Friday Mar 21, 2008

Public Data should be free

A government study has concluded that it would best to stop charging for public data, reported in the Guardian yesterday. In the 80's the UK Government established 'trading funds' for a number of its statistical and data management bodies including the ordnance survey (Maps), DVLA (Road Vehicles and Users), Companies House, the Land Registry, Met Office and Hydrographic Office, and required them to charge for access to data that had either been payed for by the taxpayer, or it was mandatory to provide to government.

This research, conducted by a team from Cambridge University, discovered that freeing the information creates greater value in the economy than would be lost through charge income.

The campaigners for free information argue, firstly that consumers have payed already for the information through their tax payments, and secondly that the government is often in a monopoly position as the only body capable of collecting some of the data through its power of compulsion. The state monopoly makes it very hard to determine a market price, particularly as the marginal cost to supply is zero. The report also denies the argument that participation in a market encourages innovation in the supply chain, because of both the lack of regulation, and the monopoly position of the government. It should be noted that some of the government's "income" is paid with taxation, since the government agencies cross charge each other.

The release of this public information would in all probability lead to innovation in the use cases as more people seek to add value to it, with different approaches and use cases, and its this innovation that will crete real economic value. This is a very real case showing that welfare optisation occurs when information and knowledge is charged at marginal cost, which for digital information is zero or virtually zero.

We'll have to see what the Treasury does.

You might also want to see the Guardian's Campaign Page, Free our Data and the Open Knowledge Foundation's web site.


Wednesday Aug 29, 2007

Discover remarkable things, in a remarkable way

One of the neat tools that are being incorporated into Sun's web site came from a product called slynkr, which the authors created a tag line, "discover remarkable things". Jamey Wood, the author blogs here, and there is a trial public version which I have been using to post news articles.


slynkr's home page


The public version's category list gives you an idea of the interests of the correspondnents/users of this trial, as does the tag page, which is pretty cluttered these days as there is no lower range threshold. But the tag page looks like this!


the slynkr tags page


It is now open source and has a home at

It is a database of URLs and comments. Basically there is a form to register a URL and users can vote, comment and tag the URLs. Slynkr will also browse to obtain any public tags on the URL. All this data is held in a database and can be viewed/consumed via HTML pages or RSS. (There is a query screen, that generates the RSS URL for filtered queries and one can for instance create an RSS feed for all your own posts, or those voted for by people). For instance I consume and republish my posts at slynkr at my planet as I don't always post these URLs to; I post those I want to read again to and those I want other's to read to slynkr. Actually, there's a bug in the way my aggregator consumes the RSS feed and I need to work out if its my planet, or the slynkr instance.

At first sight, it seems very similar to digg, but its now open source and allows a managed community. I think that federation would be a major asset in the product and now its open source, its in my hands, after I learn Java. Basically, it allows a follows/followed relationship and with strong categorisation and tagging functionality, people can build thier own relationships based on tag usage.

Digg though has some great visual representations of their RSS feed. See and the three visual tools, big spy, stack and swarm.

Big Spy talks for itself, it represents on the screen in real-time, which is good because digg is really busy. It optimises the display for popular.Both size and colour are relevant.

big spy at digg labs

Stack uses flash's animation to show the popularity of articles, votes drop in on the blobs and change colour, while the most recent articles are scrolled underneath the stack line.

{short description of image}

Swarm is another representation that uses flash's animation to show the relationships between stuff and uses colour and grow/shrink on hover.

swarm at digg labs

It would seem these are not open source, but I wonder how one might build such a thing.

Many apologies, I have used a table to format the html version of this blog, the pictures were taken on 25th August.





« July 2016