[OSM-dev] duplicate segments

Artem Pavlenko artem at pavlenko.uklinux.net
Sun Feb 11 13:56:09 GMT 2007


For everyone interested, here is list of OSM ways (IDs) that have some 
problems with them (e.g. duplicate segments):

http://artem.dev.openstreetmap.org/files/notsimple_ways.txt.bz2

Also, JOSM seems to be quite happy about ways like that:

http://artem.dev.openstreetmap.org/files/josm.png
( self-intersecting)

JOSM developers: Can we have an extra 'testing for self-intersections' step 
in 'Create Way from segments' tool? 

Cheers,
Artem


On Sunday 11 February 2007 12:26, Artem Pavlenko wrote:
> On Sunday 11 February 2007 12:09, Jon Burgess wrote:
> > On Sun, 2007-02-11 at 11:11 +0000, Artem Pavlenko wrote:
> > > Hello ,
> > >
> > > As a attempt to imporove text labels rendering on OSM I was looking at
> > > the original data and I found _lots_ of cases like this:
> > >
> > >  <?xml version="1.0" encoding="UTF-8"?>
> > > <osm version="0.3" generator="OpenStreetMap server">
> > >   <way id="4253585" timestamp="2007-01-27 15:00:40">
> > >     <seg id="21468107"/>
> > >     <seg id="21468110"/>
> > >     <seg id="21468110"/>
> > >     <tag k="highway" v="unclassified"/>
> > >     <tag k="created_by" v="YahooApplet 1.0"/>
> > >     <tag k="name" v="Southampton Place"/>
> > >   </way>
> > > </osm>
> > >
> > > Note there are duplicate segments. I was trying to use JOSM to fix this
> > > but in JOSM I can only see two segments for this particular way.
> > >
> > > What is the best way to fix this?
> > > Can we have some constraint on master OSM database to ensure ways don't
> > > have duplicate segments? Can OSM editors (prior uploading) check this?
> >
> > As a workaround, osm2pgsql could discard any duplicate segments in a way
> > fairly easily.
>
> Yes, absolutely!
>
> Still it is important to address this in the original dataset, IMHO.
> Artem
>
> > 	Jon
>
> _______________________________________________
> dev mailing list
> dev at openstreetmap.org
> http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/dev




More information about the dev mailing list