[Openstreetmap] The bigger picture

[Tom Carden]

On 2/14/06, Lars Aronsson <lars at aronsson.se> wrote:
> Tom Carden wrote:
>
> > Could the £10 you suggest be split between OpenStreetMap and a
> > commercial satellite/aerial photography company perhaps, with
> > the latter providing pay-for access to imagery that would allow
> > you to sanity-check the OSM data?
> > [...]
> > Note that I'm not talking about restricting access to the API,
> > although that again is something that could be a pay-for service
> > for commercial use.  Basically, so long as people are always
> > entitled to access OSM's raw data for their own personal use,
>
> This whole discussion is a provocation!

Not really.

> Unless the list owner
> makes it clear that this line of thinking is impossible,
> I'm on my way out of here.

That would be silly.  Nobody wants to take anything away from anybody.
 We're not "doing a CDDB", for sure.

> Some people obviously want to build a new
> Ordnance Survey, and if there is any risk that OSM is going to be
> *their* platform, then I want to make sure that I didn't
> contribute to their effort.
>

I think that being a "new Ordnace Survey" is a pointless, thankless,
uphill task and I'm not interested in re-inventing that wheel in the
slightest.  There is a bunch of stuff that will always make an
OpenStreetMap collaborative/bottom-up approach fundamentally different
to the traditional approach and that's why I'm here.  I want
peer-produced Free-as-in-Freedom maps, maps I can download, edit,
sell, give to my friends, and post on my website.

> When I spent this summer, fall, and winter drawing maps for OSM it
> was my intention that these maps should be available for all, for
> free, for ever, for any purpose (commercial or not), under the
> CC-SA license.

Absolutely, and we can't change that now.  All work submitted to
OpenStreetMap to-date, if it's redistributed at all, has to be
redistributed under that license.  But here comes my point *if it's
redistributed at all*.  As far as I know none of us is obliged to make
it available forever, on demand.  The things I'm talking about are
ideas to make OpenStreetMap sustainable.

I'm suggesting premium accounts which allow people more hits, more
bandwidth, extra features - above and beyond what the project is
obliged to provide - in order to cover costs of hosting and
maintenance.

> If OSM is going to become the kind of proprietary
> map publisher that Tom suggests, it can under no circumstances
> claim any rights to the maps I made.

Playing devil's advocate here, I'm not sure about that.  The current
license linked from the front page states:

"You are free:
to copy, distribute, display, and perform the work
to make derivative works
to make commercial use of the work"

"The work" I take to mean the maps as distributed through the slippy
tiles interface.  I hardly think the licensing situation is as
water-tight as is made out, and I don't think OSM does much to
maintain the licensing for users of its API - there is no privacy
policy, terms and conditions, user agreement.  It's a free for all.

There are lots of things that need patching up if it's going to work
like you suggest.

That said, your interpretation is absolutely in the spirit of OSM as I
understand it, and how it's currently run.

> You are free to charge for
> the map data, the same way that Red Hat and SuSE charge for Linux,
> but you are not free to restrict people to use and share the data
> for free.

So what is the difference here?  I'm not talking about charging for
the map data, I'm talking about charging for services which use the
map data, or services which provide access to the map data in a
specific format.  And I was only talking about ways in which we might
satisfy people who are accustomed for paying for such things.  In the
same email I also suggested a bunch of things which I think should be
free as in beer.

> You cannot even dual license the data (the MySQL way),
> because you don't own the copyright in the first place.
>

I think this needs clarification.  In my opinion OSM *should* own the
data, it makes distribution and attribution in particular a lot
clearer.

I'm overlapping with Steve and Imi's responses now, so I'll pick up
their points further down.

T.