[Opensim-dev] Interrelated difficult problems related to asset portability

[Kyle Hamilton]

On Tue, May 27, 2008 at 4:39 PM, dan miller <danbmil99 at yahoo.com> wrote:
>
> My question for Kyle is this however: aren't you potentially opening a pandora's box of privacy concerns?  You are basically encouraging a community of tattletales, who whisper to the powers that be any time they think they see a copyright violation.  There's something vaguely disquieting about the ramifications of such a world, even though I think it may be preferable to a locked-down shackled DRM world.
>
> -danx0r

It's going to take me a bit to get my thoughts in order on this, so
please bear with me.

There are aspects of our current legal regime that require the
copyright owner to know about issues and initiate an action before the
protections can be invoked.  However, these rely on the owners
learning about the issues in a timely fashion.

(Incidentally, this reminds me: we need a console command or something
to perform a "DMCA takedown" of assets.  However, the data needs to
not actually be removed in that event, because a DMCA takedown target
can submit a response that it's not an infringement.  Instead, set a
flag so that the asset server just returns something that won't crash
the viewer but isn't the original content for that UUID.)

I am unhappy with the idea of tattling... but the problem becomes one
of "we must find a way to comply with the current regime, or else the
regime is going to specify additional compliance measures."  I'm
proposing this idea because it brings the people back into the realm
of 'dignified beings'.  If someone is misusing content owned by
another, that other has the right to know about it, basically.  We've
had to deal with DRM because nobody wanted to tell about abuses, and
the thought of burdening servers with Yet More Processing Load (as
well as the server operators' typically-inherent disgust with the idea
of tattling) always ended up disabling any kind of notification system
in place -- if it was ever even built in the first place.

In this case, users essentially become able to make the choice that
was previously denied them by well-meaning server operators, and to be
able to use their own bandwidth (not the servers') to do so.

Since I am a fan of empowering people more than disempowering them, I
am much happier about this idea than the idea of having my ability to
use copyrighted content the way I want to pulled out of my hands.

There are some important caveats, though.  First, the data to be
squawked on must be put into the system with a notification URL in the
first place (meaning it must be made available on the network,
somewhere).  Second, the viewer must actually be told to download
bitwise copies of the content (by it being referenced in a scene that
the viewer is requested to load).  Third, there can't be much
"in-your-face" reason to make a change to the asset data.  These are
the things that make this system work -- and if any of them are
changed, no notification occurs.

This doesn't detect derivative works.  This doesn't detect changes.
This doesn't detect clips.  This doesn't detect frames of a video.
This doesn't detect anything other than bitwise copies.

There isn't much that I can suggest for the latter things.  Most of my
thought over the past 8 months or so has been related to figuring out
how to put together something that would work on the most basic level,
and to figure out how to make it work within the scope of my ethics...
and this is the best that I can suggest.

I hope that I've at least managed to spark some thought about new ways
to protect content within the realization that computers can't be the
final arbiters or enforcers, that there already is something that does
that job, and that there are much less invasive ways to do the same
kinds of things that the companies want to use the stranglehold of DRM
for.

-Kyle H