[Tagging] traffic_signals:direction=* vs. direction=*

Kevin Kenny kevin.b.kenny+osm at gmail.com
Mon Mar 20 17:47:46 UTC 2017

On Mon, Mar 20, 2017 at 12:19 PM, Jean-Marc Liotier <jm at liotier.org> wrote:

> Given how widely used the direction tag is for highway=* signs, why
> isn't it also applied to highway=traffic_signals ? Does
> traffic_signals:direction=* bear additional meaning that direction=*
> does not convey ?

If there is a traffic signal at an intersection, the default is that traffic
from any direction will be facing a signal. There's no harm in assuming
that a guidance application might warn that a driver is approaching
a signal.

By contrast, traffic signs are inherently directional. Without a direction
indication, we really have no way of conveying "traffic on the side
street has to come to a stop/give way here; traffic on the main street
can proceed relatively unhindered." There isn't quite a universal
consensus on how to convey that information. What appears to be
most popular, as you have observed, is to place the stop or give_way
near, not upon, the intersection, as a node on the way that it applies to,
and apply a direction=* tag to indicate the direction of traffic flow
that will be facing the sign.

Unfortunately, OSMAnd and whatever it uses for its routing and
navigation engine doesn't recognize this scheme, and alerts about
a STOP sign as you're leaving the intersection, as well as when you're
entering it.

I don't think with the present state of data consumers, that there is
necessarily a "right way" to do this. Ultimately, it will be the data
consumers that will decide what the "right way" is: tag your
STOP and YIELD signs correctly, and the applications will warn
drivers; tag them wrong, and they won't.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/tagging/attachments/20170320/2524ead2/attachment.html>

More information about the Tagging mailing list