They're not at all hard to find in the US. I would wager that fewer than 50% of the actual on the ground road routes in the US have any divided segment whatsoever.
What exactly is captured by two relations that is not captured by one, however? That's the part I'm not understanding. We have tools that (seem to) work fine with a single relation for both ways. They can just as easily check a single relation for the existence and completeness of component ways with forward and backward (or cardinal direction) roles as they can check for the existence and completeness of two separate relations and a super relation. Maybe I missed some earlier discussions on the advantages of the multiple relation model? -Nathan Martin Koppenhoefer <[email protected]> wrote: >2013/11/18 Nathan Mills <[email protected]> > >> I'm still confused as to why the consumers of a relation can't use >the >> forward/backward roles of the ways referenced therein rather than >requiring >> completely separate relations. Why do we need two or more relations >plus a >> super relation per road route even for undivided highways? Even for a >> somewhat experienced mapper like myself, it makes the editing process >that >> much more error prone. > > >if the highway is never divided a single route would be sufficient >(rather >hard to find this situation probably), but if you have routes that even >for >a short way do not share the same geometry it will be easier to >maintain >and to check if every direction has its own relation (and IMHO less >error >prone). > >cheers, >Martin > > >------------------------------------------------------------------------ > >_______________________________________________ >Talk-us mailing list >[email protected] >https://lists.openstreetmap.org/listinfo/talk-us
_______________________________________________ Talk-us mailing list [email protected] https://lists.openstreetmap.org/listinfo/talk-us
