From: Mu In Taiwan (mu.in.taiwan@gmail.com)
Date: Mon Oct 12 2009 - 22:47:15 MDT
>> The halting problem is not what you think it is. As far as I am aware,
>> is quite possible to show that *some* algorithms will or will not halt
>No shit Sherlock
1. Well, at least you are beginning to acknowledge that other people's
criticisms of what you are saying, are correct. A form of progress.
> if it were not true *all* algorithms would be useless.
2. Absolute rubbish. There are algorithms and heuristics that are used
despite our inability to predict their run-times accurately. The utility of
an algorithm is only partly defined by our ability to predict whether or not
it will halt.
Take the field of constraint satisfaction programming. It is fairly well
known among researchers that problems in that field tend to have the
property of being either a) easy to find answers to, b) easy to show that no
answer exists, or c) A complete nightmare that will take a hideous, and
perhaps effectively incomputable, amount of time. The problem is that we
have no effective way to distinguish between these cases a priori other than
some very broad heuristics. Guess what? This technology is used for
scheduling problems every day, all over the world.
3. I note that you have once again ignored a call for definitions of your
'fixed brains' 'non-fixed brains', etc. You also ignored most of the points
made by your critics in recent messages.
Mu
p.s.
4. I note that your claims are not even consistent between consecutive
posts:
your reply to me: "What Turing proved is that in general there is no way to
prove that any
given algorithm will halt."
your immediately subsequent reply to Robin: "what Turing proved is that in
general the only way to know what a random program will do is to watch it
for eternity and see."
These claims are inconsistent ("any given" != "random", "prove" != "know,
watch for eternity and see").
5. They're also both wrong.
What Turing actually proved, as far as I am aware, was that *a* general
*algorithm* cannot solve the halting problem for *all* possible
program-input combinations.
Note my use of emphasis on "a" "algorithm" (here, equivalent to 'turing
machine program') and "all".
Mu
This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:01:04 MDT