After two years asking questions, I have not found any robust distinction between foot and hiking. Wouldn’t mountaineering just bring the same kind of ambiguous distinction?
I believe there is a notion of scale, as in “zooming to change scale” that is not well captured in current definitions. Even in the steepest of climbing routes we can find extremely small parts that would qualify as “foot path”…
Could things be more clear if we said something like "path difficulty (e.g. sac_scale) must not be translated into a tag (e.g. mountaineering) under a certain length? It seems to me that this kind of principle would bring a series of logical consequences that would help us.
I ventured to start a fresh topic, RfC: Highway=Scramble I tried to address some of the issues brought forward here.
half kidding here: why not highway = none?
We are having a debate on the French-speaking OSM forum about removing a path that apparently has no reality on the ground but still belongs to a route (it is in a rocky place)
Please note that I am not against a new highway= value, just I found the limitations on splitting the ways unnecessary and overreaching.
Another option is ways without tags, but members of a route relation.
“Canoe routes can include waterways, untagged lines showing the approximate path in a pond or lake,…”
Especially for short sections of marked trail across rock, passable to typical people, I would still mark it as path/footway.
And would treat marks on rock (smoothened rock, vegetation removed, lichen scrubbed or trail signs) as sufficient to overtake “has no reality on the ground”. See say File:Krywan podejscie.jpg - Wikimedia Commons or another view showing marks on rock (also on top, where solid rock is present).
But note that there is Tag:highway=no - OpenStreetMap Wiki has already over 2000 uses
However, occasionally, there may be ways that one would expect to be highways, but are not. In this case, highway=no can be added to explicitly indicated to other mappers and data users that there is no highway here. This allows us to distinguish between the case of a highway type being unknown and needing to be determined, and the case of the way being known not to be a highway.
For example, in the UK, there are sometimes legal Rights of Way that are inaccessible, resulting in a legal highway that is not physically present or usable on the ground. Due to administrative errors, unauthorised constructions or lack of maintenance, the official line of a legal route may pass through dense woodland, though walls of a building a building or though a lake. highway=no can be used in these cases to explicitly mark the way as not being a highway, while still recording the official line.
This tag should not be used to highways that physically exist but have restricted access. This use of this tag for such cases is deprecated, in favour of highway=* + access=no.
I’d say, the “path” up Mount Everest deserves “highway=mountaineering”. Such a key would be perfect, to fix the note you placed there, wouldn’t it?
not entirely sure - is there even a strict route there? Or various group using navigation markers (like dead bodies) and route is not constant?
From the frequency, there should be a path better visible than in the picture of yours from Poland. On snow, a single person can make a path of great visibility. From what I know, on a lot of the ascent to Everest there are fixed ropes, so people do not stray so much. You certainly have seen the photo of the queue. Visitors only use it seasonally, no idea, if there is a path out of season.
PS: for OSM trail_visibility key values trail blazes do not factor in. This was deliberately decided when coining the key.
“Path or next marker always visible, but sometimes has to be searched for” - Key:trail_visibility - OpenStreetMap Wiki
Also, it seems that it used to be described differently - Talk:Key:trail visibility - OpenStreetMap Wiki
Funny, the SAC scale https://www.sac-cas.ch/fileadmin/Ausbildung_und_Wissen/Tourenplanung/Schwierigkeitsskala/Wanderskala-SAC.pdf does not mention this. They only talk about the “Trassee” (trail) visibility on the ground (OTG), and that does not include markings. Many taggings in OSM then are wrong, as only ways that are not trail_blazed can be anything but “excellent” or “good” in OSM.
So I messed up here. But I never understood, why the sac_scale tag and the trail_visibility got separated, it was very late in the proposal process, to make tagging more OTG, seems to be the opposite came to be.
Broadly I’m in favour of the idea, although probably in the highway=mountaineering form. However, there are other similar “routes” which aren’t continuous marked paths, but may be extremely well known by people interested in them:
- Traditional alpine routes. Hoernligrat, Matterhorn Nordwand, Brenva Arete, Eiger Nordwand. These would typically be graded by traditional alpine scales F/PD/AD/D/TD/ED (French) /WS/ZS/S/… (German) or their equivalents in other countries. Some at the lower end may also be graded with SAC scale, or parts of the approach could use this tag.
- Expedition style ascents in the Himalaya, Andes etc. I’m not aware of any traditional grading for these, and their nature is completely different if one is making the first ascent in season, leading, guiding or being guided. The classic example is Everest via the Western Cwm and South-West Ridge.
