At 10:11 AM 13/08/2010, 80n wrote:
>On Thu, Aug 12, 2010 at 11:47 PM, David Groom 
><<>> wrote:
>b) Ignoring the Yahoo data, but taking any data that may have had a PD or 
>CC-BY-SA clause that has be used in import, since these are general 
>permissions given and they do not explicitly mention granting rights to use in 
>OSM, I cant possible agree that I have EXPLICIT permission to use them. I have 
>permission by virtue of they are PD or CC-BY-SA, but not EXPLICIT permission 
>to do so.
>David, I don't think that CC-BY-SA is compatible with ODbL, nor with the 
>Contributor Terms.  If you have added content that is licensed under CC-BY-SA 
>you cannot agree to the Contrbutor Terms.  
>I'm sure you know that but your statement above suggests that CC-BY-SA is 
>compatible with OBdL and CT.  It is not.

I have moved this from "[OSM-talk] Voluntary re-licensing begins"  to legal 
talk as it is worth further discussion in view of dilemmas faced by our 
Australian community.  I understand that CC-BY-SA is currently a preferred 
vehicle for releasing government data.  I am inclined to agree with 80n, though 
in the context that CC-BY-SA licenses on data are just too potentially broad in 
their virality. I present this for the purposes of discussion and do not see my 
conclusions as immutable. I focus on Share-Alike, though Attribution is also a 

I would also like to note that I am having an email dialogue with Ben Last of 
NearMap of Australia (  They allow use of  their 
PhotoMaps to derive information (e.g. StreetMap data) under a Creative Commons 
Attribution Share Alike (CC-BY-SA) licence. They are being most cordial and 
helpful. They are submitting the ODbL for legal review from their own 
perspective.  I hope they will share some of the conclusions they reach, both 
for the perspective and the authoritative opinion.


To grossly paraphrase, a GNU type software license it works like this:

Write a word processor  --X-->  Write a book with the software.

Virality remains in the software, it is NOT transmitted to the book. It IS 
possible to use other non-compatible software to make the book.  But if the 
software is improved to write the book and software is published, then software 
improvements must be available Share Alike.

ODbL is slightly stronger:

Create map data --X--> Make a map

Virality remains in the data, it is not transmitted to the map except in 
reverse engineering out the data. It is possible to use other non-compatible 
data to make the map under certain conditions.  But if the data is improved and 
the map or the data is published, then data improvements must be available 
Share Alike.

But if CC-BY-SA license is used to try on information rather than the virus can 
potentially just keep on going. It all depends on what the original publisher 
feels they want to exert(?).

Here is a real dilemma being faced by the Australian community:

Aerial imagery under CC-BY-SA  -----> Create map data with some imagery tracing 
-----> Pull out a single lat/lon and put it in a book; make a map; ...

ODbL breaks the chain at the second "----->", either because the extract is not 
substantial or because the right-hand item is a Produced Work. CC-BY-SA does 
not, or at least you'll need to clarify with the original publisher(?).

Personal conclusion: The CC-BY-SA license are great on fully creative works.  
It was never intended to be applied to highly factual data and information, and 
if it is, it is vague and confusing.  If you believe strongly in  pandemic 
virality, then it is a good thing.  If you believe that all the chain of 
Share-Alike and Attribution should be far more constrained, then it is just 
dangerous and should be avoided. Which is why most of us want to move away from 
it as our own license. Our primary goal is disseminating data we collect 


legal-talk mailing list

Reply via email to