Norway Discusses PSI Licensing Model

Norway Discusses Licensing Model

tonzijlstra's picture
Share this: 

The Norwegian government is working on a set of licenses to use with open government data sets. It is modeled somewhat like the Creative Commons licensing vehicle where authors pick and choose elements as building blocks and so build a complete license.

On the blog of the current thinking about open government data licenses is explained and readers are invited to give feedback. Importantly, the Norwegian government takes 'free use' (i.e. no license terms attached) as the default for published data sets. At the same time the design is such that data owners have large measure of flexibility with how they formulate license terms for each data set.

Some General Terms are proposed regardless of any other decision the data owner makes upon publishing data. The three elements, also valid under 'free use', in the general terms are:

  • You may not pretend to be a service of the government department when you build a service on data from that department.
  • You may not use the data in ways that are in conflict with Norwegian law.
  • The source of the data has a right to be treated with respect (but this does not preclude criticism).

The first set of optional terms proposed is to safeguard what are the 'moral authorship rights' in copyright:

  • Attribution of the data source
  • No warranty, data source disclaims any responsibility for the use of the data
  • Obligation to register, so that data sources can see how their data is being used.

For terms of use the following aspects are proposed:

  • No changes to data, meaning you cannot change records and pass it of as original. In a very stringent form, in conflict with open data definitions it also may disallow the combination of data with other data sources.
  • Share Alike, meaning you may pass on the data only with the same conditions attached to it as you got it.

Access Terms are also suggested:

  • Interval, to be able to specify the interval in which the same data can be requested again. Mainly to limit the load on data servers. Interval is suggested to be set according to the update frequency of the data itself.
  • Update, which specifies how often the data reuser should update the dataset. This to ensure that data reusers will not be working with outdated data.

Finally there are Delivery Terms:

  • Commitment, specifying for instance how long a datasource guarantees availability in the same place and to which extend (think of guaranteed up-times, update frequencies)
  • Versioning, specifying that updates to data will have a version number in the URLs used and also to prevent reuse to fail because the datasource suddenly changes formats for instance.

The idea is that data owners can pick and choose the building blocks they deem appropriate for a specific data set. Specifying the conditions as "Free Use" means only the General Terms are applied to the data set.