2009/2/24 sly (sylvain letuffe) <[email protected]>: > >> We could also create a preprocessor that "unifies" OSM data. > > I've been thinkink a lot about an osm pré-processor, but not to solve the > problem you mentionned. > > The main problem, to my point of vue, in the duplicates tag case, is not at > developper/renderer side, but at begginers learning to contribute to OSM, and > that is not solvable by a pre-processor ;-) > > I mentionned the fact that solving it for begginers will also solve it for a > cheap price at dev/renderer side. > >> This seems to me far superior than trying to meddle with the OSM >> database itself through bots & scripts. > > For data uses yes, but not for data feeding. > > The idea I was after with a pre-processor was to explicit those so many > default values than need to be hardcoded in every routing software in order > to guess that a motorway is unaccessible to pedestrians. > But that's another story. >
Have you played with the TagTransform plugin I wrote for Osmosis a while back? http://wiki.openstreetmap.org/wiki/Osmosis/TagTransform It only works against osmosis-0.29 unfortunately... I'll have to fix that. I wrote it to do some more complex duplicate removal for the cycle map (as a preprocessor before osm2pgsql). The key one being highway=path,bicycle=yes which as far as I can tell is a duplicate of highway=cycleway. It's not actually rolled out yet though I keep meaning to do the necessary tweaking of the tool-chain. Anyway, point is it's a fairly abstract framework for removing/adding/changing tags on OSM objects that may be of interest. Dave _______________________________________________ talk mailing list [email protected] http://lists.openstreetmap.org/listinfo/talk
