[OSM-legal-talk] ODbL comments from Creative Commons

[Simon Ward]

On Sat, Mar 21, 2009 at 07:02:09PM +0100, Ulf Möller wrote:
> Thinh Nguyen of Creative Commons has posted detailed comments on the 
> ODbL on the co-ment website.

I think I’ve seen many of those arguments from the Science Commons
project before, and they still gloss over the question of “what if the
data provider wants share alike?”.

They concentrate on whether the licence meets the requirements of the
Protocol for Implementing Open Access Data assuming it’s right for
everyone.  No doubt they believe it is, but that doesn’t mean it should
be so.  (Note, the background of the comment does cover this, and states
in advance that rest of the comment is about the applicability of the
Protocol and the ODbL.)

They fail to note that if there are no rights in some jurisdiction, then
the database is free anyway, and so are derivatives.  If there are
rights in some jurisdiction, and the database is made PD, then others in
that jurisdiction may use those rights to avoid sharing derivative
databases.  Public domain dedication for *all* databases would be ideal,
but you lose a good portion of the benefits when people are allowed to
exercise restrictive rights over derivatives.

Much of it is about complexity and cost, but that hasn’t been too much
of a barrier to the adoption of reciprocal free software licences.
Granted, it is more of an issue for databases and data because rights
are not given in some jurisdictions, or are less well‐defined.  The way
to avoid complexity and cost would be to assume you have to follow the
licence, if in doubt assume you have to reciprocate, and not to try and
fight the licence and go through the expense of working out if you can
get away with it not applying to you.

In the university researcher example, if the researcher has the choice,
he should make his work free.  If it is down to university policy, then
the university should allow the work to be free.  If it doesn’t, the
researcher can explain that he would have access to the data if the
university relaxed its policy (it’s a shame that many universities seek
to keep their research somewhat closed, especially when there is
potentially money involved).  However, I don’t believe the share‐alike
necessarily comes in to action:  The researcher should be allowed to use
data for research, even substantial extractions from an ODbL licenced
database.  Only if the researcher provides a derived database do the
terms start to get in his way.

I’m unsure about having contractual obligations in the licence, and
agree with some of what the CC say, but again, the way to avoid issues
would be to meet the requirements of the licence.  The other issues,
such as enforcing the contract on downstream recipients, and the
question of whether a “click‐wrap” agreement is necessary are being
debated, and people are trying to come up with the answers, rather than
just run away from the issues.

It is a fact that share‐alike causes interoperability issues.
Share‐alike in free software doesn’t get away from those, which is a
reason why the list of licences at gnu.org[1] exists.  People are
willing to put up with these interoperability issues, giving up some
freedom and sometimes used by all permissive licence supporters to brand
share‐alike licenses as “non‐free”, for different freedoms: those that
allow recipients of derivatives the same rights as they would get with
the original.  If some other database licence does not allow it to be
distributed under the terms of the ODbL, then there is either a bug in
the ODbL that can be rectified, or the terms of the other licence are
too restrictive (or rather, don’t have the same freedoms).

[1]: http://www.gnu.org/philosophy/license-list.html

CC itself has already shown how we can deal with complex licences, and
help users to understand them.  They provide quite easily understandable
summaries of their licences that indicate how they can be used.  They
concentrate on the core requirement, and don’t cover edge cases.
Something like that, as well as assuming you should share the database
if in doubt, can alleviate much of the complexity and avoid people being
bitten through not keeping to requirements when they should have.

Maybe I should add this to the discussion page for this comment… I
probably will do if someone else doesn’t beat me to it.

Simon
-- 
A complex system that works is invariably found to have evolved from a
simple system that works.—John Gall
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 197 bytes
Desc: Digital signature
URL: <http://lists.openstreetmap.org/pipermail/legal-talk/attachments/20090321/ad7f3a10/attachment.pgp>