Since Data.gov.uk went public in January of 2010, it’s attracted worldwide attention for its usability, its user-friendliness, and its underlying mission to provide citizens with the information they need in the format that they need it. British civil servant Richard Stirling, who inadvertently has become one of the more recognizable faces of the open source/open data community, took some time to talk about the challenges of data presentation, including curated versus uncurated data, the most efficient methods of presentation, and app contests. In our conversation, he also briefly touched on why data.gov.uk chose Drupal over Joomla, and the wonderful unpredictability of what the future may bring.
JI: I wanted to ask you about data visualization. Would your ultimate goal be to present all of your data in form that has a signature look, like the World Bank?
RS: It’s not something that we have as an explicit goal for all of the data on the site. If nothing else then what we’re doing is dealing with lots of different types of data sets. We have some data, like National Statistics, which is quite well curated and has pretty good standardization and there are some common things which you can pull together. But we also have some other stuff, like admin data, it’s pretty noisy. It wouldn’t fit into that visualization mechanism. We have real time energy data coming from the headquarters buildings of the government that’s surfaced on the site. Combining those two things into one viewer would be really difficult and a bit of a hard challenge. What we are focusing on is how we can make the whole process of using data.gov.uk much easier, much simpler. Some of that will be around making the data easier to explore and navigate, and bringing some of the serendipity back in. If you vaguely know what you’re looking for, it will jump out at you. Additionally, we want to make it easier to show what is on that data. That might include some visualization of the data sets with ways of giving a sample of the data that’s there, or indeed surfacing some of the good work that other people have done. We have an app gallery on the site precisely for the lay user to get into the data set. We’re working on how to tie together the data set and the applications.
JI: You ask the public to submit their applications in terms of displaying data. What is your favorite app, or at least one that surprised you with how it tied together different information strands?
RS: Obviously, they’re all my favorite. The one that was not something that I expected is an application that lets you get a bit of sleep or work done while you’re on public transport. Essentially, your phone knows where you are. We published the locations of all those public transport nodes, including train stations, bus stops, that sort of thing. Really simply you tell your phone where you’re going, and it sets off an alarm 8 minutes before you’re due to arrive. I do a lot of train travel, that’s invaluable, but certainly, not something I had in my mind originally.
JI: In terms of open data sets there’s been, practically every day, another country or region in the world that is trying to put together a new open portal. It seems like your site in particular gets cited as an inspiration for them to move forward. Is it fun for you to watch this international proliferation of open data? It seems like you’re turning into a leading figure in that movement.
RS: It’s fantastic to see so many different countries opening up their data. I spotted the other day, I think it’s Kenya, has an open data site, which is something I hadn’t looked for, and it’s good. I’m happy. One of the things we’re trying to do with our site is be completely open about what we’re doing and open up our code. So if anyone wants to follow our particular design pattern or have a look at the code to see how we’ve build stuff, they can do. The front end runs on Drupal, the back end runs on an open-source registry called CKAN.net. We’ve open-sourced the code that joins those two things together so there’s a Drupal module that queries CKAN and exposes it to the Drupal. That’s open and available on our blog and will be on Drupal.org.
JI: Why Drupal and not Joomla, for example?
RS: It seems a long time ago now. It was just a case of we did a quick feature comparison. Drupal came out on top on that. We had some experience in deploying it from within the team. It seemed like the right fit.
JI: In terms of open source code, do you think that will enter another phase of innovation or do you think people will start to try and carve it up and make it proprietary?
RS: The history of open source is one of continuous innovation and development, with people always trying new business models and putting things together. I don’t see the rate of innovation slowing down. You look at where Linux was ten years ago and where it is now, I certainly don’t feel like the pace of innovation of open source has slowed down, if anything I feel like there are more people involved, and the pace is picking up.
JI: Do you worry about trying to meet longer term goals? Do you have an idea about what the site will look like in 10 years, or you’re just open to whatever comes in and excited about that journey?
RS: We think about the longer term in terms of what are the strategic problems that need to be fixed. What is the role that we should be aiming to do well? At the moment the development focus is very much on how do we make people’s lives better now. What’s the problems we should be focusing on now: how do we improve our offering, how do we make people’s lives easier when they want to find data, when they want to work with data, when they want to see what other people have done with it. When they want to surface their data to the public, how do we make that journey easier? Those long term goals are about making it findable, usable and opening it up people so they can create good things. In terms of setting out on the wireframe what the site would look like in ten years, that would be very challenging, yeah?
JI: [Laughs in agreement.]
RS: If you were trying to sit down ten years ago and set out what the site would look like today, I’m not sure you could do that. Or if you did, it would look completely different. For a start, you wouldn’t have people in there. There wouldn’t be quite so much focus on joining up the user base, as we’ve done at the moment. ∞