Oh no, definitely not! A new tag specifically meaning “pathless walking route” should be developed.
I only linked to highway=footway as an example of a tag based on a word with a specific meaning. The specific definition of the word footway has helped highway=footway maintain a relatively specific meaning in OSM, while the multiple broad definitions of the word path have caused highway=path to be used so many different ways that it has very little meaning left. A highway=path with no other tags could be a rural walking path, an off road motorcycle trail, a narrow street in a dense city where only motorcycles can fit, a wide paved walking & cycling path in a park, a pathless route in the mountains, a pathless walking route along a beach (under water at high tide), or even an underwater scuba diving path.
This would work for me but seems like a combination of tags that mean one specific thing.
On the other hand, a more specific alpine_path would maybe work better. Other path types could also be separate (I.e. that scuba diving route). Would that be a better option than a new PathPath that will need to be separated later again?
(“underwater offtopic”)
My view on that hasn’t changed since last year…
For the avoidance of doubt, in “beyond the edge” cases a relation (even a route relation) isn’t limited to highway=path. This is an example of something “isn’t really a path” but is part of a route. It’s across open moorland that is also access land so (rarely for England) there are relatively few restrictions about where it’s legal to go.
I suspect that these “underwater paths” are more about tagging for the renderer than anything else, and the mapper there could benefit with a little help with e.g. uMap
At this moment, the poll Alpinist routes marked as footpaths - #39 by Hungerburg is at 6:4 - far from the 3:1 a proposal needs to get accepted. The – a bit related – highway=scramble proposal reached similar result. The community is split on both issues.
From what I observe, when looking at maps or routes produced by consumers of openstreetmap data, they are not sufficiently aware of that highway=path gets used to map mappers ideas of how best to traverse probably dangerous terrain with no on the ground indication.
Is it time to update documentation? How to word that? Who starts the topic?
Anybody volunteering to tell @pnorman that a highway=path does not imply it is 2m wide and paved? Minutely updated vector tiles demo - Last time I read, this is an OSMF funded experiment – I like it, but it has a looong way to go
I do not think one should assume any vehicular travel without other tags. I understand a path as “An unspecified way of passage across unspecified terrain traversable by possibly a rather fit person on foot. Might or might not be traversable on bike, horse etc, but not in a car. Hopefully other tags can explain further!”
… or by horse or bicycle, either. A “well known cycle router” was caught out in the UK by trying to send cyclists down things tagged highway=path with no other tags. It didn’t end well.
Highway=path is commonly used for combined foot/bicycle paths.
I would say access defaults to only foot and bicycle, i.e. access=no + foot=yes + bicycle=yes. All other access has to be specified.
Special requirements (fitness, skills, gear, use of hands), limitations (height of passage, width, surface, terrain, roughness, hazards, incline/decline, banking etc.) default to none, i.e. all have to be specified with extra tags. You can’t solve that with different mapping systems, you need QA to find the errors and fix them.
Just nitpicking, bicycle counts as a vehicle too :-).
Wholly agree. Most highway=paths I walk on are not actually traversible on a bicycle unless you are an expert MTB rider (and even then…).
One more reason to treat paths without any additionl tags with caution. A user needs to employ other information liky topography and common sense when using the map.
Fine with me. However, the worldwide default access table currently specifies foot=yes, bicycle=yes, horse=yes. This matches my experience in non-mountainous terrain in Europe.
Outright impossible to go there on a bike would be bicycle=no, same as if it were illegal. If it depends on skills, gear, type of vehicle and conditions, other tags should be used to indicate the possibilities. Same for foot access and horse access. If none of these can use it, it’s not a path!
I think legal access should be a separate thing from practical access. Even when something is illegal, it might be widely practiced at different places, or one might be in an emergency when the legal restrictions might not (even accordign to the law) apply etc.
However, upon further reflection, https://wiki.openstreetmap.org/wiki/Key:mtb:scale covers for bicycles what sac_scale covers for hikers, so I think to both the same thing applies. Might be easily passable, but without further tags it is impossible to know if is hard (or even impossible, ie. vertical ladders for bikes). Not at all sure about horses.
There is even some mapping between sac_scale and mtb:scale, which can also be assumed.
I agree! I don’t think one should assume much of anything without other tags. With all the different things highway=path has been used for, the only things one can assume are:
it’s not a road
Some kind of travel is possible
it’s used by some subset of pedestrians, horses, bicycles, and motorcycles (maybe other things too)
These assumptions are all far too general to be of much use to a data consumer.
