Proposal talk:Crossing:signed

Sign types to qualify

The US examples are warning signs. There are actual less common non-warning signs. MUTCD/R#R1:_Stop_and_Yield https://mutcd.fhwa.dot.gov/htm/2009/part2/fig2b_02_longdesc.htm
—— Kovposch (talk) 19:50, 26 February 2025 (UTC)

Thanks for that info. I'm afraid that restricting the tag to the yield signs only would make it rather useless considerng how little they are used. I guess a sign that warns about a crossing at this spot (the ones with the arrow pointing down) counts as a sign that tells you that there is a crossing at this spot, too? --Bauer33333 (talk) 20:17, 26 February 2025 (UTC)

@Kovposch: There was a discussion on the forum about these signs a couple years back. [1] Personally, I think it would be more useful to record these signs as highway=stop, highway=give_way, or some variant thereof, rather than as a property of the crosswalk. Part of the purpose of these signs is to point to the stop line; if we only record the presence of this sign somewhere else, then that defeats the purpose. – Minh Nguyễn ^💬 05:12, 4 March 2025 (UTC)

Distinguishing signs by audience

As the to-do section states, a major open question is whether crossing:signed=* pertains to signage directed at crossing users or cross traffic. My first impression is that there's a lot of need for a key like this, but unfortunately it's easy to get confused about the types of signs that are in scope. A tag on the crossing about signing would be useful for a kind of sign that doesn't serve as intersection control but nonetheless alerts people to the presence of the crossing.

At a road crossing, the sign we most need a tag for is one that tells a road user that they're coming upon a crossing. Of course, it's possible to map this sign as a traffic_sign=* node, but we'd still need something to affirm the absence of such a sign. Meanwhile, a sign directed at crossing users usually isn't designed to help pedestrians or cyclists spot an upcoming crossing. Instead, it may have a more complex message, for example with instructions on how to use a call button (button_operated=yes). When a more substantial trail approaches a dangerous road crossing, there may be a stop sign or yield sign, which we already map as highway=stop or highway=give_way, respectively. When a sidewalk crosses a roadway at a signalized intersection, some jurisdictions would refrain from signposting the crosswalk, because road users shouldn't mind the crosswalk independently of the traffic signals.

Railroad crossings are a bit of a quandary. I don't know that there's a strong real-world consensus on whether the railroad is considered to cross the roadway or the roadway is considered to cross the railroad tracks. If we were to go by change of surfacing, then stereotypically a sidewalk would cross railroad tracks, which would in turn cross the roadway. In OSM, neither is split and tagged as a crossing per se, but I do map highway=footway footway=crossing along footpaths that cross railroad tracks. Regardless, most rail traffic isn't stoppable, so a warning sign directed at trains would usually be pointless, or it would need more specific tagging to clarify its nature. crossing:saltire=yes already indicates whether a sign is directed at crossing users, but it's another form of intersection control, and its very specific name causes uncertainty when the sign isn't shaped like a crossbuck.

Of the signs that are currently in scope for this proposal, the only unifying principle I can find is that they're directed at the party that a) lacks the right of way (priority) and b) is vulnerable to the other party. Do you agree with that idea? If so, maybe stating that upfront would help focus the discussion. On the other hand, I have a feeling we'd get into the weeds about traffic rules in various countries, since there can still be a duty of care irrespective of right of way.

 – Minh Nguyễn ^💬 05:54, 4 March 2025 (UTC)