From: gtoal@an-teallach.com (Graham Toal)
Date: Sat, 26 Mar 94 08:15:33 PST
To: cypherpunks@toad.com
Subject: Re: Digital Cash
Message-ID: <199403261653.QAA25502@an-teallach.com>
MIME-Version: 1.0
Content-Type: text/plain


: : Let me make a point here: I run an Amiga, and Amiga games do go 
: to town on copy protection. They generally rip the OS out of memory, and 
: install a custom one, do custom formats of the disk (as in the structure 
: is normally unreadable) and have further hard protection such as laser 
: holes, etc.

Listen, trust me, I know about this stuff, OK.  You're talking to the
guy who designed a locked-motor dual disk drive for doing direct analog
magnetic copies straight from head to head.  (When I was working in the
protection area for Acorn I mean - I was never a cracker, honest guv ;-) )

We used to hack our own protection at Acorn to see how good it was.
I've spent many contemplative hours pushing pins into floppies to
force bad sectors in exactly the right places...  I tell you though,
that sort of stuff is trivial compared to factoring the product of
two even medium-sized primes...

: : This is not kids' stuff - this is serious, and these people are 

Of course it's for kids.  I know these guys.  I never met one who was over 21.
It's just one of those hormone things that you give up when you discover
women.  (Or men as appropriate)  The only cracker I can think of who kept
up his skills after the age of 21 was Jeremy 'Jez' San, and he just did
it because he was working for Acorn on copy-protection schemes as a paid
job, which isn't the same thing at all.  Mostly they're around 18, plus
or minus.  A lot of the kids I knew who used to do this (back when I had
to keep up with the technology) were in high-school.  One of the best I
knew was a 14-year old.

: carcking it within hours. Let them loose on any program with their own 
: hardware, and they are capable of doing some quite serious things to the 
: rights of the author.

Sure, but what's that got to do with cracking crypto?  If you can't see
the difference between STO protection schemes and mathematically hard
algorithms, ...

Don't let your hero-worship blind you to reality.  No amount of skill
in disassembling or knowing the CRC tricks of a 1771FDC, or the six different
versions of an 8251, is going to help in breaking a secure crypto scheme
or something derived from one, like a secure e-cash scheme.  It's a
*completely* different set of problems.

(Oh - on disassembling - one of the tricks we used to get round the stuff
that was used to stop debuggers being able to place breakpoints was to
run in a completely simulated machine.  The program *could not tell*
that it wasn't running live on real hardware.  We'd do this stuff right
down to the level of scheduling the data arriving off the disks - some
schemes would do sneaky stuff like have the cpu execute a long stream
of INC instructions, and get hit by an int when the data arrived, and
use the value of the incremented register to check that the data had
arrived at *exactly* the right time conforming to the sneaky way it had
been written to disk.)

Acorn's R&D division knew what they were doing in the protection area
despite some laughable protection on the games side by Jez and the boys
from Acornsoft - we actually did a design review and costed out the
expense of doing our own DES chips for a dongle.  Fortunately we never
went down that route... (I'm not a fan of dongles as a substitute for
proper licensing arrangements)  We also came pretty close to putting
serial numbers in our CPUs, on-chip.  We decided against it in the end
for the same reasons we decided against dongles... it just caused *way*
too much customer ill-feeling. (For instance, if your CPU died and had
to be replaced, suddenly your mission-critical CAD package would stop
working...)

G
PS It's because we looked into doing our own DES chips in quite some
detail that I've always taken it for granted that Wiener's cracking
engine was not only possible but probably existed.  By our own back
of the envelope calculations in the early 80's, these machines were
feasible then.