DMM: Digital Me Management

Empowering technologies for identity, opacity, better sharing

This document has been submitted to the W3C Workshop on the Future of Social Networking, 15-16 January 2009, Barcelona.

Background

In this paper, we define content or work as any kind of production including text, music, video, photograph, etc created by an individual.

The Web owns its success to offering the possibility for users to share content on Web sites. The first generation of hosting platforms proposed FTP services which allowed its users to upload their works. The evolution of online advertisement allowed hosting services to provide free hosting, paid for with data mining or the display of advertisement at the top of hosted Web pages. Another evolution of hosting technologies saw Web User Interfaces make hosting one's own content accessible to more users: Web applications evolved and now offer editing interfaces to put to your content online, such as blogs and photographs hosting sites. The growth of these platforms created islands with specific features, commonly called silos.

Content hosted on silo services has a number of issues:

1. Who owns the work? For practical and legal reasons, social network sites impose terms of use whereby uploading content also involves giving the hosting company a full license on the uploaded content. The legal reasons are understandable, as it seems very hard for any commercial or non commercial platforms to use or reuse work without entering into legal troubles, and a default agreement forfeiting property rights to the hosting company appears to be a safe choice.
2. What happens to the data? A movement has recently arisen to free your data, promoting the possibility for users to move content from one platform to another. This is a step in a right direction, which will help users move their content from an online service to another, but does not address the technical hurdle of hosting one's content: silos are for the time being one of the best way for the user to have a kind of control on the work.
3. How can a user control who sees or uses their data? As of late 2008, the only reasonably easy way for a user to host content online, with any granularity over access and reuse control, is to use one of the large online platforms – walled gardens or similar –. Any self-hosting solution involves technical and complex access control mechanisms. Furthermore, the often-used method chosen by users to express their choice in content reuse – either display of a copyright mark or display of a creative commons license mark – do not provide any enforcement mechanism.

How do we create an environment where the users can own their work, have control on them and share them on their own site or on a hosting platform?

Position Statement

A few issues around content access

Social Networking Sites have a very limited set of features for sharing work. Most of the time, there is a 3-levels system (inherited from unix). A user can keep his/her work for oneself, share with friends (group) or share with public. Once content is public, technical methods for the enforcement of distribution licenses are scarce.

In the case of personal hosting, the issues are even more acute, unless you are in a very small minority of technical-savvy system administrators:

• obscurity: The user puts work on one's own server and shares it with another person through an URI, but it is basically accessible to anyone. Security through obscurity is a very weak mechanism, but is the only one available to most non-technical users.
• robots.txt: the robots exclusion protocol is an old protocol used to forbid access to content by indexing robots. This protocol is extremely limited: only one such file can be used for which makes it not practical for hosting platform. It is also an explicit declaration, which makes it impractical for content that the user prefers to hide from public view. Harvesting spam bots, badly configured bots, and most of feed readers do not respect it.
• sitemap.xml: The sitemap protocol, more recent, refines the concept of robots.txt by informing search engines about pages on their sites that are available for crawling. The sitemap protocol is also explicit declaration, and does not, at this point, provide any meta-information about the licensing terms for all available content.
• .htaccess: The Apache Web server provides a way for users to specific localized access rules through the usage of the .htaccess configuration file available. .htaccess lets users define who can access a whole section, or individually for each resources. It is also not explicitly declarative. The server decides on the access, not the client. It could be an effective way, but it is only available for one platform. It is too technical for users to configure. It is too powerful that system administrators do not authorize its usage for security reasons.

A few proposal for controlling sharing

There are mixing pieces that would help create systems around the control on work sharing. CreativeCommons has created a set of licenses to give options for users on sharing content. Unfortunately, it relies on the good will and good behaviour of those exploring this content. The Non Commercial license is loosely defined and creates issues. Ex: A work aggregated on another site which is displaying advertisement.

DRMs are used by big companies for protecting their own work. It raises a lot of issues because it is often for blocking usage more than promoting usage in accordance with some criteria. We could imagine a system where the users have full control on their content sharing.

User agents (bots, clients, aggregators, etc.) have to be identified and having a clear declarative usage policy for machine to machine negociation. Ex: I'm an aggregator interested in work that I can reuse in a commercial context.

Servers need a common and simple way of controlling the access to the work. Ex: This work at this URI has this license and is accessible by this person with this particular credentials.

Authors History

Karl Dubost has been working for W3C for 8 years and is now independent. He is interested by all issues surrounding data access granularity and privacy. Olivier Théreaux has worked for more than 5 years on W3C's open source tools and services, and is the lead technical architect of the ArtBeat network of web sites.