On Oct 17, 2006, at 12:17 AM, Raj Singh wrote:
This is a crucial point, and where the rubber really meets the
road. How do we enforce digital rights in an open, interoperable
software environment? Well, we don't have the answer to that yet,
but that's our goal.
And as an industry consortium comprising every type of organization
from commercial, open source, academic, govt. etc, nothing less
will do.
You can't.
We went down this road years ago with the early stuff i worked on.
Nobody can control the propagation of bits of data unless they
control every piece of equipment and software that it propagates to
and through...
... and that's only an option if the RIAA/MPAA model is followed: (1)
proprietary protocol; (2) every producer of significant, desirable
content has to buy into it; (3) you've got to license (and then
enforce) this DRM in everything that touches the content.
That could be done by a massive application of force by a
concentrated lobby. But there isn't such a lobby, and geodata's not a
DVD where a patented protocol can be licensed to device makers (and
what is their motivation).
Besides, successfully accomplishing the above would only mean that
some data, probably from paranoid local governments, will be
permanently walled off.
The best and probably /only/ way to tiptoe up to this is the same one
we drilled down to in every discussion of this subject:
1. Define an XML standard for a licensing-policy clause over geodata
2. Think again about #1. How is this different than policies over
other open data? If it's not, find an existing XML spec that defines
redistribution policy over data. Creative Commons already has those.
3. Understand that a policy clause in data is a REQUEST, not an ORDER
4. Encrypt private data using well known methods, when necessary, or
keep it out of public repositories altogether
Licensing the content (use Creative Commons) would be the way to give
some teeth to the policy request. I see nothing wrong with licensure
of geodata as opposed to any other collection of data. If there are
problems, the fact of the license won't be the issue, but that the
wrong license was chosen. That's a transient problem that can be
corrected or eventually worked around. Lots of desirable data will go
out with a license anyway, better to embrace that reality.
I haven't worked in this stuff in a while but I think we got a few
things right many years ago...
Two simple rules headed off a lot of complex protocol and design issues:
- If you are a content provider, offer as much meta-info about the
nature of the content as you can, so a client can decide whether to
pull down the content or not
- If you are a client, skip/ignore anything you don't recognize or
just can't understand
This approach allowed for any, all, or none of:
encrypted data, proprietary encodings, very simple clients
... that nonetheless interoperate, at least a little.
- jim
_______________________________________________
Geowanking mailing list
[email protected]
http://lists.burri.to/mailman/listinfo/geowanking
