a blog about things that I've been thinking hard about

Web 2.0? We Haven't Finished Decentralising Yet.

22 December, 2006
web news and story site history

The voted web web has progressed, somewhat, from more centralised to less centralised.

Slashdot, then K5, then Digg, then CrispyNews (which didn't take off).

Also worth a mention, even though it's not specifically a "story site", Google.

tags:

A History of the "Voted Web"

SlashDot

The only thing "voted" and decentralised about Slashdot is that moderation and meta-moderation are used to filter the comments. Stories can be submitted by anyone, but the choice of which stories to post onto Slashdot is entirely that of the site's operators.

K5

K5 lets you submit stories, and K5 users vote on which stories should be promoted to the "Front Page".

Decentralised: Any K5 user can write a story. Any K5 user can vote on a story. K5 does not require exclusive rights to submitted content, nor does it require first publication.

Not so decentralised: You have to be a K5 user in order to submit or vote, and membership is ultimately controlled by the K5 site owner ("Rusty"). Although K5 does not require exclusive rights on stories, the front page links only ever point to the copies of the stories submitted onto the K5 site. Votes are located on the K5 site and the vote aggregation process is controlled by the K5 site.

Digg.com

Digg lets you submit stories, and Digg users vote on which stories should be promoted to the "Front Page". More voting then occurs to determine which stories stay on the front page, and for how long.

Decentralised: Any Digg user can submit a story. The submitter may or may not be the writer of the story. The location of the story is not required to be on the Digg site.

Not so decentralised: You have to be a Digg user to submit or vote, and membership is ultimately controlled by the Digg site owners. Votes are located on the Digg site and the vote aggregation process is controlled by the Digg site.

CrispyNews

The idea of Crispy News is that anyone can start their own voted news site.

Decentralised: Any Crispy News user can start a news site, and any Crispy News user can vote on a story.

Not so decentralised: You have to be a Crispy News user, and membership is ultimately controlled by the Crispy News site owners. Votes are located on each Crispy news site, where each news site is controlled partly by Crispy News, and partly by the Crispy News owner who administers that site. The vote aggregation process is controlled by the Crispy news site (with some control given to each administrator).

Google

Google is an implicit voting system, where website links count as "votes". The value of link votes by each website is determined recursively by the value of link votes for each website.

Decentralised: Both the content and the "votes" are owned and controlled by individual content producers and "voters".

Not so decentralised: The aggregation process is totally controlled by Google. The Google PageRank patent prevents anyone else duplicating this style of aggregation until the patent expires.

Google's Custom Search Engine facility can be seen as a step in the direction of decentralisation. However currently the customisation process does not extend to alteration of PageRank (i.e. weighted towards the particular user), which is what would be required to achieve true decentralisation of Google-style "voted" page ranking.

A Theory of Voted WebSites

For every voting website, there are four things going on which are relevant to the voting process:

  1. Production of content
  2. Identification of voters
  3. Voting for content
  4. Aggregation of votes

Even though some of the websites mentioned above are supposedly "Web 2.0", and "Web 2.0" is supposed to mean user-control and decentralisation, we can see that complete user-ownership and decentralisation has not yet occurred. In a truly "Web 2.0" voted web, there would be four types of decentralised ownership:

  1. Production of content would be decentralised. Content and its location would be owned by its producers
  2. Identification of voters would be decentralised. Each voter would own their own identity as a voter.
  3. Voting would be controlled entirely by voters. Each voter would own their own votes.
  4. Voting aggregation would happen independently of both content production and voting. The results of aggregation would be owned and controlled by the aggregators. The same user votes could be aggregated by multiple different aggregators. Aggregators could also aggregate each others' aggregations in various ways.

On the Internet, ownership and control is determined by one thing and one thing only, which is association with domain names (this is a slight exaggeration, because you can also own an IP address, but in practice domains are what matter). In other words, content, identity, voting and aggregation will be truly "web 2.0" when they are associated with domain names as follows:

  1. Content is located via URLs. The domain name of the content URLs is owned by the owner of the content.
  2. Each voter has an identity which is defined by association with a domain name which they own (i.e. via something like OpenID).
  3. Votes are located via URLs. The domain of each vote URL is owned by the voter (and a secondary part of the vote URL should contain the URL of the content being voted for).
  4. Voting aggregation is associated with the domain name of the aggregator. The results of voting aggregation are published to URLs containing the aggregator's domain name. The aggregator may also choose to publish details about how aggregation is performed and which set of users' votes make up the aggregation.

Entanglement

The unnecessary entanglement of voter identification and vote aggregation is what makes it so hard to start your own news site. It's an uphill struggle to convince users to "join" your news site, because, since it's a new site, no one has any reason to look at the output of your site's voting, so there is little motive to contribute voting effort to your new site.

A decentralised system would remove this difficulty, because the decision to vote would be completely separated from any decision about "where" to send the vote. Sites like Digg and Reddit treat user voting effort as a precious commodity which can somehow be bought and sold (in the sense that the commercial value of the sites is a function of how many active voters each site "owns"). But votes are just a form of digital content, and like all other digital content, we already have the technology to give them away for free. And since, for the most part, nobody expects to be paid for their voting efforts, there is no reason not to give them away for free, other than the fact that doing so will break the current monopoly that Digg, Reddit and their brethren have on the ownership of "their" voters.

Google's centralised PageRank-based aggregation is also doomed to extinction and replacement by a more decentralised search algorithm, just like all other past attempts to decentralise and control ownership of content on the Internet. (Of course even if central control of recursive link valuation is doomed, Google itself is not necessarily doomed. Google's ventures into cheap domain registration and associated services with its Google Apps for Your Domain product might be a sign that Google recognises the inevitability of user ownership of everything.)

Vote for or comment on this article on Reddit or Hacker News ...