GIS and Me

Years ago, when I left university I, along with some friends, set up a company to develop a GIS product. We then went after some venture capital and just around that time rumours filtered through that Microsoft were hiring some of the best minds in GIS and we thought “Oh, bugger!” our newly formed business would not be able to compete with Microsoft. We were left in a state of shock when Microsoft release MapPoint. It was rubbish! (At least it was rubbish as a competitor to our product.) You couldn’t do even half of what our product could do or our customers were asking for in terms of new features. MapPoint was strong on route planning, but that was the only competitive edge over our own product. We didn’t do route planning at all. MapPoint offered very little in the way of spatial processing and, as far as we could see, was just a very limited way to display data on a map.

However, it looks like things are changing. Spatial data support was finally released in the November CTP of SQL Server 2008. There is no graphical front end, but the processing ability is there and that is the more important part. It is easy to put a new front end on to the data. Microsoft have some solutions in that area already such as Virtual Earth and MapPoint. There are also other vendors out there who have their own, more powerful, front end solution.

My only reservation about these new abilities, at least in the UK, is that spatial data sets are very expensive. Unlike the USA the UK government holds the copyright on the data it produces and it then charges a king’s ransom for access to it. This stifles business’s ability to use spatial data effectively and it will have to resort to poorer quality datasets or being restricted in the ways that it can use the data.A business could, for example, use map data generated by OpenStreetMap. Doing that stifles the way the business can use the data as OpenStreetMap is released under a Creative Commons BY-SA (Attribution Share-Alike) license. This means that anything a company does with OpenStreetMap data must be attributed to OpenStreetMap (not an onerous condition, if they use OS data they must include a Crown Copyright notice) and the derivative works it creates must be covered under the same license – How many companies are going to be willing to do that?

According to the OSM website “CC-By-Sa do not force you to make any source data available. You are only restricted in the license you choose for distributed data”. On the face of it that sounds like companies can use the license so long as the map is kept internally. Well, I can’t really see how that is possible. If one employee creates a derivative work, say a sales chart, and passes it out at a meeting then it has been distributed and must therefore be under the CC-By-Sa license and once in that state the company cannot stop its distribution. Obviously I am not a lawyer, but that is my interpretation of the license.

Currently, I’m looking at ways to put together demos and articles on using the Spatial data types in SQL Server 2008, however I cannot us OS data because it is too expensive and I cannot use OpenStreetMap because I’m not willing to release my articles, presentations and demos (my derivative works) under that license. I’d also like to use UK data because that is where I’ll be doing my presentations/demos.

If anyone has links to vector data that I can use then I’d love to know about it.


  1. Have a look at this article and sign the petition to end Crown Copyright…'ve already paid for it through our taxes so why should we have to pay to use it!!!

  2. Hey! I see Colin has already signed it. Time for all your readers to do likewise.

  3. Unfortunately the petition is closed. I have yet to receive a reply from the government regarding it. As an alternative, it might be useful to write to your MP regarding this.

  4. Steve Coast says:

    You want your cake and to eat it. I think you maybe don’t know the difference between collective and derivative works, and that you over-rate the value of your ideas and data.How many companies are going to be willing to do that?A entire industry arose around GPL software like Linux. A few like you said it would be impossible, but I think an entire industry will also arise around open data.

  5. @Steve Coast: Well, if you are a lawyer then please explain where my interpretation is wrong. From the example on the OSM website a webpage is collective as it is a collection of files. The map can be extracted easily as it is effectively in a different “layer” from the rest of the page and other images on the page. Derivative is where the layers cannot be separated.If I create a powerpoint presentation and include OSM data in it then that is derivative. I cannot distribute the power point file without also distributing the map as it is integral to the file format. The same goes for any demos that I create. The database I’d need to use would be an amalgam of OSM data and other data (not necessarily my data either. It might be licensed from some other source in a way that is incompatible with CC-BY-SA). It would not be easy distribute it as two parts as the relationships within the database would necessarily break.If that is an incorrect interpretation then please explain why the power point presentation or demo is a collective work. Also, if you want on side of people you might want to work on your people skills. Stating that I “over-rate the value of [my] ideas and data” isn’t the best way to get someone to your side of the argument. I have had to defend my intelectual property on a number of occasions so I know there are people out there that have (and potentially will again) take my work and claim it as their own.On to the business aspect, you are right that some companies have sprung up around GPL. However, my example was of some internal function within a company using OSM data with their own to create a derivative work (again, explain why this example is not a derivative work if I have inadvertently misinterpreted it). If the derivative work contains some company confidential data then the company has instantly lost control of that data. The moment the work is distributed the company has lost control of it. Sure, the raw data that was used is still intact, but control over the the aggregate result is now irrevocably lost. I don’t know many companies that would allow that kind of thing to happen.I don’t think even companies that revolve around Linux and GPL software would want their internal sales figures, HR data (which would be even more serious) and the like potentially exposed (or exposable) like that. They may decide if they keep the distribution internal then that is okay, but all it takes is one disgruntled employee…

  6. Tim Waters says:

    Internally, a company can use OSM data and their own. It’s only if the data is *distributed* publicly, then that data should also be licensed by cc-by-sa. There’s is absolutely nothing in the license that requires a company to distribute. A company therefore is safe using their “company confidential” information with OSM data for internal use.Also, if a map is made and distributed externally, there’s absolutely nothing in the license that requires that the source data should also be distributed. If they do distribute a map, then in the case of confidential information, then I would ask, why if it’s so confidential, is it being distributed?If the data is not confidential, (i.e. location of retail stores) then perhaps the data should be released under a similar license?The problem would be, as you pointed out, mixing other licensed data, and then wanting to distribute the final map. I would recommend looking at the source of the other data, as it is precisely because of restrictive data licenses, that the openstreetmap project was created.

  7. Licence issues are a perennial in OpenStreetMap. Open data licensing is very much in its infancy and, by and large, the pros and cons of applying “creative works” licences (e.g. CC) or software licences (GPL) have not yet been satisfactorily worked through. The derivative/collective works question is one of those that occurs repeatedly.The OpenStreetMap Foundation is currently making a determined attempt to resolve some of these issues, potentially through a new definition of derivative/collective as it applies to OSM data. We hope to have definite progress to report in the New Year.Best wishesRichard Fairhurstpress & publicity, OSMF

Leave a Comment

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s