[OSM-legal-talk] Why is the data protected?

[Anthony]

On Sun, Nov 28, 2010 at 2:51 PM, Andreas Perstinger
<andreas.perstinger at gmx.net> wrote:
> Isn't the content the users provide just facts (at least the
> coordinates, some tags could be questionable)?

I don't think it's quite that simple.  If I draw a complex
intersection on a piece of paper, that's not a fact, but an expression
of a fact.  Likewise, if one creates a complex intersection in OSM,
I'd argue that it's no different from creating it on a piece of paper.
 It's an expression of (a large number of) facts.

Is the expression obvious?  In most cases, probably.  Just trace down
the center of the road, right?  To the extent that's all people do,
even the expression, while not purely factual, is uncopyrightable (*).
 But then, add in a merge between two roads.  Ask 10 different people
to map that situation, and I bet you get more than 2 different
answers.  When do you start the merge?  When do you complete the
merge?  Try it different ways some time, and see how it renders.  Or
just look at the OSM data at onramps/offramps and notice how different
people chose different ways to express the facts.

I think the answer is that the substance of OSM is almost completely
factual.  But there's that other tiny fraction of content which I'd
say is copyrightable (*).  So I think it would be wrong, legally as
well as morally, for OSMF to simply ignore the underlying license.

(*) Probably, in most jurisdictions.

On Mon, Nov 29, 2010 at 4:04 AM, Andreas Perstinger
<andreas.perstinger at gmx.net> wrote:
> I would argue now that before the new CT (since March 2010 if I'm right)
> there wasn't a real CT at all (for me the simple sentence used is just too
> vague and let many questions unanswered). Therefore I think even if you have
> any problems with the new license, as a mapper, you should welcome the new
> CT if you want that OSM in future is build upon a stable fundament.

I would welcome *a* new CT, along with the old CC-BY-SA.  Both the 1.0
and 1.2 CTs have problems, though.  A proper CT would be something
like:

1) Everything you submit is either completely owned by you or on a
list of exceptions deemed by OSMF to be compatible with OSM (**).

2) You grant OSMF with a license to your contents under CC-BY-SA (2 or
later), and you further grant them permission to sublicense your
contents under CC-BY-SA (2 or later).

3) OSMF grants you (and everyone else in the world) a license for the
entire contents of OSM, under CC-BY-SA (2.0).

4) OSMF promises to always license everyone's submissions under
CC-BY-SA (2 or later).

The purpose of this CT is to help alleviate the questions about chain
of title which otherwise would plague (and except for the CT,
currently do plague) the ability of commercial users to use OSM.  CT
1.0 goes too far, to the point where it is being blatantly ignored.
CT 1.2 doesn't go far enough, as a vague "I think my stuff is
compatible and you can remove it if it isn't" provides little
assurance at all.  Maybe something in-between, in the form of a list
of exceptions, editable only with OSMF approval, would be doable.

(**) Exceptions would mainly be imports from people/organizations
which themselves agreed to the CT, or compatible data where the chain
of title was extremely clear (e.g. data collected by a government
entity, released as PD or CC-BY).  In extreme cases, exceptions might
be made for imports from sources which didn't agree to the CT, because
to not allow them would greatly hinder OSM (e.g. if a major player
forked OSM and refused to sign the CT), but this would have to be
decided on a case by case basis.