Skip to main content

Over the last month or so, the team have begun to engage more seriously with the API community. We’ve written blog posts, and talked at Open Data Day. We wanted (and still want!) to hear from people about what kind of information our API should carry, and how it should be structured and work to meet best practice.

We’ve been very happy with the response. :) It’s been great to see the community step forward and offer its huge knowledge and experience.

There’s been some comment in the last few days, though, about the fact that we’ve taken our API down for a few months. We did warn people it was going to happen, but we realise that we should have been more explicit about it, and the reasoning behind it.

The ‘old’ API was built on the beta version, and data models, of the website. Since the full public launch of the site, we’ve realised that those data models weren’t right for what we wanted to do. A major problem is that they weren’t extensible at all.

We’ve been putting a lot of work into changing the site’s data models and making them extensible — work, admittedly, that’s not obvious from outside.

Effectively, we’ve been futureproofing the code behind the site.

One of the major pieces of work that’s just been completed has been to change how the site deals with the roles, organisations and people that are part of the Government A–Z directory. And this change is, of course, heavily tied in with our API and ‘downloadable government contacts’ functionalities.

One of our core development principles is ‘release early, release often’. With limited resources, we had to decide whether to get the roles/organisations/people work, the API and the downloadable contacts all working properly before we released the website’s next version into production. This would have taken many months, however, and was felt to be a Bad Idea.

We thought long and hard about which was the right way to go: a very large, very delayed release, or smaller iterations. We chose the latter. One of the key factors we took into account was whether anyone was using our API for their business or organisation. We’re not aware of any, which is one of the reasons we’re talking with people around what sort of information our API should offer.

We agree that taking down APIs is far from best practise. Our new data models, and the new API, are and will be built in such a way that it won’t be necessary again :)

We’ve been really glad to hear from the API community about the takedown and best practice. It’s great to see that people are engaged. We haven’t heard from anyone using the API for their business or organisation, though — if you or someone you know is, please do get in touch!

And, as we say on our site, we do want to keep the conversation going. So please don’t hesitate to get in touch :) You can find the team on Twitter, the Open Government Ninjas forum or the Web Toolkit blog post.

Utility links and page information