version in RPython to be compiled by PyPy. The download and docs links are there to show what the website will look like eventually.

It should have bigints, though.

Elliptic Curve Primality Proving is way faster in practice, and although it's running time is not deterministic, it does give a proof that a number is prime.

Since Corporations are People, how would UMG be charged with a felony?

IANAL, but I would think the person who filed the takedown notice would be charged. AFAIK, even the automated mass-takedowns are in theory signed off by a lawyer. At least, there's someone named as responsible for it.

Actually, UMG could be charged; if it is convicted it would likely be fined.

(Proving that any given corporate officer had the requisite intent to commit a crime (the 'mens rea') is difficult to impossible, which is one reason why corporations can be tried in criminal court.)

Allowing corporations to be sued is exactly why corporate personhood exists, isn't it?

Not sure you could imprison a company though. That'd probably be the same as the death penalty.

Not that I don't want to support you guys for doing something awesome like this, but I wish the "Like" button showed appropriate meta-content on my facebook wall, instead of just a generic description of SendWrite. Also, donated. :)