How complex can this Derived Data thingmy be? 11

Gosh we seem to have been talking about Derived Data for such a long time and still we don’t seem to making much progress. Every time government reviews Ordnance Survey it stumbles on the thorny subject of Derived Data.

Ken Field's OS License Shirt (pic by me)

Nearly 18 months ago when the short lived new business strategy for OS was launched by a now departed minister there were promises that the issue would be resolved soon. I consulted for OS on GeoVation last year and had several conversations with people across the organisation all of whom explained how complex the problem was and assured me that they were working on reducing the extent of the problem by providing clarity on what was and wasn’t Derived Data, this is the Derived Data Dilemma – which I may refer to as DDD for the rest of this post. October 2009 was a target date but that slipped, then the PM launched his Making Public Data Public initiative and whacked a great hole in the OS ship by freeing up quite a lot of OS data. Many people raised DD as a/the key issue in the consultation at the beginning of the year, some were even foolish enough to expect a resolution in the response to the consultation (come on guys this is complicated). What we actually got was what sounded like a fairly strong commitment in the government response:

“3.24 Ordnance Survey will also be proposing changes to the derived data policy for the commercial sector, including ‘Free To Use’ data, as part of its work on revised pricing and licensing. Further details of when these changes will be implemented will be communicated later in the year. In addition, Ordnance Survey will work with Cabinet Office and the Office of Public Sector Information (OPSI) to agree guidelines and processes which would permit certain other datasets created by public bodies and containing licensed paid-for Ordnance Survey data to be made available for unrestricted, including commercial, re-use. In agreeing these guidelines, the parties will need to balance a number of issues, including the significance of the Ordnance Survey data taken, the impact of release of any dataset on Ordnance Survey’s commercial business and any legal or regulatory consequences for Ordnance Survey. This reflects concerns expressed in the consultation responses about licensing of Ordnance Survey products, and in particular derived data restrictions.”

Now read quickly this may not have sounded too bad, a bit woolly but the sentiment and general thrust were encouraging. Now read my italics again, look who is doing the proposing and also note the mention of a DD policy – the core of the DDD is that there is no policy or not in the sense that most of us would understand a policy, a document that we can refer to that perhaps includes definitions and procedures. Much of the confusion that exists around DD is due to the lack of clear definition of what is and isn’t derived data and what you are and aren’t allowed to do with it if it is derived.

So cut to yesterday and after a bit of a PR warm up on Twitter the OS publish an announcement about DD on their blog. You can decide for yourself whether this constitutes a resolution of the DDD for public sector bodies within the PSMA and any commercial organisations that want to use data that has been created by a PSMA participant using OS licensed mapping as a background. Sure DEfRA can share some of their data with farmers but what about releasing it to a voluntary or commercial body who wishes to analyse the data and either hold DeFRA to account or sell services to farmers to help them get the maximum possible subsidies and grants from DeFRA?  In this era of Making Public Data Public and increased transparency one might have hoped for a simple statement enabling people to get on with using, sharing, reusing and even adding value and possible making profit from public data derived in part from an OS map. Something along the lines suggested by Ed Parsons a while ago (wildly paraphrased by me) “if it isn’t on the map then we make no claim to it” would have been good. Perhaps I am being a little harsh and there is much more clarity available to PSMA participants, if so please OS can you share with partners, consultants and others who need to understand and advise our clients what is and isn’t possible in this new liberalised DD policy – a plain English document maybe?

An irrelevant picture from our holidays just to lighten this rant or perhaps to point out that anything is possible if this 2CV can travel 1500 miles from London to the Alps and back!

Mentioning Ed Parsons brings me to the second part of the OS blog post, despite claiming that the PSMA heralds a new era of data sharing and usage the same old roadblock remains if a council or police force wants to present its data to the public or internally using Google Maps. OS and Google have been unable to reach an agreement for several years now over the conflict between Google’s licensing and that of OS. I am not going to repeat the debate here if you don’t know what they are arguing about you can get a sense of it in the comments on the OS blog or on Ed’s blog. A substantial majority of people would prefer to use the Google Maps API to build their mashups (it’s free for public sites, has a lot of powerful functionality, is very scalable, has global coverage etc), they don’t want to be told that they can use another API which will cost them money or offer less capability (NB statement of interest – I have an involvement in a company that creates webmaps on both Google and Bing’s API’s). There must be a resolution to this, I can’t believe that Google wants to misappropriate any OS IPR and I suspect that the clause in the Google API that OS are so concerned about exists in other API’s but in a slightly different format. I agree with the sensible comment from Thierry Gregorius:

“Would OS and Google care/dare to draft a joint statement on the derived data matter? If it is purely about interpretation of facts I don’t quite see what the difficulty is, except that communication via keynotes or blog posts is probably not sufficient to make it happen.”

