[sldev] Re: [ARCH] Permissions

[dirk husemann]

John Hurliman wrote:
> Argent Stonecutter wrote:
>> From: John Hurliman <jhurliman at wsu.edu>
>>> To get something close to that would require a change in how content is
>>> delivered to clients.
>> [...]
>>> The discussion about trusted/untrusted domains really has nothing to do
>>> with this as the weak link in the chain is the clients
>>
>> What part of "I know that, this is not a 'technical solution', it's
>> like a 'speed limit' sign" do I have to explain again THIS time?
>
> Speed limits work not because there is a sign on the side of the road,
> but because there of patrolling officers and radar/laser guns and
> court fines. There is a very good discussion to be had here about
> social solutions to the problem, but I don't think it affects the
> evolution of the architecture technical spec. As long as the system is
> extensible enough that people can implement a fake permission system
> on top of it there should be no worries.
i think the way the discussion has been going (at least my perhaps
limited understanding of it), we all agree that we won't be able to
achieve an airtight permission system --- because in the end it's up to
the client for a lot of the stuff. what might be protectable (might) is
the script logic as that (currently) runs on the servers and can be
"protected" via "contracts".

yes, in the end "DRM doesn't work". but the permissions system would
allow the content creators to signal intent, i think of it as a "license
digest".

it affects the architecture to the extent that we need to have answers
to these questions because they are one of the first ones to pop up when
discussing this with people.

    cheers,
    dirk

-- 
dr dirk husemann, pervasive computing, ibm zurich research lab
--- hud at zurich.ibm.com --- +41 44 724 8573 --- SL: dr scofield