Topic Alpinist routes marked as footpaths was created in this community forum instead of going edit war. So far so good. In my opinion, the subject deserves community attention:
Is tag
highway=pathappropriate to map pathless Alpinist routes over UIAA III terrain, intended for use by expert mountaineers?
I opened a poll there, Alpinist routes marked as footpaths - #39 by Hungerburg – Please vote!
Do we have a photo of representative location?
Does voting “no” to mapping these routes as highway=path mean:
At the upper T5/T6 levels, the nature of difficulties can be so different (climbing, exposure, glaciers, loose terrain…) that it’s very hard to make such general statements without knowing specific cases, like the one mentioned. That path “in some places presents third-grade difficulties”. If these are only few and short, without exposure or can be easily secured with a rope for someone with less climbing experience, it may be much easier than an extremely exposed ridge hike.
This video shows the section from the Bivacco Ceco Baroni to Bocchetta di Levade that according to the changeset comment “in some places presents third-grade difficulties”. I couldn’t find anything else showing the eastern section beyond Bocchetta di Levade except the photo in the other thread.
To me, voting no just answers the question of whether highway=path is appropriate for these routes. I would say a no vote takes no position on how these routes should be tagged instead or if they should be omitted from OSM. That would need to be a separate poll question.
I would vote "no, except
sac_scale and trail_visibility"Isn’t the fact that a way is part of a hiking route adequately recorded by it being a member of a hiking route relation? Why should this also indicate highway=path? I see no issue with an untagged way (or some other non-highway=path tag) being included in a hiking route relation.
Certainly, if you are of the opinion, that openstreetmap data should only contain reproducible information on facts about what is on earth, these routes have no place in OSM.
Nevertheless, such stuff is in the data, the very case is exactly about that. So my question in other words: Is highway=path a proper and accepted method to map stuff, that has no On The Ground Truth to it?
I did not yet look at where the poll stands, from remembering last time and from reading along here, this seems widely held fine. Link to poll.
PS: In a recent topic in the Austria category on a similar concern, all but one commentor agreed: It is the trampling that makes the path, not the ideas of people creating a route.
Meanwhile the Austria topic got a new “pro” message – Weglose ‘Wege’ im Hochschwab
It is written in English, likely not from a local. The gist of it to me: Everywhere anybody ever rambled or scrambled a path can be drawn. If there was not a path there, then how could they have passed the area?
PS: In my local area, recently a path was mapped – by a non-local, according to changeset comment from Strava heat alone – there indeed showing a faint broken line over non-trivial terrain. A month later it was deleted by somebody whom I trust to map from on the site knowledge.
Obligatory xkcd.
I don’t think that was the consensus there. Something that is clearly marked or where people inherently know where to walk because of widespread common knowledge (in the relevant community) can also justify being mapped. Track marks on the ground are a good indicator that there’s a path there, but are not always obligatory (e.g. on rocky terrain where people don’t leave tracks).
I would add that on grassy areas there aren’t usually any stones for building cairns.
On wide areas people seem to choose whatever direction they please, therefore not beating the path in the exact same location as in the narrow passes. Hence the beaten paths in the narrow-pass sections are much more visible and obvious.
A “fact about what is on earth” can also mean that you can get from place A to place B easily and without having a clear line on the ground to follow. I think one of the main points of a map is exactly that (think “treasure map”, to illustrate the point).
I would also add that, in English, there are terms like “path to success”. These are obviously not clearly visible, either. A Path in general language is a bit broader term than just a line in the sand.
This is the heart of this poll: Is it OK to map purely ideal things highway=path? C.f. [RFC] Feature Proposal - Flowlines
BTW: The message from CAI Salò is fully sensible from a mountaineering point of view. Yet, I question that such should be filed under same OSM category as commune maintained 3m wide paved so called shared foot/cycle/horse use infrastructure.
Indeed, it is very broad. This unfortunately has made it a poor choice for an OSM tag. I’ve come around to avoiding it in favor of more specific options wherever possible.
Yes, absolutely. For that specific case it seems that highway=scrambling would be a much better fit.
Edit: The fact that there are SAC Scale, surface, visibility, and other tags available, which further clarify what type of path it is, makes that point moot.
“Primary” tags such as highway= do matter though, for at least two reasons:
Consider highway=living_street for example - the differences compared to highway=residential are relatively subtle and presumably could have been expressed by extra tags, yet it has its own primary tag. The difference between an off-trail mountaineering route and a constructed path seems more important than between a residential and living street.
I absolutely agree with this and is the point I’m also trying to make in my posts. This whole question was triggered (it’s obviously not the only reason but was the trigger) to open the poll and, prior to that, the whole discussion at Weglose 'Wege' im Hochschwab.
I mapped a difficult path on the terrain rarely treaded by mountaineers, with a sporadic (but incredibly helpful and useful) cairn, because I would like the map to serve the purpose of showing the way, and not just display what’s already visible and obvious. In a fog, or even just by looking from a distance, it is not always easy and obvious to know that there is a pass there. Just like in the CAI case, linked in the original post. I find this incredibly useful information and the reason to actually use the map. I don’t need the map to tell me the obvious stuff.
The initial path was based on GPX tracks created by people who went down the route earlier (from a touring web site), later confirmed in person. This is when other pointers were identified. To me, the path/way was clear. I’m pretty sure it would not be to about 90+% of the general population, though. I would have had much harder time had the cloud not dispersed later in the afternoon.
So, I presume, there are really two questions here:
One of my concerns is that, on this particular mountain, there are dozens of difficult paths on OSM that come from the valley to about 100-200m from the plateau and then disappear. To me that’s highly misleading. It can mean that either there is no way further up due to cliffs, or it may mean completely the opposite - now the path is no longer visible but still possible and, after 100m you reach the plateau and can walk to a nearby marked hiking path. I would like to go to some of these routes and it would help to at least know that they are not dead ends, which is at the moment not possible, just based on the map data.
I fully agree that there is a need for another primary tag. One suggestion was to use highway=scrambling, which fits in some instances. My particular case wasn’t really scrambling but a “relatively ok, T4” route. But it is not marked nor very obvious, and goes between cliffs and stones which can be highly dangerous. It is partly on steep terrain, but doable and not harder than many marked routes which are mapped and have a higher grade.
The wiki does not really stress it, but I think highway=footway is mostly used in urban settings with clear access. Using it for remote paths in the mountains is I think quite a distinctive use (and when I see it somewhere, I tend to change it to highway=path, as it has so far always been an exception to the rule with other trails in the area marked as paths). It does not help that some renderers assume it is mostly paved (or very well built), it leading to very weird renderings of streets on ridges. If a new tag for “invisible” paths is adopted, I think it definitely should not be highway=footway.
Would it be a good idea to introduce something like a prefix (tbs:, to be surveyed) that can be added to highway=path that makes these paths invisible to renderers and routers and thereby make them harmless? The prefix could then be removed (automatically?) when sac_scale and trail_visibility tags are added later.
We could also contact general-purpose renderers and routers (and recommend on the wiki) that highway=path with trail_visibility=no and sac_scale=demanding_mountain_hiking or worse should not be shown and be routable.