devlog

I was glad i caught the first Open Knowledge Forum last week, featuring developers from theyworkforyou, public whip, iCan and writetothem.

Between them, they have good points and proven use cases for why we just need the raw data, in a structured format, openly available without supplication.

• Public information providers save money.

Dropping raw data on a web or FTP site via an automatic process is inestimably faster, cheaper than handling requests for information on a per-use, permissions basis. Public sector publishing requirements are fulfilled without lavish information “viewing” sites which are better provided in the public domain…

• Skilled enthusiasts build query tools that are easier to use.

All web-facing hackers know that services with freely available structured data, and/or machine interfaces, propagate faster as people grow “niche” services round them - think flickr and mappr, or the google API and directionlessgov.

I enjoy these arguments, they don’t involve ungroundable statements about civic good, or radicalised “information wants to be free” sentiments of which i am always fond. Look what we can build, this gang are saying, in a few months of spare time and no funding, with open standards, and simple free software tools. They have very local proofs-of-concept, which works a lot better than pointing and shouting at FEC->FundRace, SEC->TheyRule and so on.

The implementors have all begged for info licenses through partnerships with public bodies; or borrowed information from the net, and only come to licensing terms when the service is too useful to put back in the box (cf theyworkforyou / hansard).

More poignantly for me, all of these civic services are dependent on spatial information for their workings: type in your postcode, find which borough or ward you are in, what institutions are nearby. The implementors seem annoyed, at the least, at the prices they pay or hoops they jump for access to geodata.

I’ve been helping put together the next Open Knowledge Forum, on Open Geodata which should be in London on April 14th. We should have some good speakers, and the open mapping projects coming together here should offer a good local proof-of-concept.

Analogise the situation we’re in with geodata; it’s as if Hansard only
published human-readable PDF images of scans, or handwriting, and that in
tryng to transcribe them, we risk copyright infringement for over-use of the
word “road”