> Our understanding of how computers work, on the other hand, is pretty
> solid. We designed them from scratch based on simple axioms. If your
>axioms are true in the real world, then your proofs reflect real-world
> truth as well. When we prove something uncomputable, we really know
>that you can't compute it on our computers.
In my experience, the axioms are never completely true in the real
world. I have had the following discussion with a client on a numerous
occasions:
for (i = 0; i < 20; i++)
{
Client: I want feature X
Me: I cannot get it done within a reasonable time.
}
Finally, I get tired of this and start probing the client with
questions. Why do you want this? Is this assumption correct? What
about this one? Will the following work?
Then I realize my axioms of the real world were wrong. The problem the
way I understood it was indeed unsolvable within the given
constraints. However, time and again I would find a way to give my
client what he wanted with less than 50 extra lines of code.
--
Sasha Pachev
AskSasha Linux Consulting
http://asksasha.com
/*
PLUG: http://plug.org, #utah on irc.freenode.net
Unsubscribe: http://plug.org/mailman/options/plug
Don't fear the penguin.
*/
