Thinking About the Crypto Unthinkable

[Bill Stewart]

At 08:11 PM 08/26/2001 -0700, Tim May wrote:
>This said, I wouldn't advise _anyone_ to study "policy"
>(or its earlier incarnations, "Operations Research."
>"Systems Analysis," or the utterly execrable "General Systems," a la 
>Bertanlanffy).

Hey, I resemble that remark (Undergrad and Master's degree on Operations 
Research.)
Cool subset of applied mathematics - it touches on enough different fields,
including the algorithm-analysis stuff that overlaps computer science
and complexity theory, probability and statistics, simulation, scheduling,
inventory theory, graph theory, measure theory, abstract stuff like matroids.
Good for looking at systems design, and it worked well for me,
though you risk being too generalist and not specific enough at anything.

Unfortunately the whole field of Linear Programming changed just about
the time I left college :-), with Karmarkar's work showing that
LP could be done in polynomial time (though with a big ugly constant multiplier
that means that the theoretically-exponential Simplex algorithm tends to 
converge faster.)

There was work from operations research that was on the rather bogus side,
like the stuff that encouraged development of square pineapples because the
cans fit tighter on shelves than round ones....