[odc-discuss] Copyright status of OSM map data

[Richard Fairhurst]

Ed Avis wrote:
> The broad conclusion is that OSM data does fall within copyright, even in
> the US.

I've been trying to stay out of this issue on the OSM lists, but I'm
afraid I think the report is significantly flawed.

When the memo asserts "Labeling and annotating could be described as
highly discretionary tasks" this may be true for the (generally frowned
upon) non-verifiable minority part of OSM mapping, such as the
‘smoothness’ tag (http://wiki.openstreetmap.org/wiki/Verifiability).

But it appears to me that Ms Gellis has been ill advised here. Street
names, in most localities, are not discretionary. Nor are road numbers and
turn restrictions. The road outside my house is mapped with a geometry
following its centreline to the limitations of the source material
available, tagged with a name clearly marked on a street sign, and with a
road class ("highway=") chosen by following clear precedent and criteria.

“Labeling” (tagging) is an unoriginal activity based on observation of
“what’s on the ground”, that cardinal OSM rule. These simple street
attributes and basic connectivity are the most commercially valuable part
of OSM; therefore the most likely to be infringed; therefore the most
necessary to protect.

Ms Gellis also cites City of New York v. Geodata Plus, LLC. Beth Noveck, a
law professor and United States deputy chief technology officer for open
government (http://en.wikipedia.org/wiki/Beth_Simone_Noveck), has publicly
cast doubt on this:

"It is questionable from the facts set forth in the judge's order whether
the act of digitally ‘tracing’ a map to convert a paper-based to a digital
image rises to the requisite level of originality to satisfy the Supreme
Court's demand for originality under Feist. While the underlying facts in
the paper-based maps are non-copyrightable (though the Second Circuit has
held that a municipality may retain copyright in its tax maps to the
extent necessary to create an economic incentive to create them and so
long as the public's notice of the law is not impaired), the Eastern
District emphasizes the laborious and ‘painstaking’ process that goes into
doing the tracing and finds that the vectorized maps are certainly
copyrightable. This is despite the fact that the Supremes rejected the
‘sweat of the brow’ doctrine as a basis for copyright. Originality, not
hard work, is the sine qua non of copyright and tracing seems little more
obvious than the alphabetizing which defeated copyrightable for Feist's
telephone directories."

(from http://cairns.typepad.com/blog/2007/10/feeding-bureauc.html)

It is not for me to say whether an eminent law professor or a district
judge is correct. Ms Gellis’ citation of the case is interesting but it is
clearly not unanimously believed to be definitive.

The wider misperception is that the change to ODbL is principally driven
by a belief that no protection would be available in the US without a
contract-based "licence". That is not the case.

Rather, there is a recognition that protection (and means of protection)
vary very widely across jurisdictions and are frequently changing. As an
example, just a couple of weeks ago, a European Court of Justice judgement
appears to have inter alia weakened the sweat-of-the-brow copyright
protection available in the UK.

Richard