[OSM-newbies] Footpaths again

[Richard Welty]

On 3/19/10 11:39 AM, Andre Engels wrote:
> On Fri, Mar 19, 2010 at 4:27 PM, James Ewen<ve6srv at gmail.com>  wrote:
>    
>
>> We are told that we can not even look at things like Google Maps to
>> check the spelling of a road name due to copyright infringement
>> issues.
>>      
> Which is silly. The name is not copyrighted by Google, so you don't
> break their copyright by getting it from them.
>
>    
the big problem is that maps often contain errors, sometimes 
intentionally introduced.
the intentional errors are there to create evidence of copyright 
infringement.

i have some excellent maps i get from a local source, they're 
invaluable. but they
do contain errors, i've seen enough errors to bring home the point that 
copying from
maps is not kosher for OSM, both because we want it to be right, and we 
don't
want to be sued for copyright infringement.

using them to support research on the ground is perfectly acceptable, 
but it's important
to do some level of direct, personal verification. i have reviewed every 
single rural
primary, secondary and tertiary road in Rensselaer County, NY. it took a 
while, but they're
all right. if i trusted other sources, they wouldn't be. there are 
mistakes in the other
sources.
>> I asked specifically about using a Google Streetview image to read the
>> street name off of the signpost, and the consensus was that such an
>> action would be in violation of the OSM directive of not creating
>> derivative works from a source protected by copyright.
>>      
> That's a ridiculously extreme definition of 'derivative work'. Why
> would you be allowed to use the signpost (with a name that may be
> copyrighted, and a design that may be copyrighted) but not a picture
> of it? Is Google's copyright so holy that we may not break its
> copyright even in the case where it doesn't have copyright, but we are
> free to break other people's copyright?
>
>    
i concur. the fact of the road name itself is not subject to copyright. 
the image is
subject to copyright. however, because we don't know certain other
things, using a street view to get this info is problematic.

for example, a couple of months ago i mapped a small extension of an 
existing
subdivision that wasn't in OSM:

http://www.openstreetmap.org/?lat=42.66856&lon=-73.62842&zoom=17&layers=B000FTF

one of the new roads was named Teresa Terrace. i was back over that way 
a few
weeks ago, and the name had been changed to Tiffany Place. Google maps still
has the previous name. the resolution in the street view doesn't let me 
see what
the sign is, but there's a fair chance it's the old name.

if you could read the sign, and depended on Google Street View to be 
right, you'd
be introducing an error into OSM.
>> It is stated that one should not create GPS traces with
>> "snap-to-roads" enabled, as the resultant data would be derived from
>> the database contained in the GPS device. Obviously the derivatove
>> trace would not contain ALL of the data from the database, and as such
>> by your definition, would be acceptable for inclusion into OSM.
>>      
> No, I'm not saying that any part will do. The issue is what
> information is copyrighted and what is not. To be copyrighted it must
> have a certain minimal level of creativity. The location of a road may
> well be copyrighted, as you have to make decisions as to whether or
> not include certain curves and such. The name of a road is just a
> simple fact, there is no creativity in calling Main Street Main
> Street.
>
>    
the issues with snap-to-road are the same as with copying from maps: we 
want OSM
to be right, and we don't know why the road coordinates in the map on 
the GPS are
wrong -- simple error, or setup for copyright claims.

richard