[OSM-legal-talk] New site about the license change

Ed Avis eda at waniasset.com
Tue Nov 16 21:45:31 GMT 2010


Richard Fairhurst <richard at ...> writes:

>My strongly-held belief is that, just as it's generally accepted that to
>discriminate against "fields of endeavour" with a non-commercial licence is
>not "open" (e.g. see
>http://blog.okfn.org/2010/06/24/
>  why-share-alike-licenses-are-open-but-non-commercial-ones-arent/),
>the application of a creative works licence to a data project creates
>similar discrimination against fields of endeavour: traditional 'creative
>works' built on the data are required to be shared-alike, but other works
>(including even data) aren't.

I think it's pretty clear that data, if derived from the OSM data, would need
to be distributed under the same share-alike terms.  The only way this would
not be the case is if the law (copyright law in particular) grants no
exclusive right over map data - in which case we are back in the hypothetical
world where everyone can trace over Tele Atlas and OS Master Map, they can copy
from us, and there's nothing anybody can do about it.

>My view is that an artistic cartographic map
>using OSM data is as independent a creation as a routing application using
>OSM data, and the same rules should apply to the cartographic art as to the
>source code of the routing app.

This is an argument in favour of the ODbL's provisions that you can make a
'produced work' and it doesn't have to be distributed under the same terms.
That is in my view the one big point in favour of the ODbL.  I rather liked the
idea that printed maps produced from OSM would have to be freely licensed, but
this was an emotional reaction.  It would probably help the project (as well
as certain companies) if map tiles could be redistributed without restrictions.

I am unhappy, though, at the claim that although these map tiles or printed maps
need not be distributed under ODbL, it would be prohibited to trace over them
and make your own map data from them - because this would 'regenerate the
original database' or something like that.  So the tiles are not really
unrestricted after all; they come with a no-reverse-engineering clause.  This is
one of those (in my opinion) unenforceable claims which restrict ordinary people
without a lawyer on hand, but have no real legal force.

>>If the current licence can be trivially circumvented, people
>>would be doing so by now and we'd see Google or Tele Atlas copying 
>>the OSM data with impunity; yet there are no such examples.
>
>Oh, there are plenty of infringements: yet another one whizzed by on #osm
>today and no doubt there'll be another later this week. Those who have to
> care about their PR (Google, Waze) will abide by the spirit of the licence,
>albeit in retrospect; the cost of the negative PR outweighs the minimal
>saving in geodata licensing. For some the equation balances the other way,
>so they won't attribute or share-alike.

Agreed.  But do you really believe any change in licence text will affect
that?  Those who ignore CC-BY-SA today would ignore any licence.
 
>But I'm not really talking about infringements per se; I'm talking about
>circumventing the spirit of CC-BY-SA within the letter of CC-BY-SA. The
>"computer-generated derivative" previously discussed here and on
>cc-community is the obvious example; you can avoid having to share if you
>combine on the client rather than the server.

That's more interesting.  Yes, you can run a program on your local computer to
download data (or any copyrighted work, really) and make manipulations to it.
You don't have to share the result with anyone.  In my view trying to forbid
this is a step too far and would make the data non-free; everyone should have
the right to make local changes, and the right to make computer programs to
do it.  But this is truly a matter of opinion.

>>As for PD, I'm not sure that the 'community consensus' on that has ever 
>>really been measured.
>
>Not formally, no. Certainly I based my decision to actively support a move
>to ODbL rather than a move to PD (or attribution-only) on the grounds that
>every time anyone even tentatively suggested the latter, the mailing list
>storm was so vast that I could never see it being remotely realistic. Of
>course, I hadn't realised that a storm would be provoked when anyone tried
>to suggest _anything_. :(

Yup.  The most curious feature of the whole process is the assertion that there
is a consensus for a share-alike licence, but somehow not for the share-alike
licence we are successfully using; and that a public domain (or CC-BY) map
is a practical impossibility, but that the much more complex ODbL+DbCL+CTs
setup is workable.

-- 
Ed Avis <eda at waniasset.com>




More information about the legal-talk mailing list