[Openstreetmap] Re: money and what to do about it

[Matt Amos]

On Friday 18 March 2005 04:31, Schuyler Erle wrote:
> * On 17-Mar-2029 at  5:29PM PST, SteveC said:
> > IP wise, whats to stop me reconstructing the image from mapserver
> > tiles and selling it in competition with $satellite_reseller? Or,
> > giving it away free hence lowering the market value of their
> > product? Unless they have extremely liberal license here, or they
> > really want to view the person scraping as the infringer... there
> > could be problems?
>
> I believe that Space Imaging's license for the IKONOS data
> prohibits redistribution of the raster data unless downsampled to
> 16:1 or less, but explicity permits unlimited redistribution of
> derived vectors.

umm... would putting it on a web server for the general public to 
annotate count as "redistribution"?

> Anyone volunteering to do vectorization would probably have to sign
> (click?) some kind of legal agreement to the effect that they
> understand the terms of the license under which we acquire the data
> and agree, for the sake of the project, (a) not to copy or
> redistribute the satellite imagery, no matter how tempting it might
> seem, and (b) that the vectors they create based on the imagery
> become the property of the sponsoring organization (e.g. OKFN), to
> be published under a suitable Open license that everyone agrees to
> ahead of time.

there are issues here w.r.t. caching (e.g google images, http proxy 
caches, internal browser caches, etc...) which are violations of 
"thou shalt not copy". but these are technical problems and can 
probably be solved (e.g: cache-control: no-store).

> > I'm flipping between thinking its really easy or really hard to
> > do properly. It depends a lot on the image scale and what kind of
> > topography its looking at. 'Course maybe the image comes with
> > coordinate overlays or other useful data.
>
> It will be easy. Any imagery we purchase will have to be
> georeferenced and orthorectified.

does this include compensation for a flat image of a curved earth, or 
are the pictures small enough that this doesnt matter? i know this 
was a problem with the landsat data, as the patches never quite 
joined up at the edges.

> > > Where can we get a chunk of 1m / comparable aerial photography
> > > to experiment with?
> >
> > As a proof, just steal it from somewhere.
>
> Actually, Space Imaging offers samples. Someone just has to
> overcome the tedium of drawing vectors over some random part of the
> world no one cares about and then see what the vectors look like.
> If it works (and professional GIS houses do it all the time, so I
> expect it will) then we can splash out on the imagery, in the
> knowledge that it will be useful, and then experiment with
> automated extraction or what.

apparently this is how OS does it anyway, so it can't be all that bad, 
right?

cya,

matt
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 190 bytes
Desc: not available
URL: <http://lists.openstreetmap.org/pipermail/talk/attachments/20050319/359164df/attachment.pgp>