Michael Collinson "Mike"
The OpenStreetMap geodata license
This is my purely personal view on the need for an OpenStreetMap license change and what that license should be. I am no lawyer. This is not an intellectual or scholarly essay though I do try to overview where we are now and what I see as the main factors to consider. I respect other views but this is mine.
When thinking of the license, I always go back to the first page of our wiki:
"... free geographic data such as street maps to anyone who wants them. The project was started because most maps you think of as free actually have legal or technical restrictions on their use, holding back people from using them in creative, productive, or unexpected ways."
These two sentences got me started in OpenStreetMap. I love OpenStreetMap, I love mapping for OpenStreetMap. I love tag discussions. I love Potlatch. I love JOSM. I hate the current license.
It is great for photos. It is great for Wikipedia. It is great for our own wiki. It just does not work when we actually want folks to go and mix our data as much as possible with other things ... other geographic data, school projects, books, Wikipedia, databases, software, games whatever. It is not for factual data and so it is too vague for us. It stops the good guys with "If I ..." questions but not the bad guys who can interpret it to suit themselves. And the good guys have to and ask each of up to 270,000 contributors their question to be fully considerate. Yeah, right. For OpenStreetMap use, the current license sucks.
So what sort of license should we have?
Software started the Open Intellectual Property movement. By the late 1990's licenses matured into two main variants, non-viral BSD Licenses versus viral GNU copyleft licenses. GNU licenses work very well with baseline software such as operating systems. But data is not software. Software has a clear edge around it. With a GNU-licensed word-processor, there is an obvious distinction between using it to make and sell your award-winning novel (A GNU license makes no claims on that) versus improving the software and selling it (A GNU license definitely has something to say about that). Using versus improving OpenStreetMap data is blurred because you will be often be doing both together.
The next pioneering wave was open licensing of creative works such as photos and pieces of text. Again there is a choice between viral and non-viral licenses personified by the excellent Creative Commons suite of licenses. Again, viral licenses work well. Wikipedia licensing helps maintains a single open reference source and makes commercial rip-offs fairly pointless. But again, data are not creative works. Creative works have less of an edge round them than software, but nevertheless it is trivial to take a quote from Wikipedia, put quotes around it, mention the source and so maintain both the integrity of your own project and that of Wikipedia. Not so with OpenStreetMap data. What is the project information that should properly be given to OpenStreetMap and what is none of OpenStreetMap's business?
So, we come to a third pioneering wave of Open licensing. How to license a publicly made collection of factual or mostly factual data ... a database? It so happens that OpenStreetMap IS the pioneer here. There are no obvious precedents, no long-standing open project track histories to point at or compare.
So viral or non-viral? Or put it another way, Share-Alike versus PD (Public Domain)? Which will make OpenStreetMap stronger in the long term?
The truth is, we don't know yet.
Now to be frank, I personally really preferred the PD approach. CC0, http://creativecommons.org/publicdomain/zero/1.0/, is less than half a page long. Really easy to understand and implement. Then I and we can just go and do some mapping. Cool. It clearly meets that criteria on the OSM front page. School kids can use OpenStreetMap data without their attorney being present. Google can present maps made from our data on their website. So can Microsoft. Cool. I've no bias against commercial use. To me, anyone means anyone. To me, trying to tell people how they can and cannot use our data is holding back people. Data is not software. It is not easy to snatch it (and the hundreds of thousands of hours that went into collecting it), add something neat and copyrightable to it and then sell an unassailably superior product. The OpenStreetMap response is simple, just collect more data. Then give the poor, wretched company the job of trying to merge it all back in. I've worked on trying merge UNIX software trees, its a nightmare.
Problem 1. What if I am wrong in those last sentences above? What if there IS an advantage of Share Alike that I am missing? Time will tell. We have Share Alike now. We can only give it away once. Once it is gone it is gone.
Problem 2. And I think this is the most important one. A lot of my good friends in OpenStreetMap are very attached to Share Alike and will offer intelligent reasons why it is a Good Idea. I've focused so far on USING OpenStreetMap data. But to use OpenStreetMap data presupposes that someone CONTRIBUTES data, continue to contribute, continues to improve it, continues to encourage others to contribute. That only happens if folks are happy contributing. Fun is actually a serious business for OpenStreetMap. We sometimes forget that.
In our current stage of development, sustaining and increasing contribution is still more important than encouraging use. I see that window continuing for another two years.
So, I summarise our current condition as:
- We have a By Attribution, Share-Alike license.
- It may not actually protect our data.
- It is very hard for potential end users to interpret when applied to OpenStreetMap data.
- We need to do something about that.
- As a community, we are split on whether we should by PD or Share-Alike.
- A single unified happy contributor community for putting data in is as important as the license for taking data out and using it.
- If we give away Share-Alike, we can only do that once. Now may not be the time, we will only know that in hindsight.
- Keep the By Attribution, Share-Alike philosophy but make it directly applicable to databases and to OpenStreetMap data.
- Think of the longer term. Poll all voters on their attitudes to different types of license but don't get into a fight now.
- Clarify what does and does not trigger Share-Alike provisions
- Tighten up the ability for OpenStreetMap to get actual data back via Share-Alike provisions
- Liberalise the ability to make maps from mixed-license sources to encourage use of OpenStreetMap and encourage innovation.
- Strengthen copyright only with copyright, contract and database rights to the extent that they are applicable in different countries around the world.
- Everyone owns the data they contribute as before but they additionally they license it to OSMF to, in turn, license and publish collectively. If end users have questions about the license intent, the OSMF can consult the contributor communuty and provide one answer.
- Lock in the requirement for a free and open license, so preventing commercial hi-jacking.
- Define a practical, workable process for future evolution of our free and open license without dictating to our future selves what that should be.
- Ask OSMF members if all this makes sense. OSMF members represent a wide range of opinion, have generally followed and are well-informed about the complex issues involved and are passionate and care about mapping and about the ideals of OpenStreetMap ... whether or not they agree with me.
- Implement the new license and contributor terms.
- Try and make share alike work. If it does, great. Continue to discuss alternatives in case it does not.
GNS Placenames to OSM Conversion Script(offline version, Perl, 0.5 API)
Country Bounds - a list of 249 countries and dependencies with rectangular bounding and centre lat/lon coordates. Generated by taking the maximum and minimum values for each country from the US government GNS Placenames database.
PGS Coastline to OSM Conversion Script (offline version, Perl, 0.4 API - OUT OF DATE)