True. If a cracker has access to your executable, it WILL be cracked
given enough time and motivation.
Instead of coming up with some brilliantly obtuse crack-detection
scheme, the realistic goal of copy protection should be to delay the
inevitable crack as long as possible, especially since the majority of
your sales will probably happen in the first two months.
Looking at it this way encourages you take a holistic approach to copy
protection; for example, sometimes social engineering is even more
useful than real engineering to delay the appearance of a crack. For
example, you could release your own flawed 'crack' with the game shortly
after ship. If you make sure that the flaws aren't immediately
noticeable, you'll probably delay crackers from starting work on your
game for a while. Similarly, add crack detection routines to the game,
but don't immediately halt the game upon detecting the crack. Instead,
subtly alter the game so it is unbeatable. The Game Developer article
for Spyro: YOTD mentioned that they actually removed some of the 'eggs'
you needed to collect to beat each level, so you couldn't get very far
in the game on a cracked version. This works because most crackers
don't actually play through the game to test the crack fully.
If you have several layers of such protection, you'll cause the crackers
a lot of false starts, buying your title more time on the shelves
without a working pirated version.
[mailto:gdalgorithms-list-admin@...] On Behalf Of Jon
Sent: Tuesday, January 28, 2003 2:39 PM
Subject: RE: [Algorithms] An intelligent analyst
> I have a killer (simple) idea for copy protection but i need a little
> perspective first :)
The hardest copy protection I've seen so far relied on the fact that
the I-cache is not synced up to the D-cache unless explicitly flushed.
But, whatever you do, it's likely that the protection will be broken
at some other level. The harder you make it, the more of the crackers
will try to break it. They are more plentiful than you are. Because
the machine is in their hands, and they have logic analyzers, they
WILL win in the end. The best copy protection is to make sure you're
satisfying paying customers.
This SF.NET email is sponsored by:
SourceForge Enterprise Edition + IBM + LinuxWorld = Something 2 See!
GDAlgorithms-list mailing list GDAlgorithms-list@...