[ckan-discuss] Licensing open data

Fri Oct 25 22:23:35 BST 2013

On 10/02/2013 01:35 PM, Thomas Levine wrote:
> First, you might be interested in my recent discovery that a lot of 
> open data isn't openly licensed.
> http://thomaslevine.com/!/open-data-licensing 
> <http://thomaslevine.com/%21/open-data-licensing>
>
> Second, I'm slightly curious as to why this is. I'm sure that some 
> people have thought a lot about whether/how to license open data; 
> could someone point me to some literature on this?
>

Hey, Tom. I meant to reply to this earlier.... I'm going to reply from 
two points of view.

Putting on my advocacy hat-

I've been working with Eric Mill (Sunlight), Jonathan Gray (OKF), and 
others on a fairly technical policy document about how the U.S. 
government should do data licensing. Namely, don't license: use CC0, and 
we provide suggested language for how to do that. While it's written for 
U.S. gov agencies, some of it is applicable for any open data project. 
Our latest final copy is here:

http://theunitedstates.io/licensing/

Our recommendations have already been used on two gov open source 
projects. We're working on a draft for a version 2.0 here:

https://github.com/unitedstates/licensing/pull/1

and would be glad to have anyone else sign on to the document. Also 
anyone should feel free to submit issues and pull requests on the 
document. I'd love also to see a fork for an international version.

Putting on my gov contractor hat from experience working on HealthData.gov-

There are a lot of reasons why a dataset wouldn't be listed as openly 
licensed. The most common reason I've seen is that the question of 
licensing wasn't a question at the time a dataset was entered into the 
data catalog, and datasets are not often updated once they are entered. 
 From there, data owners at agencies may not know the licensing status 
of a dataset because they may not know where the dataset came from with 
enough certainty to be able to say it is public domain or something 
else. There are also a good number of datasets produced in complex ways 
through public-private partnerships and, of course, gov contractors 
which, at least in the U.S. makes ownership and licensing extremely 
complex. Often those datasets are really not open. So it's varied, and 
in some cases actually really difficult to figure out what to do.

-- 
- Joshua Tauberer
- http://razor.occams.info

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.okfn.org/pipermail/ckan-discuss/attachments/20131025/d793fb89/attachment.htm>