[Talk-GB] OS OpenData Licence update (WAS: Finding Unmapped public rights of way)

Michael Collinson mike at ayeltd.biz
Tue Jul 30 13:58:50 UTC 2013

On 30/07/2013 11:49, osm at k3v.eu wrote:
> Robert,
> On Tue, 30 Jul 2013 08:57:13 +0100, "Robert Whittaker (OSM lists)"
> <robert.whittaker+osm at gmail.com> wrote:
>> ...
>> OSM takes a conservative line on copyright and licensing issues...
> I agree with Rob 100% on this, it is pretty obvious that the Government
> intends for this data to be freely usable by businesses and projects
> like OSM. This has been covered to death a number of times in the past.
> There is a lot of external data in OSM that requires attribution. The way
> the project handles that seems to work pretty well regardless of what
> the amateur lawyers may say.
> It is very hard to imagine the circumstances where OSM would face any
> issues from using these datasets and if that were to occur then the
> data can be removed as has happened when other data sources have been
> challenged in the past.
> If you don't want to use this data in OSM then don't use it but you are
> not the arbiter deciding what others may do.

I would add to this that as Robert W is quite right that "OSM takes a 
conservative line on copyright and licensing issues", the Licensing 
Working Group formally made the Ordnance Survey aware of our (then) 
intended use under ODbL and explicitly pointed out where there where 
potential incompatibilities.  The upshot was that the OS kindly made a 
formal declaration that they had no objections to such use for all 
OpenData product where they have complete IP control, i.e. everything 
except CodePoint data. Since that time, the OpenStreetMap Foundation, as 
formal publishers of the database, have had no communication from the OS 
rescinding that for future releases of OpenData. In other words, if the 
Man Says Yes, then the Man Says Yes.

Now, the potential incompatibility with the OS OpenData License per se 
has never been removed. This means that there are problems for the OSM 
community and the general public in these *other* areas:

- Use of OS OpenData other than that described above, i.e. CodePoint. [I 
personally feel that the real problem all along is the Royal Mail and 
their apparent decision to hijack post code and address databases paid 
for with public resources into the private sector. Chris Hill has been 
working here but the LWG informally feels that the response he got is 
deliberately vague and obfuscatory.]

- Confusing use of the OS OpenData License instead of the Open 
Government License on other datasets. As I recall that is datasets from 
English Heritage.

A number of individuals have been working on these and other issues, at 
least Robert Whitaker, Rob Nickerson and Chris Hill. I apologise to them 
that the LWG has not been in a position due to lack of manpower to give 
support despite requests to do so. I therefore suggest the following:

We (all) take a simple unified stance that:

1) The Open Government License, OGL, was deliberately brought into being 
to provide a consistent, harmonious platform for releasing open 
government-funded and government-owned data. The OS OpenData License is 
clearly at odds with this and should be retired completely. Anything 
currently under the OS license should re-published under OGL at the 
earliest opportunity.

2) There is a set of other key datasets which we believe need to be 
unequivocally published under OGL for the public good:

     - PROW data, however provided.
     - "Royal Mail" address database, (I am shaky on the details on 
this, Robert Barr is the man to talk to).
    - Others you may identify.

I, and I believe all other LWG members, will be at SOTM 2013 in 
Birmingham.  I suggest that we all meet up then.  If possible, I'd like 
to make points 1 and 2 as a formal LWG/OSMF submission to ODUG before 
then.  However, I want to be sure that I get all my facts straight, and 
lack of time to read everything up is what is stopping me right now.  
Any comments/support greatly welcomed.


More information about the Talk-GB mailing list