[Openstreetmap] Re: OSM License and debian

[Chris Holmes]

Hi everyone.

First I wanted to introduce myself, I've been lurking for awhile.  I
work on GeoServer, an open source java based implementation of the Web
Feature Service (WFS) and Web Map Service (WMS) specs, among other
things.  I've been thinking about the ideas behind open street map for
awhile, and I'm quite excited to see a community already forming around
them.  For me the real interest is attempting to apply open source
principles to something other than computer software.  The idea that it
may be a new way of organizing production, and the only reason that
software is first is because computer programmers are the closest to
the medium that makes it possible – the Internet.  A quote:

'The simple logic of open sourcing would be a choice to pursue ad hoc
distributed development of solutions for a problem that exists within
an organization, is likely to exist elsewhere as well, and is not the
key source of competitive advantage or differentiation for the
organization.'
From
http://www.ssrc.org/programs/itic/publications/ITST_materials/webernote1.pdf

I very much think spatial data falls squarely under that definition, and
am interested in exploring what might be possible with a variety of
open source approaches, which my guess is will fall somewhere between
software, with relatively high barriers to entry (knowing programming,
getting commit rights) and wikipedia (anyone can edit anything).  But I
like starting with the most anarchic :) (and it is easier to implement
than building cvs for geodata from scratch)

> I welcome commercial use of the OsM data, because I want to increase
> the user of geo-referenced information.  I hope that increased
> awareness of the advantages of geo-referenced will convince decision
> makers to make the databases of the national mapping agencies
> available at no or low cost because this will improve the service
> level and productivity of the society as a whole.

My motivations are actually a bit different.  I am looking to a lot more
than just convincing national mapping agencies to release data.  What
I'd like is a geospatial commons that is maintained by more than just
governments, all of us just waiting for the day that we can convince
decision makers that there is value in freeing data.  I believe that we
all can actually help out governments, and show that by opening their
data, and letting others help to maintain it, they might actually save
money.  And indeed that there are far more data producers and potential
data producers than just governments.  For gov'ts the collaborative
maintenance certainly won't be the anarchic wiki commons type
architecture, but I think a variety of open source type governance
structures, tools, and licenses could lead to workflows that make it so
private enterprises, citizens, and government agencies alike can all
contribute to a commons of geospatial data.  In the states we have lots
of free data, but it is out of data and of inferior quality.  I'd like
to see an open source process that allows it to rival (and surpass)
commercial data providers, and save replicated energy and money for all
of us.  And indeed in countries that don't allow any, then start from
scratch.  But I would see success as more than just Ordnance Survey
making their data free.  I would see it as them adopting some of the
principles of feedback and user contribution to geodata that make OSM
possible.

So that's where I'm coming from.  I meant for that to be a shorter
introduction, but brevity is far from my forte.  I also have feelings
about basing this all on open standards, leveraging other gis
programming toolkits from the beginning, and making the architecture
compatible with traditional gis, but more on that later.


> Realistically, we're going to have gigabytes, if not terabytes of data
> before we're satisfied. Therefore I think that the idea of
> "distributing" OSM data with any OS or OS distro is probably not worth
> considering.
..

> Consequently I don't see that a CC license for data is incompatible
> with a GPL license for the code, or that a CC data license would
> seriously interfere with people's access to the data in a way that
> would make the license (or the licensed data) unusable for our
> purposes. People understand CC as applied to content and (to me) it
> solves more problems than it creates.

I very much agree with this, and indeed find the idea of applying GPL
and BSD and other software licenses to data very odd.  If you read the
GPL it constantly talks about software, and I just don't see it working
well with data.  And I don't understand why it would be included in a
linux distribution – I think most people would browse the data on the
web, and probably use web apps with it, and then _maybe_ download the
bits that are relevant for the area they care about.  Yes, some people
will want the full database, but not someone installing an operating
system.

I believe that eventually we should seek specific geo-data licenses. 
And I think Creative Commons would be the starting point, and provides
a great license to use for now.  I do think that there are enough geo
specific points to warrant a set of geo-cc licenses (things like
printed maps vs. online, how the attribution is done, if analysis of a
dataset needs an attribution).  But more than that I think it's
important because GIS types like to believe that they're 'special', and
that their problems are unique.  I personally don't think they really
are, but it will make many an individual, company, and national mapping
agencies happy if there is a license written particularly for them. 
And I think when it comes to that we should definitely get in touch
with the CC people.  Until then I think a general CC license is
definitely the way to go.  The debate should be on share alike or no –
making it viral like GPL or a free for all like public domain/apache.

Best regards, and keep up the good work,

Chris


----------------------------------------------------------
This mail sent through IMP: https://webmail.limegroup.com/