[Tagging] How to tag oneway restriction applying to pedestrians?

Jmapb jmapb at gmx.com
Sun Jan 12 17:10:52 UTC 2020

On 1/11/2020 7:13 PM, Jarek Piórkowski wrote:

> On Sat, 11 Jan 2020 at 18:18, Jmapb via Tagging
> <tagging at openstreetmap.org>  wrote:
>> https://www.openstreetmap.org/way/97010406
>> - It was originally a vehicle route but was changed to pedestrian with
>> painted bike and foot lanes. For motor vehicles, only emergency and
>> specifically permitted delivery traffic is allowed.
>> - It was *always* one-way, and the one-way signs are still there.
>> Bicycles and permitted motor vehicles are required to follow the one-way
>> signs.
>> - Pedestrians can move in either direction, and this is explicitly
>> indicated by painted marks in the pedestrian lane. (Thus there's a
>> oneway:foot=no tag, and it's worth noting that OSRM respects oneway:foot
>> and routes pedestrians "backwards" but GraphHopper does not.)
> That's a good counterexample - thanks.
> I was thinking of a somewhat similar example of Stanley Park Seawall
> in Vancouver, which is also one-way for cyclists, but is mapped with
> separate ways for footway and cycleway. However the Seawall has a
> physical separation in form of a small curb between the two modes, so
> that's defensible. From Esri imagery it looks like Prospect Park ways
> are separated by mode only with paint, so having separate ways for the
> modes is not as elegant or arguably correct.
> So it looks like we will indeed need a new tag to specify one-way-ness
> for pedestrians.

Correct, the Prospect Park drives have paint separating the lanes, but
nothing physical. So mapping separate ways would be unorthodox.

Personally, I have no problem with oneway=yes having different
implications depending on the value of the highway key. In general I
would expect the oneway value to align the predominant use of the
highway in question.

More specifically:

  - I would expect a oneway=yes tag apply to foot traffic on footway.
  - I would also expect a oneway=yes tag to apply to foot traffic on
pedestrian, path, and cycleway -- unless explicitly nullified with a
oneway:foot=no tag.
  - I would not expect a oneway=yes tag to apply to foot traffic on
track, service, unclassified, residential, or any larger roadway, unless
made explicit with a oneway:foot=yes tag.

Of course I understand that from a data consumer's point of view it's
irritating when a tag has different meanings in different contexts --
especially if these differences are not formally documented.


More information about the Tagging mailing list