Guardian Open Platform

The Guardian today announced the Open Platform, comprising two products: a Content API built on top of their existing search engine allowing other people to build applications on top of the Guardian’s content, and the Data Store which is a collection of public and Guardian-owned data sets made freely available for reuse.

Open Platform sticker

I was invited along to the announcement, and took some hurried notes during the introductory talks. They’re by no means comprehensive but I did manage to capture a few quotes from some of the speakers and most of the Q&A session too. Here are the highlights…

Tim Brooks, Managing Director

Before the web, we reached around 6 million people per week. We now reach 33 million in a good week.

Emily Bell, Director of Digital Content

We take risks. … Handing this over to you guys is a risk. But one that I’m sure will pay back in multiple dividends in terms of the creativity it unlocks. It’s a significant step towards the idea of ‘Guardian Everywhere’.

Mike Bracken – Director of Technology Development

showing Chalkboards

We can’t do everything ourselves … We want you to show us how to improve

Stephen Dunn, Head of technology strategy

shared a timeline of some of the Guardian’s online activity, pointing out that today’s announcement is a step in a journey.

  • 1995 – Guardian first on the web
  • 1996 – Guardian New Media Lab
  • 1998 – Unmoderated Talkboards launched
  • 1999 – Guardian Unlimited launched with registration sustem, but was removed in the same year.
  • 1999 – RSS feeds and headlines distribution service
  • 2001 – first Guardian blog
  • 2006 – Free Our Data
  • 2006 – Comment Is Free
  • 2007 – RSS Everywhere
  • 2008 – Full feeds (with ads)
  • 2008 – First guardian hack day

and their web principles:

  1. Permanent – (a cool URI is one that does not change)
  2. Addressable – (resources are about something, ready fort the social web. We live in ‘the age of Point-at-Things‘)
  3. Discoverable – (multiple routes to content. Tagging drives discovery)
  4. Open – (hackable URLs, using and contributing to open source tech)

Matt McAlister, Heard of Guardian Developer Network

Today the Guardian announces the open platform. It’s the suite of services allowing partners to build services with the Guardian

Some more detail from Matt on the terms of use. You get rich, tagged article content. Full content, not just the headline and abstract. And you can publish it in your apps

We decided that the best price point for fueling growth is free. We want you to be able to make money on your site too.

Looking at the site, there are Commercial partner programs to enable commercial use. The FAQ says that “The Content API is free to use. There are some terms that you must adhere to for the free access level. For example, our default limit on queries per day is 5,000 calls, and we will in the future ask partners to display advertising from our ad network on pages with our full content.”

Explaining that it’s a beta trial, and they’ll be approving API keys on a limited basis (which has pissed off Dave Winer), Matt said

We’ll be doing that slowly so that we can understand what the dynamics are, and what you want to build … But we do plan to open more widely soon.

Simon Willison

Simon demonstrated the API explorer (the first app written using the API) and how each set of results also specifies a set of usable filters which are available in source, and shown in explorer.

New concepts can be prototyped in less time than the meeting you’d have needed to discuss it in the first place

and demonstrated this point using several demos including (from Chris Thorpe) and APIMaps from Stamen Design. Pleasingly, AMEE are another launch partner for the Data Store too. Exciting stuff.

Q&A session

Q – (Jon Slattery from Press Gazette) – Are you happy for news organisations to use your material?
A – (Emily Bell) – They do already! Frequently. The free model is built with developer community in mind. If other news orgs want to use the data because it’s better then theirs, that’s fine within the terms.

Q – (James Cronin) – Can we store data in your data store?
A – (Matt McAlister) – we’re publishing data into the store at the moment, but tell us what you’ve got, we’d love to look after it.

Q – (Stefan Magdalinski from Moo) [re 5000 hit daily API limit]
A (Simon Willison) – 5k hits is what you can ask from us in 24 hours. You’re allowed to cache that data though, so could serve as many pages as you like. We’re also serving smart caching headers with 50 min expiry.

Q – Are you exposing photos & media?
A – (Matt McAlister) – Not yet. But we return headline and description and URL for our photo pages.

Q – (Andy Pipes) – Any plans to share personal data, like the New York Times?
A – (Matt McAlister) – No. None at all.