The road properties are mapped on the road. The relation (also for motorways) should not impose this. The E34-relation contains segments with traffic signals, so it's not all freeway/motorway. We should differentiate between fiets-o-strade-the-cycleway and fiets-o-strade-the-route. The latter can contain cycleways and residential roads imho.
Maybe we should invent something like highway=cycleway;cycleway=fiets-o-strade for the former. On Fri, Feb 2, 2018 at 8:47 AM, joost schouppe <[email protected]> wrote: > @Ben: I was also thinking about proposed, but I don't like how this is not > in the main tag. Instead of route=bicycle + state=proposed, I would prefer > something like route:proposed=bicycle. Then you automatically break data-use > that hasn't heared about these 2000 objects worldwide that have a breaking > extra tag. I can see how that would be annoying. Still, the specialists > (cycle layer, waymarkedtrails) seem to know about it, so I suppose that ship > has sailed. > > @all: so it would be OK to map "official detours", if signposted. > > Subquestion: But a Fietsostrade is more than a cycle route, it also comes > with certain assumptions about road properties (much like a car motorway > does). The official detour would definitely be missing those features. I > would say that as long as there is no one-on-one between the route and the > road features, data users should not make assumptions about "this is part of > a fietsostrade route, hence it will have fietsostrade infrastructure". And > we should map those properties that make it special on the segments only. > The alternative would be to have two types of relations or two types of > roles within a relation. > > -- > Joost > > _______________________________________________ > Talk-be mailing list > [email protected] > https://lists.openstreetmap.org/listinfo/talk-be > _______________________________________________ Talk-be mailing list [email protected] https://lists.openstreetmap.org/listinfo/talk-be
