Re: [computer-go] how to use GTP in place of GMP

[Richard Brown <rbrown@xxxxxxxx>]

Jim O'Flaherty, Jr. wrote:

> What's to say your DB can be loaded into memory as a RAM drive, in your
> case, reducing your I/O bottleneck substantially?  Then the penatly you
> currently suffer would disappear, right?
Yes, that might actually be true in my case.

But what about the program whose yet-to-be-written algorithm takes a long time?

I know that many of today's programs do enormous amounts of lookahead, or
"reading" in a very short period of time.  So, there are well-written programs
that already exist, which don't take very long to play a move.  I grant you that.

But suppose that the _winning_ (as I said, yet-to-be-written) algorithm
simply takes longer than 250 ms to arrive at the hand-of-god move?

Suppose it takes 25 seconds.  Is that to be allowed?  Thirty seconds?

I just don't think that using the amount of time the program takes to
respond is a good measure of whether or not it's suspected of "cheating".

It appears to me that the only way to prevent such cheating attempts is
to require the programs to be physically present at the tourney site.
(You could still use networking _software_ to hold the tournament, on
a private network of course.  Nothing wrong with sockets, per se.)

But to prevent human "ringers" from playing for the program, I just can't
see any way around inspecting the hardware, to determine whether there's
a wireless card in the box, and if there is, disallow that hardware.

I admit I haven't really thought very long or very deeply about this.  It's
been a long week, and I'm tired.  Maybe there is some sort of digital signature
trickery that can be used to prove that a program is who -- rather, what --
it claims to be, even across the globe, networked from afar.  But someone
besides me is going to have to figure that out, if so.  I just don't see it.

--
Richard L. Brown             Office of Information Services
Senior Unix Sysadmin         University of Wisconsin System
                            780 Regent St., Rm. 246
rbrown@xxxxxxxxxxxxxxxxx              Madison, WI  53715

_______________________________________________
computer-go mailing list
computer-go@xxxxxxxxxxxxxxxxx
http://www.computer-go.org/mailman/listinfo/computer-go/