[okfn-discuss] CKAN questions

[Jonathan Gray]

Francis,

It would be great if you could add it to CKAN!

While the focus is on datasets that are 'open' as in the OKD - we've 
been registering non-open stuff too. Especially with large and important 
datasets that are freely accessible but have either no terms of use, or 
terms of use that restrict redistribution/re-use - such as many large 
institutional datasets.

I too am particularly interested in the datasets held by the UN and 
other international institutions. A little while ago I met someone who 
worked there who said he'd try and put me in touch with whoever he could 
to discuss issues relating to how their data is licensed, whether they 
assert rights, terms and conditions etc. I haven't yet followed this up 
- but perhaps we could start a mini-project on UN data?

Talking of institutional policies, I've long been interested in data 
produced by US government agencies (such as NASA), much of which must be 
in the public domain and effectively open. It would be great if relevant 
departments could clarify the legal status of the data they 
produce/distribute. It would also be good, as there's so much data out 
there, to try to map out what government departments publish data - and 
what kinds of data they produce.

I wonder if any of the recently announced open data protocols and 
licenses will be adopted by international/governmental data producers:

http://blog.okfn.org/2007/12/17/good-news-for-open-data-protocol-for-implementing-open-access-data-open-data-commons-pddl-and-cczero/

Anyhow, regarding your questions:

 > Do I just add it as "Non-OKD Compliant: Other" on CKAN? And try and
 > contact them to complain about the lack of a license / public domain
 > assertion?

This is pretty much what we've been doing so far. Personally I've not so 
much been complaining as enquiring:

  * what license/terms of use apply and/or whether they are asserting 
any rights in their data;
  * what their attitude towards re-use is;
  * if they are restricting re-use to 'personal' or 'noncommercial' use 
-  what the rationale behind this is;
  * whether they have considered making their data open (and why doing 
so could be a good idea).

Now, in relation to the last point, as well as pointing out the Open 
Knowledge Definition [1], the Open Data web buttons [2], and the Guide 
to Data Licensing [3], we can flag up the new protocols and licenses 
mentioned above (PFIOAD, ODC PDDL, CCZero)!

[1] http://www.opendefinition.org/
[2] http://www.opendefinition.org/buttons
[3] http://www.okfn.org/wiki/OpenDataLicensing

I'm thinking of revamping our communications page soon and will 
certainly bear this in mind... Perhaps we should also put something 
about this on CKAN.

 > What do we do in a world without the culture of good licensing?  Both
 > to make that culture, and to deal with it meanwhile.

(I've taken 'good licensing' to mean making it clear what others can do 
with data, rather than being liberal about what others can do with the 
data...)

As well as contacting those responsible for the data, for larger data 
producers I think it could be interesting to start mini projects on a 
'per organisation/government' basis (as mentioned above).

E.g. mapping out what data is produced by UN agencies, or US government 
departments, how this may be re-used, and what legislation/policies are 
relevant. In other words - helping to document what kind of culture 
currently exists around data sharing. Having a load of entries on CKAN 
is a good start. Perhaps this would be a good area of commonality with 
some of the recent open government projects/groups.

(N.B. This is a bit similar to what a group of us have tried to start 
doing for environmental datasets with: 
<http://www.okfn.org/wiki/OpenEnvironmentalData>)

In terms of encouraging a culture of sharing - I guess its always good 
to have more/better 'exemplar' projects that show what can be done with 
the data if others are allowed to re-use it. (One idea I've been 
*really*  keen on for a while is in visualising public data a la 
gapminder, swivel, many eyes, etc.)

Regards,

Jonathan


Francis Irving wrote:
> Just found a new data set (pharmaceuticals prices for international
> development), and thought I'd add it to CKAN.
> 
> http://www.who.int/hiv/amds/price/hdd/index.aspx
> (you can search with all wildcards to get all the data, although it is
> a huge page)
> 
> The problem is I don't know the license. The WHO site isn't very
> helpful, to the extent of not even having a contact form.
> 
> Now, clearly this is still a useful dataset. It would seem a shame
> not to add it to CKAN just because it has no clear license. Because
> most good data won't have such a license in our current data licensing
> climate.
> 
> Do I just add it as "Non-OKD Compliant: Other" on CKAN? And try and
> contact them to complain about the lack of a license / public domain
> assertion?
> 
> What do we do in a world without the culture of good licensing?  Both
> to make that culture, and to deal with it meanwhile.
> 
> Francis
> 
> 
> _______________________________________________
> okfn-discuss mailing list
> okfn-discuss at lists.okfn.org
> http://lists.okfn.org/cgi-bin/mailman/listinfo/okfn-discuss