[Tagging] Deprecate ref=* on ways

[Paul Johnson]

On Mon, Jun 2, 2025 at 6:48 PM Rob Savoye via Tagging <
tagging at openstreetmap.org> wrote:

> On 6/2/25 4:58 PM, Paul Johnson wrote:
>
> It probably wouldn't be a bad idea to start formally deprecating ref=* on
> ways.  Ref to describe routes on ways violates the "one object, one OSM
> entity guideline," and is one of the reasons relations were created as a
> primitive in OSM in the first place.  If the ref=* on ways dinosaur was
> allowed to die the death it should have 15 years ago.
>
> I'm not sure if this would be an improvement for those of us focused on
> mapping in remote areas. In the US, many of these were imported into OSM
> from TIGER ages ago, others traced using imagery. In OSM there is a large
> percentage where the highway geometry is completely wrong, going the wrong
> way at a fork. After fixing the geometries, and updating the name and USFS
> refs here's my question. USFS refs change frequently, often at every
> intersection, but then name stays the same. The name would I assume be the
> relation name, but making each member way (with no tags) in the relation
> could loose the USFS reference, which is actually important. In that case
> the route relation could have the name, but each way would need the ref.
> And some have multiple refs, USFS, and country or state, but the name is
> the same. Out here in the western US, the USFS and BLM refs are super
> important for those of us responding to emergencies in these remote areas
> for wildland fires or backcountry rescues.
>
I'm pretty sure 4 digit Forest Service roads that are a single way and have
no reason to be divided (since they tend to be forestry tracks of only a
few hundred meters) really can't violate the one object, one OSM entity
guideline by tagging ref on way, so that'd be a no-change situation either
way.  I'm focusing on all routes that have 2 or more member ways, for this
reason.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/tagging/attachments/20250602/f1904bf1/attachment.htm>