I’ve written a guide about applying licenses, waivers or public domain marks using the Frictionless Data specification. Your feedback is very welcome.
Applying licenses, waivers or public domain marks
Applying licenses, waivers or public domain marks to data packages and data resources helps people understand how they can use, modify and share the contents of a data package.
In this guide, “license”, can be read as “license, waiver or public domain mark”.
It is recommended to apply a license to a data package. This license applies to all the data, files and metadata in the data package unless specified otherwise.
You can optionally apply a license to a data resource. This allows a license that differs from the data package license to be applied to the data resource. If a license is not specified, the data resource inherits the license from the data package.
Specifying a license
The Frictionless Data specification states that a license must contain a name property and/or a path property and may contain a title property.
name: The name MUST be an Open Definition license IDpath: A url-or-path string, that is a fully qualified HTTP address, or a relative POSIX pathtitle: A human-readable title
You can specify the location of a license using a URL or a Path.
Specify a license using a URL
To specify a license using a URL, use the fully qualified HTTP address as the value in the path property, e.g.
"licenses": [{
"path": "https://cdla.io/sharing-1-0/",
"title": "Community Data License Agreement – Sharing, Version 1.0"
}]
Specify a license using a Path
To specify a license using a path, use a relative POSIX path to the file in the data package as the value in the path property, e.g.
"licenses": [{
"path": "LICENSE.pdf"
}]
In this example, LICENSE.pdf would be in the root of the data package folder, e.g.
folder
|- datapackage.json
|- LICENSE.pdf
|- README.md
|- data
|- data.csv
|- reference-data.csv
It is recommended that the licence is provided in markdown format to simplify its display in data platforms and other software.
The license can be a separate file or included in the README.md file. If license information is included in the README.md file, it is recommended that it follows the guide for formatting a README file.
Applying a license
These scenarios apply to either the data package or a data resource.
- Apply an open license
- Apply a non-open license
- Apply a waiver
- Apply a public domain mark
- Do not apply a license
Other considerations:
- Provide additional license information
- Copyright belongs to multiple parties
- License may become legally binding
- Software may not fully support the Frictionless Data specification
Apply an open license
For an open license, use name, path and title, e.g.
"licenses": [{
"name": "CC-BY-4.0",
"path": "https://creativecommons.org/licenses/by/4.0/",
"title": "Creative Commons Attribution 4.0"
}]
name must be an Open Definition license ID however note that some license IDs are placeholders or have been retired and should not be used, e.g. other-at, other-open, other-pd, notspecified, ukcrown-withrights.
Apply a non-open license
To apply an non-open license, use the path and optionally the title properties. It is preferred that the license is published at a URL (a fully qualified HTTP address), e.g.
"licenses": [{
"path": "https://creativecommons.org/licenses/by-nc-nd/4.0/",
"title": "Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0)"
}]
If the license is not available at a URL, you can specify a license using a path.
Apply a waiver
You can indicate that copyright has been waived by referencing a waiver at a URL in the path property, e.g.
"licenses": [{
"name": "CC0-1.0"
"path": "https://creativecommons.org/publicdomain/zero/1.0/",
"title": "CC0 1.0"
}]
If the waiver is not available at a URL, you can specify a waiver using a path.
Apply a public domain mark
You can indicate that there is no copyright in the data or that copyright has expired, using the public domain mark or other public domain dedications, e.g.
"licenses": [{
"path": "http://creativecommons.org/publicdomain/mark/1.0/",
"title": "Public Domain Mark"
}]
If the public domain dedication is not available at a URL, you can specify the public domain dedication using a path.
Do not apply a license
If you have not decided what license to apply but still want to publish the data package, describe the situation in a file in the data package, e.g.
"licenses": [{
"path": "README.md"
}]
Other considerations
Provide additional license information
It can be helpful to data consumers to provide additional copyright or attribution information such as:
- copyright notice - this property allows a data publisher to specify a short copyright notice that can be used directly
- copyright statement - this property allows you specify a URL for the copyright statement
- preferred attribution text - the text to be used when attributing the creator of some data
- attribution URL - this is the URL to be used when building an attribution link
This is explained in the ODI Publisher’s Guide to the Open Data Rights Statement Vocabulary and Re-users Guide to the Open Data Rights Statement Vocabulary.
Some licenses require that data consumers provide the copyright notice in the attribution (e.g. CC BY 4.0 Section 3).
Some data publishers may waive some of their rights under a license, e.g.
Noosa Wedding Locations data by Noosa Shire Council is licensed under a Creative Commons Attribution 4.0 licence.
Noosa Shire Council waives the requirements of attribution under this licence, for this data.
You can include this information, either:
- in the file containing license information (e.g. README.md)
- as metadata properties in the datapackage.json
The data package specification supports adding addition metadata properties to the datapackage.json, e.g.
{
"name" : "coastal-data-system-near-real-time-wave-data",
"title" : "Coastal Data System – Near real time wave data",
"licenses" : [{
"name": "CC-BY-4.0",
"path": "https://creativecommons.org/licenses/by/4.0/",
"title": "Creative Commons Attribution 4.0"
}],
"copyrightNotice": "© The State of Queensland 1995–2017",
"copyrightStatement": "https://www.qld.gov.au/legal/copyright",
"attributionText": "Science, Information Technology and Innovation, Queensland Government, Coastal Data System – Near real time wave data, licensed under Creative Commons Attribution 4.0 sourced on 26 December 2017",
"resources": [
{
"path": "https://data.qld.gov.au/dataset/coastal-data-system-near-real-time-wave-data",
...
}
]
}
Copyright belongs to multiple parties
Sometimes data in a resource may be combined from multiple sources that are licensed in different ways. You can indicate this by placing two or more licenses in the licenses property. Further explaination should be given in the README.md.
"licenses": [{
"name": "ODC-PDDL-1.0",
"path": "http://opendatacommons.org/licenses/pddl/",
"title": "Open Data Commons Public Domain Dedication and License v1.0"
},
{
"name": "CC-BY-SA-4.0",
"path": "https://creativecommons.org/licenses/by-sa/4.0/",
"title": "Creative Commons Attribution Share-Alike 4.0"
}]
License may become legally binding
The specification for licenses states:
This property is not legally binding and does not guarantee the package is licensed under the terms defined in this property.
A data package may be uploaded to a data platform and the licenses applied to the data resources and publicly displayed. This may make the license legally binding. Please check your specific situation before publishing the data.
Software may not fully support the Frictionless Data specification
Be aware that some data platforms or software do not fully support the Frictionless Data specification. This may result in license information being lost or other issues.
Always test your data publication to ensure you communicate the correct license information.
As examples:
-
Data Curator only allows the user to select from a limited set of open licenses to describe the data package and data resource licenses.
