[OSM-legal-talk] Database and its contents

[Ed Avis]

Emilie Laffray <emilie.laffray at ...> writes:

>As always, the standard reality check applies: if you believe that maps or the
>data they represent are not covered by copyright, please start large-scale
>photocopying of some commercial maps, or copying the information from them into
>another format that you then publish.
> 
>See that's where you are confusing maps representation and the underlying
>data. If I was to start a large scale photocopying of some commercial maps, I
>would infringe copyrights of the printed map. This doesn't infringe on any
>possible underlying rights since you are stopping at the map itself. This
>doesn't imply anything on the underlying data.

See the second part - 'or copying the information from them'.

Copying the underlying data is one of the things that we do in OSM.  We trace
over other maps, copying only the underlying facts (this street is there, and
it has this name) and not the representation.

When can we not do this?  When the map is under copyright!  Have a look at
<http://wiki.openstreetmap.org/wiki/Out-Of-Copyright> which has a long list
of old maps which are or are not okay for copying into OSM.  Most of them date
from long before any database right was introduced, so if the data itself is
not covered by copyright, we should be able to trace from them with impunity!

Again, reality doesn't agree with the assertion that copyright doesn't apply.
Even though you are dealing only with 'data' or 'facts' and not merely
photocopying the representation of them, you are still infringing copyright,
and there have been recent court cases backing that up.
(In the UK, for example, Ordnance Survey versus AA - copying of factual
information from OS maps triggered a copyright action, which succeeded.)

I agree with your point that database right exists and we need to license it.
I wanted to address the particular point, often asserted here, that map
information is somehow exempt from copyright so we cannot use copyright to
protect it.  That's just not backed up by the real world.

>Since we are here, can I ask you two questions on pure licensing? The answer
>needs to be short else, it will be drowned in words to hide the true belief you
>have.1) In which camp are you?2) Do you believe that CC-BY-SA would protect
>the project legally?

Yes, I believe it is a simple licence based on well-understood law which
provides an effective way to deal with violations.

More than that, it has proved effective in practice; big organizations such as
Microsoft or MapQuest use our data respecting its terms; and no map vendor
would ever take the risk of contaminating their product with licence-infringing
data.

For the first question, I don't really mind as long as the project uses a free
and open and licence which is workable in practice.  I am reluctant to accept
the ODbL as truly free and open because of the contract law provisions; to me
it resembles more an EULA which attempts to limit through contract restrictions
the rights which you would otherwise have.  I believe a licence should be a
pure grant of additional permissions.  (These are also easier to enforce, since
the other party cannot argue that the licence 'does not apply' - if it doesn't
apply, then they do not have any rights under it.)

I also believe that the ODbL/DbCL combination is unclear and cumbersome in
practice, attempting to shoehorn a map into a distinction between 'database'
and 'contents' which makes little sense, among other issues which tend to
cause lots of diverging interpretations on this list.

I used to support share-alike terms but they seem to cause trouble as people
think of loopholes (real or imagined) and add more and more legalese.  From
a practical point of view I think public domain or attribution-only would be
superior for this reason.  The status quo of CC-BY-SA is working well but it
is hard to stick to the rule "if it ain't broke, don't fix it".

-- 
Ed Avis <eda at waniasset.com>