Please guys, there is a mass of public geodata and creativity that are just waiting for the opportunity to combine for the public benefit, help us to make it happen.

And now it is time for a cup of strong coffee ..

11 thoughts on “How complex can this Derived Data thingmy be?

  • steven

    OS have just posted a blog update doc at

    Seems to be recapping the same stuff re Google T&C’s and setting out the changes or clarifications that they would like to see.

    They say

    “We hope this clarification is useful, pending the release of more detailed guidance on the PSMA and the development of our pricing and licensing review later this year.”

    Not sure that it is?!?

  • Elliot Hartley

    Sometimes when we talk about rights, licensing and derived data we forget that whilst private (sort of) the OS is wholly owned by the Government and therefore the taxpayer by extension.

    It seems absolutely ridiculous that other governmental departments and organisations, who supply some of their data to the OS have to buy it back. Or get in a panic about using it to provide services to the citizen. At the very least all tiers of government should be getting this for free and not worrying about licensing.

    Wasn’t the OS founded to provide data for the government (initially for military purposes) and not to make money?

    Whilst I take what Kevin has said seriously are we not forgetting that the OS is property of the UK and not an independent private company? This obsession with licensing is unnecessary, the OS can have a different remit from a private company and I think the UK gov and OS should be using it as the basis for innovation in all sectors.

    The OS is a fantastic organisation with a data product everyone wants to use, I just wish I could say let stop asking them what derived data is and start creating new services and datasets.

    Apologies if it read like a rant, it wasn’t meant to!

  • Kevin

    You may be right – obviously you would have much better perspective. However as you know, the issues of revenue and allocation of risk are closely linked, at least in the commercial marketplace. May not be an issue in this case.

  • Kevin

    My thoughts (I am a lawyer who works with geospatial companies):

    1. When you are talking Derived Data you are talking license agreements.
    2. Most organizations will have their lawyer review a license agreement before it sends it out.
    3. Lawyers see license agreements as more than a transfer in rights of IPR – also see them as an allocation of risk associated with matters such as liability, ownership, privacy, national security, IPR, etc.
    4. From a legal standpoint, you can’t avoid these issues by simplying ignoring them in the license agreement – ignoring them simply means you let someone else (courts, laws, etc.) allocate the risk for you. Lawyers really don’t like that.
    5. Many areas of the law with respect to spatial data are outdated, unclear and/or confusing.
    6. You technical folks are moving so fast in this area that lawyers are unable to keep up. (Spatial law is not talk in law school – yet!)
    7. Consequently, many lawyers really don’t understand what spatial data is, what it is being used for and how it is being used.
    8. Defining derivate products for spatial data in a license requires a good understanding of both the law and how spatial data can and will be used – both of which are limited at this point
    9. Given the confusion in the law and uncertainty with what spatial data is and does, it is easier for lawyers to say no than yes.
    10. Lawyers will get in more trouble if they say yes and they are wrong than if they say no and they are wrong.

    • steven


      I am sure that you are correct about the legal issues around risk and liability. However I don’t think the long standing roadblock around DD in the UK is due to these issues, it seems to be principally about protecting IPR (Crown Copyright) and maybe maximising revenue

  • blabyboy

    Elliot hits two nails with the last questions in his post.

    I would suspect that the OS bods are working on different permutations of definitions and then passing them:
    a) past lawyers to identify loopholes, pitfalls
    b) sales team to identify potential revenue loss against products affected – what definition costs them the least?

    IANAL etc.. so correct me if I’m wrong, but can’t policies be challenged in court? i.e. if it were seen to be discriminatory in some sense. Has no-one ever challenged the OS in court on this, to at least get a definition of the policy to be produced for consideration by the court against the plaintiff’s concern(s)?

    If we look at the part quoted by you, we should at least get an answer when the revised pricing is completed – a work that realistically cannot go in perpetuity… or can it?? 😉

  • Ken Field

    you’ll get me into trouble printing photos of ‘that’ shirt! but I’m pleased to see you’ve used a smart automatic blurring technique to avoid showing my face…the back of my head is far more photogenic!

  • Elliot Hartley

    Thanks for this post, I was initially heartened by the OS blog post, it seemed that maybe someone would start talking plain English in regards to Derived Data. Your take on it has made me sceptical again!

    I wonder if anyone has been brave enough to just start doing it and using PSMA members data (that may or may not be considered derived data). What would the OS do, especially in light of this ‘announcement’?

    Is the reason that the OS don’t release a clear statement on what is and isn’t possible with DD really that they themselves either don’t know or haven’t decided? Do they think they will open the flood gates and lose revenue?

    • steven


      A few organisations have already published some of their data using the Google Maps API (best if I don’t point them out). The earth still seems to be rotating smoothly around the sun and lawyers have not got any richer.

      Hope I haven’t invoked the curse of “speaking too soon”

Comments are closed.