[Tagging] Is a sidewalk always adjacent to a traversable road?
simone.saviolo at gmail.com
Thu Aug 25 13:24:46 BST 2011
2011/8/25 Josh Doe <josh at joshdoe.com>
> On Thu, Aug 25, 2011 at 4:30 AM, Simone Saviolo
> <simone.saviolo at gmail.com> wrote:
> > Considering the above, I can hardly see how a crossing (either for
> > pedestrians or cyclists or horsemen) would deserve its own way
> (especially a
> > highway=*). The only "need" for it - and mind it, it seems quite a good
> > point to have it IMHO - would be routing. For sure, if it seems fit to
> > the way, it should not be a highway=footway (using pedestrians as an
> > example).
> A highway=footway is a route that pedestrians are intended to use or
> use on a regular basis, it doesn't always have to correspond to
> physical infrastructure. In other words if there's a well worn path
> through a field or forest that should be considered a highway=footway
> (or possibly highway=path, but that's irrelevant here). I don't see
> how there's any difference between a route that a pedestrian follows
> through a field and one that crosses a street, what we're doing is
> describing the *network* of ways that pedestrians travel on.
So far I've thought of highway=footway as a "path" (not in the OSM sense)
that is dedicated to pedestrians, but it's not a road closed to motorized
traffic. For example, a pedestrian passageway through a park, or a
pedestrian bridge over a railway. But if it's like you said, then I guess
that a crossing might fall under that name.
> > I suggest using highway=crossing for that. Using it on nodes is only
> > appropriate to mark at what point of the road vehicles may encounter
> > pedestrians trying to cross. Secondly, highway=crossing ways would be
> > clearly indicated as routable segments that are not actually part of, for
> > example, a cycleway network.
> For footways crossing streets we have the footway=crossing tag. I
> suppose we could get away with a highway=crossing tag instead, but
> then of course it's critical to add foot/bicycle/etc=yes/no. And all
> data consumers would have to change to work with this new highway
> type, not that it's something I'm opposed to, but it's worth
> considering if the distinction between highway=crossing + foot=yes and
> highway=footway + footway=crossing is worth the hassle.
This is nice too. I have no particular preference for any of these methods,
and I wait to see what other people think.
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the Tagging