Graeme Wilford wrote on 09/06/2010 11:05:
> All this talk about licensing and attribution caused me to question 
> the following:
>
> If I modify existing (or add new) OSM primatives using other 
> primatives as a guideline and the guiding primatives are attributed, 
> my changes might be considered derived works. Should the attribution 
> be inherited?
>
> To give a couple of simple examples;
> - a building shape is added from, say OSSV and appropriately 
> attributed and then I move an existing POI (or apply a new one) on to 
> the building
> - a lake is added from say, OSVM or an aerial survey and I draw a new 
> or re-align an existing footpath around the lake
>
> Any views?
>
> Thanks,
> Wilf.
>
> m...@home
>
> http://facebook.com/gwilford
>
>
>
>    
It is the impracticalities of this that mean that there is a morally 
correct answer - which has the potential for you applying a source of 
local_knowledge;OSSV;survey;GPS to properly attribute the many things 
that you amalgamated in making the edit - or leaning on the licensing 
wording which talks of reasonableness a bit harder, where you consider 
only the major issues. I'd say aligning a footpath around a lake is 
really not depending on the source - you probably are not making a 
judgement as to accuracy and depending on it, you are making a 
cartographic judgement of a pragmatic nature rather than depending on 
the data - e.g. you have made a judgement that the footpath source is 
less reliable than OSSV in this case or you'd be hacking the lake not 
the footpath.

Legally, specifically thinking of OSSV, this is why it is important to 
resolve the fairly straightforward issue of the correct overall 
attribution. Either OSM accepts that it has content derived from OSSV 
and applies the minimal and reasonable attribution to the product, then 
the issue of individual attribution is not legal but interesting 
information for other editors to judge their edits against when making 
amendments (e.g. my source is local knowledge, I will defer to OSSV or 
not depending on, say, other landmarks); or alternatively OSM as a 
project refuses to acknowledge the use of OSSV etc. as a source, and the 
issue is void, as you cannot comply with the OS licence and any 
facilities to use OSSV should be stripped from the project. An end user 
is never aware of the individual attribution of a feature.

If the project rejected OS data, with its most mild attribution 
requirements, along with a licence to use the derived data in any way in 
perpetuity, then I think it also starts having a problem with using out 
of copyright data, as it would really be saying that it is only 
interested in data collected through members' own efforts, and in that 
regard, using out of copyright OS maps seems little different to 
generously licensed OS data.

Spenny

_______________________________________________
Talk-GB mailing list
[email protected]
http://lists.openstreetmap.org/listinfo/talk-gb

Reply via email to