I just realized that in OSM there are sometimes streets with coinciding nodes. I use the OSM data for matching gps position data on the OSM road network and these street segments with zero length disturb the matching process. So currently I just filter them out by myself. This is fine for me and does not really bother me.
I would like to suggest to remove such nodes automatically on the serverside. I think this isn’t a big deal, but it would improve the map quality for further processing. Where can I propose such a change?
Well this issue has been discussed already years before, do some internet search about “OSM duplicate nodes” or “way with A-B-B-C nodes” … some quality assurance tools can reveal them, see OSM wiki about it.
So a road with duplicate nodes is indeed false and should be corrected? I found some reasons for duplicate nodes, e.g. for multilevel roads (mentioned here: http://wiki.openstreetmap.org/wiki/Duplicate_nodes_map). But then the duplicate nodes belong to different roads.
As I have no experience with automated edits in OSM, I will first edit the OSM data locally. I will then propose the automated edit in the mailing list of the corresponding country.