There were several discussions in the past regarding the mapping of ‘invisible’ paths, but sadly they did not result in any action or even tagging proposal.
I hope to push it to formal proposal after requesting for comments here.
Basic ideas:
“highway=path” + “trail_visibility=no” is an ugly hack and is undesirable
“highway=path” generally shouldn’t be used unless a path is marked somehow (“truth on the ground” - if you look and don’t see a path, then there is no path there)
introduce a tag route=alpine_climbing for alpine routes (applicable to relations and, in simple cases, to ways), e.g.:
I strongly disagree. There are many cases where it makes good practical sense to map such paths.
What matters is if a map feature is verifiable. “Truth on the ground” is just one way of verifying a map feature, but it doesn’t exclude others.
This may be worth considering for ways beyond sac_scale=demanding_alpine_hiking For ways that can be tagged with sac_scale, we have a working system that however is not sufficiently applied by us and not always implemented as we intended by data consumers.
Well, an example would be welcome, but at least in the mountains this is something really undesirable, I think - see, for example, Alpinist routes marked as footpaths
Yes, that is the idea behind the proposal:
Q: Why not use sac_scale=*?
A: sac_scale is for hiking paths, where no belaying/rope usage is required. Alpine climbing generally starts when you need a rope. Also, hardest sac_scale=difficult_alpine_hiking only “includes climbing pitches up to UIAA grade II”. Alpine climbing routes are generally above that grade and, although there could be easy sections (easy ridge with UIAA 1 difficulties for example), they aren’t normally accessible by a hiking route.
Thank You for asking, what people think. Quick thoughts:
Call it route=mountaineering. I know, alpine does not stand for the Alps but for the style developed by Austrian mountaineers in the Alps that spread worldwide, but mountaineering is the British top-term (easily translated Bergsteigen, Hochtouren), and why exclude British expedition style where the routes are the same?
Personally, I would not mind if it ate a bit into highway=path space and included some of what today is mapped as T6 hiking by the SAC – which sometimes overlaps with their “Hochtouren”, it is just a question of equipment brought there: If roped its mountaineering, if not its hiking (therefore much easier?) But this is openstreetmap after all and a lot of people will try to keep a hold of their well-earned privilege of having this mapped as path. So better not make it part of any proposal – dual tagging to the rescue?
What else to pay attention: This will not only cover already mapped stuff (Mt. Everest? a fixed rope climb with not much “technical” difficulty [certainly not a hike though] for most except the Sherpas that plant the ropes at start of season. And a few other mapped routes too.) This may be used to add new routes: So verifiability has to be accounted for.
I use trail_visibility=no mostly for sections of hiking trails that are not visible but are part of an otherwise visible trail, and for connecting visible paths for routing purposes to indicate that walking from one to the other is possible.
OK, thanks for the clarification. I am not interested in mapping these, but don’t mind if others want to map them as long as it’s clear that they are too difficult for most hikers.
I recently learned that footway=link can be used to achieve much the same? Please excuse if I my spoil this topic, but I just cannot help and suppress that observation of mine.
I don’t quite get it. At the moment it implicitly overlaps with T5 (=climbing:grade:uiaa=I) and T6 (=climbing:grade:uiaa=II).
My suggestion is to use common sense/truth on the ground for the tagging scheme selection - if the trail is marked somehow (paint marks/cairns) or is simply clearly visible (well-beaten), then it is a legit hw=path with sac_scale.
If not - why tag it as a path? This is tagging for the renderer.
It is to cover “paths” on Mt. Everest as well (and other similar “paths”).
No ‘path’. Main route according to Strava heatmap 2022-2023.
This clearly needs fixing.
Anyway even a fixed rope “path” exists there for just a month (in spring, in May - not sure if people ascend in autumn), for the rest of the year it’s just snow.
what I like here is the idea to use route as a primary tag and to “eat into the highway=path space” as someone put it.
what I like less is the use of mountaineering or any term ending in -ing indiscriminately for relations and ways, because it would add up to the confusion between the features of routes and the activities that they support.
