MS DRM OS

[Michael Motyka]

>On 20 Dec 2001, at 9:20, Michael Motyka wrote:
>
>> My thought is that it is not novel in any way save that it witholds
root access
>> from the owner of the machine.
>>
>
>I think it does a little more than that.  "Deny the luser owner
>root access" is sufficient to explain how the luser is prevented
>from copying or modifying the trusted content, but it doesn't
>explain how "trusted" apps can access the data.
>In essence, deny the luser root access + all programs signed
>by microsoft automatically run as root. Neither piece alone
>would be innovative enough to be patentable, but maybe the combo
>is.
>
>George
>
I think the combo is used regularly  : e.g. version control logfiles
that are not directly accessible to a user but the user can have ci/co
access via applications that run with the correct permissions. A user
may fuck up a version of the file but may not make things unrecoverable.

We could get into a long discussion of the details of data source and
sink types ( sockets, disk files, other devices ) but the basic
principle is the id/permissions one.

I just don't think the concept is particularly novel but I have no doubt
it will be implemented and protected viciously. I use the word viciously
because that is the only way to make DRM work and there is a great deal
of money at stake.

Mike