Re: [boost] [geometry] robustness approaches

Subject: Re: [boost] [geometry] robustness approaches
From: Arash Partow (arash_at_[hidden])
Date: 2009-03-17 09:39:59

Hi Luke,

I really appreciate the effort you're putting into this library,
just skimming over the code and I can see it will be something
I'd definitely be using.

But in regards to some of the comments you've made about robustness
and such, I would like to suggest you have a read of the paper:

"Robustness In Geometric Computations" Christoph M. Hoffmann 2001

There are some geometric software design concepts such as "folding"
(not protein folding" etc) that really need to be taken into account,
as most geometric computations are layerings or repetitions of each
other, understanding these concepts makes suggestions like epsilon
tolerance based robustness approaches inadequate, but further more
reliance upon type precision essentially becomes a bottomless pit
of problems.

I have some other comments regarding interfaces for the library
that are concerned with online algorithms, but I'll leave that
for later as I'm still trying to determine some working examples.

But again as before, thank-you for the effort!

Arash Partow
________________________________________________________
Be one who knows what they don't know,
Instead of being one who knows not what they don't know,
Thinking they know everything about all things.
http://www.partow.net