Should an EU opendata project use Google, Bing Maps or OpenStreetMap? Does it even matter?

Google, Bing and OpenStreetMap (OSM) all have dedicated followers in the mapping fraternity, and there seems to be a healthy level of competition between them. Last week, a well-known OSM’er blasted Google for stealing their ideas and exploiting their open community for commercial gain. These arguments are not new and, whatever the case, all players deserve recognition for bringing mapping to the masses. But I wonder how many people really appreciate the differences between the different systems, not just in terms of technical usability but also legally & commercially.

(photo from authors own Flickr collection)


I’m currently acting as a part-time data usability adviser on an EU-funded geohazard project. The aim of this 3-year project is to create satellite-derived terrain motion datasets for major urban areas across Europe, and to make the data available in the most accessible, understandable and reusable form. The instigators of this EU project are mostly geology and remote sensing experts and so, being more familiar with data creation than data publishing, they are planning to build some kind of INSPIRE-compliant map portal… uurrgh.

Portals date from prehistoric internet times, when people were still chained to their desktops. So I’ve slowly begun to nudge the project team away from the monolithic portal idea. They are now quite excited about opendata concepts such as mash-ups, linked data, the semantic web, and the need to cater for diverse user needs and platforms with simple standards. But, besides all this, and as part of the EU funding requirement, the project still needs to build a site that displays the data in a pre-cooked way which lay people can visually grasp. So a base mapping platform is needed. Should I recommend Google, Bing or OSM?

Rather than pretend to have all the answers (I don’t – see earlier blogpost… ), I’d love to hear your opinions. Does the choice really matter or is it purely ideological? Here’s some of my thoughts to get the ball rolling:

OpenStreetMap: Probably the best choice politically, if only for the term ‘open’. Coverage looks very good and is nearing completion in many parts of Europe. But it will never be as consistent as that of other suppliers, and once coverage is complete there’s still the issue of updates (once the blank spaces are mapped, will people be motivated enough to remap them?).  Also, with OSM being an open community, there is some uncertainty around governance and future direction – it’s all very organic and could go off in multiple directions. Then there’s an unresolved issue around the Creative Commons licence which restricts re-use for commercial purposes (unless you’re in the business of giving away your intellectual property). There has been talk of a new OpenDB licence but to be honest I’m not sure how advanced this is and whether it will resolve this particular issue.

Google Maps: Love or loathe them, their cloud infrastructure is unbreakable and their market share in mass mapping is by far the biggest. This is also why some citizen mappers (albeit a minority) prefer Google Mapmaker to OSM for data collection: where infrastructure is sparse,  Google offers an out-of-the-box solution with minimum hassle – if you can live with Google’s licence terms and commercial aims. Some people are naturally worried about Google claiming usage rights over your data but in some cases – like an opendata project – this can actually be helpful as it massively broadens your audience. And Google certainly does not claim ownership of your data. The deal is basically that you provide non-exclusive access to your data in return for a free professional service. In some circumstances this is quite a good deal, and does not interfere with your ability to re-use your own data in another form on another platform e.g. as part of a paid product.

Bing Maps.  Every year I marvel at Blaise Aguera’s new innovations but in terms of global implementation, Bing Maps seem behind the curve. They were the first to launch bird’s eye imagery but are only now beginning to catch up with Street View (in collaboration with Navteq), and the lag time to implement Blaise’s magic seems quite long, at least in Europe. To my knowledge Bing’s user terms are very similar to Google’s, so there is little difference in terms of IP or re-use.  However, Microsoft recently invested in OSM, which is now included as a layer in Bing Maps (and OSM founder Steve Coast even joined them as an employee), so this could have a very interesting future.

Keeping in mind that the geohazard ‘portal’ (urrgh!) would only use one of these services as visual background mapping – like wallpaper, really – and that it will not launch until 2-3 years from now (a long time in internet terms!), which service would you recommend? Does it even matter, or is the choice purely ideological?


Why governments should outsource open data to ESRI and Google

Mapping, environment, demographics or transport data: From the White House to Europe and Canberra, it seems like governments are suddenly throwing the whole lot over the fence in the name of transparency. #OpenData is the new buzzword in the land of open web awesomeness, and everybody shouts, hooray!

In the UK the government has so far published over 3000 datasets on This is truly great news and there are some interesting nuggets in there. The media somewhat naively refer to it as a huge “database” but, in reality (with the exception of Ordnance Survey mapping), it’s an unsightly bucket of wonky PDFs, random Word documents, grainy JPEGs, and dubious spreadsheets that look like they fell off the back of a truck. That’s because they did. Which is exactly the point.

Open data reminds me of the tree house I recently built. It looks like a garden shed dropped from 30ft. But the kids love it, so it’s fit-for-purpose. It’s just not very usable for other purposes. And the same goes for government data.

But the quality of open data will improve, you say. Indeed, some of it will improve. But governments primarily exist to govern, not to create beautiful datasets. People would rather have more police on the streets and not be mugged, rather than see awesome data showing all the muggings.

But good open data will improve those very public services, you say. True, but I doubt that professional experts in public agencies need to be told by amateur Joe Blogs how to do their jobs. Transparency and accountability are great, but “like citizen dentistry, some things are best left to the experts.” It’s the same with open data. Let’s leave that to the experts too.

In the US, the White House and ESRI have done a major deal on hosting all federal geodata through a proprietary ‘open’ portal. A contradiction in terms? Anti-competitive? Who cares – both parties are very clever and you can read more about it on James Fee’s blog. The government gets a free service, ESRI gets the data format advantage and everyone else can piggy-back off it using industry standards like OGC. This nicely sorts out the 21.5% I mentioned in my previous blog post.

And while we’re at it, why don’t governments just give the rest to Google to host and manage, on behalf of everyone else, so we can actually find it, and slice & dice it off the shelf. So governments can get on with the job of governing, and the economy can get on with the job of innovating.

But of course there are many commercial and other reasons why this would be undesirable. Besides it wouldn’t pass competition or privacy laws in Europe. So instead there will be continued piece-meal data creation, management, duplication, and tendering. As someone once said, “there’s a lot of money to be made in prolonging the data problem.”

We can’t have it both ways. So just forget everything I wrote. Instead, how about you enjoy some truly beautiful data in a new book by Now that is awesome. And, interestingly, its main source is… Wikipedia.