[Talk-GB] OSM Contributor Terms vs OS OpenData Licence
mapping at sheerman-chase.org.uk
Mon Apr 18 22:42:13 BST 2011
On 18/04/11 22:23, Frederik Ramm wrote:
> I'm an outsider to all this OS business but if you guys in the UK
> should really have been uploading data that requires attributing OS in
> every downstream product then we have a problem which has nothing
> whatsover to do with the license change. I can see *no* OS attribution
> on any of the major tile providers, including our own. Of course you
> can always go to the source and see from the object history that OS
> was involved, but that is a technique that you seem to discount above.
> So either this is all a big misunderstanding, or nobody who used OS
> data until now has cared sh*t for the license.
> Now I could understand if someone has always maintained that OS data
> was incompatible with OSM and thus refused to use it.
> What I cannot understand is if someone has happily used OS data until
> now, in the full knowledge that nobody would attribute OS downstream
> anywhere, but now says they cannot sign the CT because they codify
> exactly what has been happening. Reality check, anyone?
I actually agree with you Frederik, but the entire project so far
overlooks the even bigger problem that CC-by-SA technically demands that
every contributor is attributed in every derived work. Given that we
ignore that issue, and that OS Opendata is compatible with CC-by and
that is always compatible with CC-by-SA, we have just imported OS data
into OSM because the larger attribution issue never was solved. So we
are all equally guilty.
Now along comes ODbL, which was intended to address shortcomings with
CC-by-SA. To say we were in technical violation of CC-by-SA doesn't
justify us going along with ODbL if ODbL is flawed. If anything, we
should strive to be more legally rigorous, not less. We do have an
imperfect attribution on the wiki  for CC attribution. Agreeing to
the CTs seems to be a bigger violation than our current practice,
because it declares that the contributor has unlimited rights over the
data (in order to grant OSMF that right too). Also, just because OS has
not complained so far is not a reason to continue to abuse their license.
The solution, as far as I can see, is to improve attribution
requirements (which would mean rewriting the CTs again) or to remove the
data. I know Richard thinks Opendata doesn't pose a problem, so there
may be other answers... I call for LWG to get their analysis of the
Opendata legal situation done ASAP - that might put minds at rest or
allow us to get on with fixing the problem.
More information about the Talk-GB