[OSM-legal-talk] Is the "data share-alike" road navigable?

[Richard Fairhurst]

Mike Collinson wrote:

> I think this only works for the cottage cartographer?
>
> As I understand, this is  varying the original proposal as follows:
>
> 1. Database of GPS traces - public domain, possibly with a license  
> to get around EU database rights to achieve the same effect
>
> 2. Database of nodes, etc. - share-alike licensed
>
> 3. Maps produced from data - up to the creator (proprietary, share- 
> alike, whatever), but you must release the source material used  
> back under the same database license above.

That's accurate, yes. Otherwise it's not really share-alike and you  
might as well put OSM under an attribution licence (which would make  
me happy, but not some others!).

But I think with many of the examples you give, we're really talking  
about Imi's generally accepted 'layer' clarification/exemption again  
(2.2 in the wiki legal FAQ), which says that you can overlay a non- 
share-alike layer on a share-alike OSM map.

AIUI this relies on the difference between a "derivative work", which  
has to be shared-alike under our current licence; and a "collective  
work", which doesn't. IMO the licence doesn't make explicit that  
layers are collective rather than derivative, and even if we stick  
with CC-BY-SA we should find some failsafe way of ensuring the will  
of the community is backed by the licence text.

To take them one by one:

> For someone whose value add is making good maps, it works providing  
> they are not using other licensed data or a client's private data.
>
> For the Reading University case (sorry I forget your name) it only  
> partially works.  Creating/enhancing the base OSM Reading street  
> data directly makes sense for everyone.  But if the client has to  
> be persuaded to give up extra private data, it is probably easier  
> just to use commercial data as a base.

Yes, true. AIUI Tom wouldn't object to this anyway as a share-alike  
advocate.

At some point, assuming that certain other data _can't_ be shared  
(e.g. OS), the cartographer has to make a judgement call as to  
whether it's better just to buy in OS data, or to use OSM data and  
fill in the missing 20% him/herself. When the latter becomes more  
attractive, we've won - share-alike has given us a big chunk of new  
data.

> For ITN, it doesn't work if they combined any Google data because  
> ITN cannot release that source material.

Under the current licence, the ITN use can only work given Imi's  
layer clarification. Given that a Google satellite background is in  
some ways distinct from OSM data superimposed on it, you _might_ be  
able to say they're different layers. But probably not: a video  
presentation is a "flattened" image, the layers can't be extracted  
individually.

Under a putative share-data licence, though, it's actually easier. We  
put the OSM database of nodes/segments/ways (Tom's item 2) under  
share-alike. Anything added to that has to be shared-alike. But a  
satellite backdrop isn't anything like nodes/segments/ways, so -  
depending how tightly we draw the licence - we may not need to  
require that it is shared.

> For Nestoria, http://www.nestoria.co.uk/cowes/property/buy , it  
> does not work because they have to release all their property  
> listing information displayed on their map to competitors for free.

Imi's layer clarification again. :)

> For Multimap, http://www.multimap.com, I don't know how they are  
> using OSM data. Would someone clarify?

I believe they're considering using it as a "community map" layer.  
Depending how this is implemented, it's either aok anyway, or relies  
on Imi's layer clarification.

cheers
Richard
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/legal-talk/attachments/20070309/6d20872b/attachment.html>