I did not think of traffic directions, good point! I think oneway makes sense better than traffic_direction. This would means that oneway apply to traffic whether it is on ground, water, air, rail... much more easier for routing engines (and for amphibian vehicles such as Duck tours!).

The usecase I've originaly found (but I did not check all 17596 usage) was a usage for intermitent streams (and some which are culverts!) in the middle of the desert of Colorado. In this specific case I don't think any traffic is feasible.


So the question now is: does oneway apply to the direction of the stream or to the direction of the traffic?

 * The wiki pages says oneway is used to "indicate the access
   restriction on highways and other linear features"
   (http://wiki.openstreetmap.org/wiki/Key:oneway).
 * The wiki requires to draw waterways downward (which makes sense for me)
 * http://wiki.openstreetmap.org/wiki/Inland_navigation says nothing
   about traffic directions (but I think it's a real need)
 * JOSM stays silent on this one.

So I would say that oneway is valid for traffic directions on waterways and oneway=yes means that this part of the waterway can only be sailed downward. If my understanding is right, I propose to update the waterway and oneway pages to say that oneway can be used on waterways to represent the traffic direction.

But does it means that a river should be forked under a bridge whenever traffic directions go under different arches?

Yours,

LeTopographeFou

Le 17/09/2016 à 14:45, Colin Smale a écrit :

I would expect that the situation where the flow direction conflicts with the traffic direction is likely to be quite short - under bridges, around obstacles etc. In these cases we could always call on our old friend "oneway=-1" or "oneway=reverse" to mean "traffic direction is opposite to the flow direction". Having to create a route relation is a bit of a sledgehammer to crack a nut.

//colin

On 2016-09-17 14:35, Dave F wrote:

I've seen it used on navigable canals to indicate traffic direction.
If there is a route relation I think it should be indicate with forward/backward roles.
If not then for clarity, maybe something like traffic_flow=backwards?
Adding a route relation would be preferable though.

Dave F.

On 17/09/2016 13:20, Andy Townsend wrote:
I've certainly used "oneway=yes" on inland waterways to document signed traffic flow control, so a blanket removal would make no sense. There may be places where a previous mapper has tried to use it in error to indicate water flow direction, but you'd need to ask whoever the previous mapper was in each case (or use a bit of common sense).
Cheers,
Andy
*From: *LeTopographeFou
*Sent: *Saturday, 17 September 2016 13:17
*To: *tagging@openstreetmap.org
*Reply To: *Tag discussion, strategy and related tools
*Subject: *[Tagging] Use of oneway=yes on waterways


Hi

According to the waterway=stream wiki page (http://wiki.openstreetmap.org/wiki/Tag:waterway%3Dstream):

/If a flow exists, the direction of the way must be downstream (i.e. the way direction follows the flow)/

As of today there is a very small percentage of streams (17593 ways according to taginfo, 0.23%) with oneway=yes.

Is there any undocumented purpose? Is it ok and safe to delete oneway=yes tags for streams?

The same question can apply to drains, ditches, canals...

Yours,


--
LeTopographeFou


_______________________________________________
Tagging mailing list
Tagging@openstreetmap.org
https://lists.openstreetmap.org/listinfo/tagging


_______________________________________________
Tagging mailing list
Tagging@openstreetmap.org <mailto:Tagging@openstreetmap.org>
https://lists.openstreetmap.org/listinfo/tagging


_______________________________________________
Tagging mailing list
Tagging@openstreetmap.org
https://lists.openstreetmap.org/listinfo/tagging

_______________________________________________
Tagging mailing list
Tagging@openstreetmap.org
https://lists.openstreetmap.org/listinfo/tagging

Reply via email to