Re: more copyright stuff

Matt White wrote:
> Andrew Loughhead wrote:
>   
>> But I think you are right Liz, it is relevant. I believe some osmers 
>> collect street names by exception, that is, they compare what a 
>> published map says with what street signs say, tick the confirmed ones, 
>> and when they see an exception, note it.  I think this makes their list 
>> a derived work and so is relatively high risk. 
>>     
> Derivation surely means taking data from one source and putting it into 
> another. Not including data is probably not derivation - otherwise (to 
> go to extremes), my blank sheet of paper is a derivative work of channel 
> Nine's program guide, due to the fact that I looked at it and chose to 
> omit all of it...
>   
Hmm. I think I can hear the creaking and twanging of an overstretched 
analogy here. :-).

Perhaps the method in use could be described as taking licensed data 
from a source and using it to decide where not to expend effort 
collecting original data. The effort management is then based on 
licensed data, and the question would be if new original data, produced 
through the smaller effort, is in some way tainted with the license 
conditions of that first source.  I don't know, despite my earlier 
assertion.  Maybe the story Liz referenced tells us to be risk averse 
rather than risk aware when it comes to copyright or other rights 
management methods, though.

Anyway, I think OSM needs to be a broad church and I really don't expect 
everyone to agree with everyone else's techniques, whether its with 
regard to collection of feature names, or collection of geometry, or 
tagging, or anything else.  Frankly its a good thing people don't all 
work on OSM how I do, or Canberra OSM data would be pretty, but crap, 
instead of great. 

cheers
Andrew.