- Alpine style ascents in distant ranges. Again I’m not aware of gradings, but keen amateur mountaineers may attempt peaks of 6000m & over (I know at least one OSMer who has done this).
- Mountaineering ski tours: with ascent & descent. Obvious examples of well-known routes are the Haute Route & the Patrouille des Glaciers.
- Downhill ski routes which lie beyond typical off-piste descents. The Mer de Glace is a well-known example with few difficulties, but many objective hazards. It is usually skied with a guide. In the Chamonix valley there are hundreds of these routes, some of which are not only technically very demanding, but involve severe objective dangers. A number of published guides are available for various favoured locations in the Alps, but in other locations knowledge will be limited to the local enthusiasts.
- Bushwacking hiking routes in back country. This applies in particular to parts of North America, Australia and New Zealand, but there are also plenty of places even in Europe (e.g. NW Scotland). I presume in tropical countries where vegetation grows fast there are well known routes which require use of a machete to traverse.
Ideally I’d like a solution which also encompassed the ability to tag these other options. highway=mountaineering would cover most of them (
Changed topic to trending terminology. First sentence in post above referenced the old title “Sport=Mountaineering”.
Tagging paths indeed needs some improvements. It would help if OSM database recognizes that difference among walking path, hiking/mountaineering trail and alpine climb is significant and required different type of highways instead of subtle and often hard to understand, implement and render decorations with additional tags.
Situation with tagging paths is like if we have just one type of highway for the all roads and then have do make distinction using additional tags. For roads that is not the case. Various types of roads are recognized as different highway types which is logical, useful and very helpful. Same should be with various types of paths. there should not be just one type of path highway.
I would not worry much about updating existing data. If tagged correctly, most of them could be automatically updated to new highway types.
That would also make some force one renderers to pay more attention to rendering such highways. For start, all those that do not render them properly now, would stop rendering them at all, as that would be new highways, they do not recognize - until they update, meaning pay attention.
Other than resolving issues like this which initiated discussion, it would help properly tag all other pedestrian ways, especially in rural and mountain areas. It seems like that people who designed highway tagging scheme did not pay much attention on way objects other than regular urban traffic. Tagging pedestrian ways out of urban areas is quite lacking.
So, I would gladly vote for adding new types of highways to differentiate walking paths, hiking/mountaineering trails and alpine climbs as such. Those three I meet regularly but I guess there are more types that could be considered too, as others already pointed out.
Situation with tagging paths is like if we have just one type of highway for the all roads and then have do make distinction using additional tags.
no it’s not, we have footway, cycleway, bridleway, pedestrian, via ferrata and path…
And in the ancient times of OSM the proposal for path wanted to replace footway, cycleway and bridleway so until today it’s often equivalent and the wide definition can mean all or nothing
I am constantly perplexed, why data consumers consider what is mapped as highway=path an easy walk or even something, where you can ride a bike on. From the Prosecco Bar to the summit of Mönch, OSRM gives 41 minutes - OpenStreetMap - whereas https://www.sac-cas.ch/de/huetten-und-touren/sac-tourenportal/moench-1194/hochtouren/ (watch out the nice map!) says 3-4 hours. Brouter with https://raw.githubusercontent.com/poutnikl/Brouter-profiles/master/HikeProfiles/Hiking-Alpine-SAC6.brf gives 2:01 hours. Something definitely not out of reach for trained mountaineers. Speaking as someone, who met people, that would correct me, if I said three quarter hours, to restate 41 minutes instead; Not this hike though.
PS: Aerial view of a section of the path here https://upload.wikimedia.org/wikipedia/commons/9/93/Aerial_View_of_Mönch%2C_Switzerland.jpg
Perhaps because highway=path is also used for generic “shared use” paths that are genuinely suitable for bicycles. The wiki describes a path as open to all non-motorized vehicles by default.
Presumably this path would be smoothness=horrible or even worse, I don’t know if routers would take that into account as it is directly related to wheeled vehicles (whereas using sac_scale would require interpreting what a tag intended for pedestrians implies for a bicycle).
For anyone who isn’t following all 3 threads, this topic is being discussed in other threads as well:
While the thrust of each discussion is similar, there are a few interesting ideas that have popped up in one versus another and slightly different key suggestions have been discussed (
Because of ambiquity of meaning if term path. It may be interpreted in various and sometimes contradictory ways and even that does not caver all on the ground situations.