Re: DRMTICS 2005 Call for Papers

Francois Grieu wrote:

> In article <7xd5rd2fkl.fsf@ruckus.brouhaha.com>,
> Paul Rubin <http://phr.cx@NOSPAM.invalid> wrote:
>
>
>>Bryan Olson <fakeaddress@nowhere.org> writes:
>>
>>> > Note: the above simple scheme is very vulnerable to a
>>> > collusion of users buying the same music
>>>
>>>Ay, there's the rub. [Hamlet III,i,65]
>>
>>There's an even worse problem, which is that it depends on the vendor
>>knowing the identity of each of the legitimate buyers. That alone is
>>enough to get lots of people to not buy the product.
>
>
> Agreed. Also, the decoder must know the key; this makes it
> a huge problem for a end-user device to decode the watermarked
> signal and act on it, which is what so many DRM schemes want
> to do. I have never seen anything resembling a PUBLIC or
> PUBLIC-KEY method to decode a watermarked signal that can
> seriously claim to be safe against removal of the watermark.
>
>
> Francois Gieu

There is a half way. Modern players store their programs in
PROM so most users are unable to modify the software.

Invent a method of payment that the customers find acceptable,
so few will want to cheat. Make the watermark consist of two
parts - first the song ID & singer and second the email address
of the user. Include this information in the file header. Append
a digital signature. The music can now be traced back to the
purchaser.

It is possible to change the watermark but this is proof that
something fraudulent is going on. Pass laws banning that.
Such laws do not get in the way of honest people.

The players should accept say 5 email addresses.
Received on Thu Sep 29 21:38:57 2005