From: Curtis Yarvin <Curtis_Yarvin@geoworks.com>
Date: Fri, 20 Feb 1998 11:16:28 +0800
To: ichudov@Algebra.COM
Subject: Re: Digital copy prot3ction
In-Reply-To: <199802200216.UAA09021@manifold.algebra.com>
Message-ID: <199802200259.SAA10250@ammonia.geoworks.com>
MIME-Version: 1.0
Content-Type: text/plain



> 
> I can hardly believe that any of these schemes are undefeatable.
> 
> As soon as the CPU starts talking to a video and sound board, 
> this whole thing becomes easily breakable. All one needs to do is
> to capture the signals that go to these boards and re-record them.

In general, there's no way of building a secure system that
prevents copying of information, but permits its consumption.
The two are too closely related.

The best you could do is a tamper-resistant hardware key on
the audio/video card.  (This locks you into a design where
the content is decoded on the card, which may be suboptimal.)
And anyone who can crack the crypto chip can get unprotected
digital copies and distribute them.  This is probably doable
by the same kind of people who set up pirate CD factories.

I'm sure Intel knows this.  I doubt the content companies do.
It's not Intel's goal to create an undefeatable protection
system; it's Intel's goal to convince the content companies
that it has done so.  This seems to have been achieved.

Curtis Yarvin
(not speaking for Geoworks)