Talk:Open Data License/Geocoding - Guideline

From OpenStreetMap Wiki
Jump to: navigation, search

Geocoding and reverse geocoding

The legal implications for geocoding and reverse geocoding may be different, but they are handled as if it's all the same thing: "Geocoding Results can be latitude/longitude pairs, full or partial addresses and or point of interest names." - should this be clarified or separated?

Very different use cases

For me, there are very different use cases, even if we consider only geocoding, and not reverse geocoding.

Live geocoding

For live geocoding, that's getting a single geocoded location by using a big part of the OSM DB. It's clear that this single record is a produced work (after all, a substantial part of the database is used to geocode a single address). As a produced work, the result should still mention a credit to OSM, but that's the only requirement.

Geocoding a private database

When a company has a database of clients, they don't want to share that database. But that's fine, since the share-alike requirement only kicks in when you do share the database. If you're not planning on sharing the database, then there's no problem using OSM data, and apart from keeping their stuff private, they have no obligations towards OSM.

Geocoding a database for publication

When you start with a database of records, geocode these records using OSM, and publish that data, this becomes the most interesting case.

From my reading, the individual records can still be protected by certain rights (f.e. in a database of photos, every photo can be protected by copyright, names can be protected as trademarks, free text like advertisements can be protected under copyright, ...) however, the database is a derived database, so has to be released under ODBL or compatible. When we look at the different use cases, it seems only normal that we want to free that data.

  • A commercial group that wants to geocode a list of all members of the group, and publish it:

In this case, I think it's a clear win-win when the data is distributed as wide as possible. The group can only get extra publicity from letting the data grow in OSM.

  • A tourism-oriented website that makes a selection of the best restaurants in a region, geocodes it, and publishes those:

This website typically only wants to distribute their selection through their own website, as that gives them add revenue, or perhaps users need to be a paying member. IMO, this is a typical example of a freeloader: they use our adresses (which costed us a lot of effort to gather and maintain), but refuse to give anything back. So we should be allowed to import the combined data back in OSM (in case they were missing). Note that they can still protect the photos they gathered, or the texts they wrote about the restaurants. If there's no possibility to spread the data (f.e. because they don't own the rights), then they should move on and use a different base to geocode on, either automated or manually (by visiting the restaurants). They're no value for the OSM community either, so why would we care?

The above post is my view on it, I think it complies fully with the ODBL text, and I think it's best for OSM. --Sanderd17 (talk) 13:44, 20 February 2015 (UTC)