Documents, copyright and database right

In order, because each step has a different answer in law:

1. We fetch it, where the publisher offers a route to it. Where they do not, we do not go round them.
2. We keep the copy we fetched, internally, so the same document does not have to be fetched again every time we improve how we read it. That copy is never shown to anyone and never served.
3. We read facts out of it — what is being built, roughly how much of what, which trades it implies.
4. We publish the facts, graded, so a customer can see which of them the publisher stated and which we worked out.
5. We show the customer the document’s title and the publisher’s own link to it. We do not host a copy for download and we do not serve one through this site. If a customer wants the document, they get it from the body that published it, on that body’s terms.

Copyright protects a work — the drawing, the specification, the schedule, as the person who made it expressed it. It is the author’s, or their employer’s, or the Crown’s, and nothing about us reading it changes that.

Publishing that a scheme involves a particular quantity of a particular material is not publishing the drawing. What we put in front of a customer is a set of facts and our own reading of them, written in our own words. We are not asserting any right over the underlying document, we do not claim to license it to anybody, and buying a package from us gives a customer no rights in it whatever.

So if you want the document itself, get it from the publisher. Our link points there for that reason. What you may then do with it is between you and them.

It is worth being direct about one thing, because it is the provision a reader might expect us to be relying on and we are not.

Section 29A of the Copyright, Designs and Patents Act 1988 allows a copy of a work to be made for computational analysis. It does not apply to us. It is limited by section 29A(1)(a) to analysis “for the sole purpose of research for a non-commercial purpose”, and we read these documents in order to sell what we find. We do not rely on it and we are not going to describe what we do as text and data mining under it.

The same is true of the equivalent provision for databases. Regulation 20(1) of the Copyright and Rights in Databases Regulations 1997 requires the extraction to be for teaching or research “and not for any commercial purpose”. It is not available to us either.

What permits the reading is the licence each publisher grants over what it publishes. For most of the material that is the Open Government Licence; the acknowledgements the licences require are at Tritium 3.1 and licences.

The Open Government Licence is generous and it is not unlimited. Its own exemptions section says the licence does not cover, among other things:

• “personal data in the Information”;
• departmental and public sector logos, crests and the Royal Arms, except where they form an integral part of a document or dataset;
• third party rights the Information Provider is not authorised to license;
• other intellectual property rights, including patents, trade marks and design rights.

The first of those matters more than it looks. A planning notice may be licensed while the case officer named in it is not. The licence is the publisher’s permission over the information; it is not, and could not be, anybody’s permission in respect of a person named in it. Permission for that has to come from data protection law and from nowhere else — which is one of the reasons named individuals were removed from what this product sells. Attribution does not cure it: an attribution condition is a condition on the licensed material and does not extend the licence to the exempted material.

The practical consequence for a customer is in the next section, and it is short: a name you take off a linked source page is not covered by anything we have.

Buying a package does not transfer any right in the underlying documents or notices to you, because we have none to transfer. What you buy is our compilation and our reading of published material, for your own use in pursuing work.

Your own reuse of the published material is subject to the same licences that govern ours. In particular: if you follow one of our links to a source page, read a case officer’s name off it and build a list, you are not covered by our licences and you are not covered by the Open Government Licence either, because personal data is outside it. That is your own processing and your own responsibility — see which contact details Tradello holds.

Republishing our compilation as a product of your own is a different question again, and the terms of sale are where it belongs. The operative wording for it is with a solicitor.

Beyond copyright in a document there is a separate property right in a database. Under regulation 13 of the Copyright and Rights in Databases Regulations 1997 it subsists where there has been “a substantial investment in obtaining, verifying or presenting the contents”, and regulation 16(1) makes it an infringement to extract or re-utilise “all or a substantial part of the contents” without consent.

Regulation 16(2) is the one that matters to a business like ours: “the repeated and systematic extraction or re-utilisation of insubstantial parts of the contents of a database may amount to the extraction or re-utilisation of a substantial part”. A daily automated read of somebody else’s index is exactly the case that provision describes, and the fact that each individual read is small is not an answer to it.

We record this here rather than reassure about it. Where a publisher licenses reuse — as the open government sources do — the consent required by regulation 16 is given. Where one has not been read, nothing has been established, and the honest list of those is below.

This section exists because the alternative is a page that sounds more confident than the facts support. None of the following is settled, and we are not going to write a sentence that assumes any of it.

• Most council planning portals. A survey of 127 local planning authorities found no machine route at all for 113 of them, which means no published licence anybody has read for those 113. Anything said about council portals as a group is a generalisation, not a licence position.
• The licence version on two of the national services. The Open Government Licence is established for them; a page stating which version applies specifically was not found, so no version is asserted for those two.
• Whether attribution is required for European notice data, or merely conventional. We give the credit either way and describe it as given rather than as required.
• Third-party company websites. We read published addresses from organisations’ own sites and honour their automated-access files. Whether any individual site’s terms of use prohibit it is a per-site question nobody has surveyed.
• One paid commercial feed’s terms, unread. A paid commercial feed exists whose terms nobody has read. It is disabled, it has never delivered a record, and nothing derived from it has ever been sold. It stays disabled until the terms have been read.

If you publish something we have read and you think we have gone further than your licence allows — or you are a rights holder in a document a council published — write to us and say which document and which right. We would rather be told and fix it than argue about it. The address is at the foot of this page, and help and contact has the other routes.