[OSM-talk] reusing OSM POI information

[Frank O'Dwyer]

Hi,

I've developed an iPhone app / web app for annotating/rating/tagging
locations.

I haven't formalised this yet but the intention is for the data the app
is collecting to be open data - I'm currently intending to use the
opendatacommons database license for the data collection and cc-by-sa
for the entries themselves. I'm not sure exactly which version of
cc-by-sa to use but since 3.0 is the latest I'm thinking of that.

I'm interested in reusing the OSM POI information and incorporating it
into the database. Conversely I'd like to be able to share back with
OSM the data that the app collects, as there is probably a lot of
overlap and perhaps new information for you (e.g. additional GPS fixes
per POI, etc). Other information like rating and tags and unstructured
descriptions you may or may not be interested in but of course that
would be available also.

I know that OSM has had a lot of discussion about switching licenses so
my question is if there is any problem with doing this and if my
proposed license choice would be compatible with yours? Also, if I do
this who should I attribute - the individual OSM contributors (how?) or
OSM or both? I have read the wiki FAQs on this but I thought it would be
better to check.

On the technical side I have got fields in my database per entry for
contributing user, data source (string), and data source identifier
(also a string). The first field references the record for a normal user
of the app and the other fields are intended to reference the original
source of an external POI (normally null as I don't use any external
sources yet). I was thinking of something like creating a user called
'osm' and then using the source and source id fields to reference the
osm username and node id. In this case the entries would be attributed
to OSM. Alternatively I could create dummy user records for external
contributors, which would attribute the entries to individuals. Having
the node ID should also make it easier to match up entries across the
two DBs. Also if necessary, adding to the schema is not too much of a
problem.

Cheers,
Frank