Well, there is trail_visibility, and no is an approved value thereof. And path is – I call it – greedy. I takes on anything but roads. Way: Pizzo Bernina - Biancograt (403643662) | OpenStreetMap – I’d not call visibility there horrible – Panorama Viewer – More to the opposite: One cannot get lost on such an arête.
Fair enough. What about:
leisure=route
sport=mountaineering
? That captures a bit of context on subject matter quite nicely, I think.
I’d say that in this case one follows the arête, not some path. Send this photo (especially without footprints =)) to a layman who has never been to the mountains, and ask him if he sees a path there
The problem with trail_visibility=no is that it inverts the meaning of the main tag: hw=path says that it’s a path, but then - oh, wait, actually it is not =)) it’s just an unmarked direction =(
We are free to define highway=path as we like, and can choose it to include unmarked directions if that has practical advantages (such as enabling routing). Our definition can be different from a layman’s understanding of the word “path”. That there is a difference is a disadvantage, but I think it is compensated for by the larger advantages of including invisible paths in the definition of highway=path
I understand how this hack works for linking purposes - when the gap isn’t big, but when such invisible paths span across valleys - I don’t know…
Also, as noted above, when mountaineering equipment is needed to move on such a “path”, users of the map start having unpleasant surprises.
We might be free to define anything we like, but the users of the map expect some common sense approach - and here it seems to be lacking with tagging hw=path for mountaineering routes (even with trail_visiblity=no)…
Regarding Unification: This might be just how openstreetmap works:
highway=path
sac_scale=hiking
and
type=route
route=hiking|foot
only is just as contrived when sac_scale is not “mountain_hiking” or some such or route is exactly “hiking”.
Regarding an arête is not a path; I guess that applies very well to that section: For a route it is fine to say, “Then follow the ridge|couloir|whatever_natural_phenomenon to the top”. But does it make sense to add the arete to the relation (after splitting it, of course)?
PS: Still waiting for how @rhhs would gauge trail_visibility on Bianco Grat and of course what @StC thinks about mountaineering when considered a sport.
PPPS: I have no idea if openstreetmap ever will be useful to create so-called Topos (like your Peak Partisan example.) Or, in other words: openstreemap be all that it needs to study in order to go on such an endeavour.
I agree that the use of trail_visibility=no should be limited to where it makes practical sense, i.e. the rule should be “don’t use, except when…” (followed by the description of a few cases where it can be used). In the past I’ve mapped trails along valleys based on just someone else’s GPS trace, not knowing if there is a visible path, but I’m not sure if I would do that again today. Such paths should somehow be tagged so that a user can see that they’re potentially dangerous (Temporary use tag to map potentially dangerous highway=path) and should be tagged with sac_scale and trail_visibility as soon as possible.
I also agree that “paths” beyond sac_scale difficulty should not be tagged with highway=path
That depends on how you interpret “markers” (Key:trail_visibility - OpenStreetMap Wiki). I think a “marker” should be interpreted as “any clue as to where the path is that was created by humans” (could be a worn path, cairns, trail blazes, cut line in a forest, etc.). An arête is not human made, so doesn’t count as a “marker”. The footsteps in the snow are, but are temporary so shouldn’t be considered when mapping. So I would tag the arête that’s visible on the panorama you posted as trail_visibility=no. That it is in fact tagged with trail_visibility=horrible suggests that in some places, there are traces of human use visible.
Of course, I will be biased as I teach map and compass navigation but…
trail_visibility=no
is equivalent to
map_and_compass=yes
Trail visibility is subjective. Some will need to revert map and compass while others will see the subtle path line in the landscape. Don’t assume because you can’t see it that others won’t… and consider seasonal effect with snow, leaves, drought, summer growth etc.
Yes, it makes sense - the route could stray from the ridge/couloir/etc and continue via some other feature (or no feature, i.e. just the rock face) at some point.
Generally, in mountaineering a route ends at the summit (or at least at some prominent feature like mountain_pass=yes), so the idea is to follow this approach.
It probably won’t - topos usually have way more information that can’t be described on the map (general stuff like what favorable conditions/season is best for climbing, gear needed, etc), but such detailed descriptions won’t be mandatory - basically the simplest option (a way from the start to the top) is what is needed for sure.
Complex option is for those who would want to do micromapping.
