Joho the Blog
An Entry from the Archives

I'm not sure about Scott's island analogy... doesn't Creative Commons already enable harvesting and search? (e.g. through search.creativecommons.org) Is the CC scheme missing unique IDs for works?

Hi Seb, regarding CC's search, I'd say:

1) there doesn't appear to be a way for 3rd parties to build applications / services on top of it -- it's a search site, rather than a dataset that anybody can use, however they want.

2) it won't ever include non-CC works. For a registry to work, it should draw from outside CC too (ie, from the Copyright Office, entertainment industry, public domain, etc)

-Scott

Why would I be interested in a centralized, authorized, registry?

The first thing such a registry would do is start telling me I can't access this and that, that I'm not 'authorized'.

The second thing it would do is tell me that I can't register this and that, that I'm not 'qualified' or that the content is not 'appropriate'.

The first is annoying. The second is dangerous.

Anybody can create objects. Anybody can express rights. Any system that does not respect these two principles will fail; the producer and the consumer have become one.

Here's a better idea: create rights expression models in XML using ODRL. Put them on your website. Reference these models with a pointer embedded in the object metadata and/or the object itself. Build systems that interpret the ODRL and offer functionality accordingly.

No, it's not rights enforcement, it's rights expression. Conformance is voluntary, and made easier by the tools we work with. Yes, there is slippage, but this creates the domain of fair use. Enforcement is handled where it should be, for large commercial violations, in the courts.

Hi Stephen, I believe you're in Canada? Don't you currently have some sort of system that compensates rightsholders for (some) filesharing?

As you must know, there is lots of talk about ACSes these days (alternative compensation systems). Do you favor such a resolution to the p2p situation?

If so, wouldn't that require just such a registry in order to know who to compensate? Doesn't Canada use such a registry? If there's going to be a registry, shouldn't it be open for all to use?

-Scott

A few more comments in response to Stephen...

1) I'm not at all suggesting that people should be forced to use such a registry, merely that it should exist, and that it should be accessible to all.

2) We're far more likely to wind up with a system that rejects works as not 'appropriate' if the government gets involved (ie, "compulsories")

3) When you say: "Anybody can create objects. Anybody can express rights. Any system that does not respect these two principles will fail" -- perhaps I misunderstand, but it doesn't seem like filesharing is "failing"?

-Scott

So a satisfactory approach might be to generalize the Creative Commons idea to enable the full gamut of rights to be expressed (ODRL might work there), then have harvesting spiders (à la Technorati) build databases that third parties can build upon?

Seb:

Using a spider is a really interesting idea! ;)

One concern is that my intention was for the database to be "trustworthy" or "authoritative" in a legal way -- so that new commercial and/or non-commercial services could freely and safely share info provided by the database.

I'm not quite as sure if a spider-populated database could be trustworthy in that way -- it seems to me that there would need to be a bit of friction there to keep the data correct (again, I point to the process of making changes to DNS records).

-Scott

Scott, I had considered that. The way you keep the rights file authoritative is that it tells you were to get a key, and the key is required for access to the material (this, obviously, applies only to commercial material).

My own work on digital rights management is collected at http://www.downes.ca/dwiki/?id=DDRM (due to a coding error, my wiki will ask you for a user ID and password - I can't just take it off, I'll be deluged by spam - as a temporary measure, please use UserID: Anymouse password: anymouse).

Hi Stephen,

It seems your preference is for file-level DRM?

One nice thing about an approach like DRUMS is that it totally sidesteps the DRM issue altogether.

-Scott