Hmmm aren’t you saying the exactly the same thing? Changing the tagging to something new will
workaround the overloading of highway=path for consumers that either won’t include rendering/routing for the new tag or care enough that they will devise appropriate rendering/routing,
provide an outlet for mappers that want to map this stuff that have previously being using path (the reason I suggested using something a bit more generic than scramble is because using that will likely just end up in people using path because “scramble” doesn’t quite fit),
It won’t fix, regardless of which tag we end up using, the case of people that don’t care and use inappropriate rendering/routing, which is why I don’t think the argument around which specific value we use is particularly relevant on the -consumer- side.
I’d say yes it does. Below UIAA grade 2 or YDS class 4 a route only requires simple scrambling that any able bodied person can manage. Above these ratings the route requires technical rock climbing skills. A route like this is entirely innappropriate for a person lacking these skills.
And those app developers tend to be the big guys: Mapbox, Facebook etc. etc. The smaller sites which care about these things are already parsing sac_scale and the like.
The great thing about highway=scramble is that it offers a failsafe, easily implementable, easily understood solution to every data consumer - basically unless you’re a specialist hiking map, do not parse ways with this tag. Compare to the current situation where you have to understand and parse a whole bunch of different, complex tags: most OSM data consumers simply don’t have someone on the staff with detailed enough knowledge of OSM tagging. (I write as the person who runs the site with probably the most complex OSM access tag parsing that exists.)
highway=scramble is a proposal which, if adopted, will make OSM distinctly better for its end-users in the space of a couple of months. If it isn’t adopted, then we will still be in the same situation two, six, 12 months from now, and people will still be falling off the side of mountains because their map told them to.
But I’m happy to come back to this thread in 12 months’ time to eat humble pie if one of the naysayers has managed to convince all the big data consumers.
You are being a bit optimistic about the ability of the general populace to do UIAA grade 1 but that’s besides the point.
The problem is that you are not in alignment with the actual highway=scramble proposal (that includes UIAA grade 2) and that is typical of the issues around the whole subject matter, aka there will be endless debates about when scrambling ends and when technical climbing starts. So I would rather have a generic tag that we can add difficulty scales to and leave it at that, and yes that will lead to overlap with the climbing route tagging at the higher end.
Well I didn’t say “general populace”. I said “able bodied person” . But yes, point taken.
Forgive me. As an American, I am familiar with YDS classes, but not so much with UIAA grades. Yes there will be regional variation on where exactly the cuttoff between scrambling and technical climbing is. Regional variations are not uncommon and I don’t see a way to avoid them regardless of the tagging scheme.
To make it clear I was just reacting to @Hungerburg declaring the proposal dead and pointing to some possible ways in which we could move forward. That was before @Hungerburg decided that it still may have some life in it.
I could totally live with highway=scramble even though I’m not in love with it for the already mentioned reasons and some more (for example it will lead to arguments about exposed paths that are dangerous but clearly not scrambles).
I still do not see much chance of the proposal getting approved. Main thought to extend voting was to gather more insight from the voting public while the topic is hot and watching this space here, where I heard, that there should be better ways to address some of the same stuff, without the voting result sending prejudice.
That is why this proposal. Scrambling starts, where walking ends (UIAA I, and if my understanding is complete, YDS class 2)
Certainly, no upper limit would allow climbing routes to get the “missing physical” that QA tools complain. UIAA II was chosen, because that is the limit of SAC T6 (and therefore path, at least in my reading of the spec,) and also, because it is so-called non-technical, can be done without gear, ordinary boots, no extra luggage for hikers. So it can rightfully be called “The highway=steps for trails”. If the base got upped, I’d fear the result being something similar to path, where again all the info is in the attributes.
Not quite. YDS class 2 is walkable terrain but with the possibility of occasional use of the hands for balance or comfort. Only once class 3 is reached are hands required to advance.
I’m not sure if these two comparison tables are entirely accurate, but they both show YDS class 3 & 4 as equivalent to UIAA grade I and UIAA grade II is not reached until YDS class 5.1: Mountain Project, Wikipedia
It seems to me that there are different objectives which are sometimes overlapping in this proposal and the various discussions around it, which are quite interesting. I think it might help to gather those objectives and have a look if this proposal or some of the other ideas is a solution for this.
Just some of the objectives I gathered:
“Dangerous paths” shouldn’t be part of highway=path, as secondary tags aren’t enough for non-hiking maps/routers to render them appropriately; so a disruptive new highway=? is needed.
Scrambling parts of hiking trails should be mapped, so we need a tag for that.
The definition of highway=path is too wide, so every small solution is a step in the right direction
A new value of a highway tag should be so obvious that misinterpretation is rare
The state of path related tags is alright, data consumers are the problem, nothing to fix here
Of course the list is longer (or I’m not quite getting everything), what I want to say - highway=scramble (or something else) can’t be the solution of all of them, and it would make it a bit easier to solve at least one thing if the discussion would more about objectives than semantics.
I’d appreciate you creating https://wiki.openstreetmap.org/wiki/Proposed_features/demanding_paths and file a Draft there. I chose this title, because it found some support in the voting community by people involved in the subject matter, but you are of course free to choose any other by modifying the URL. Preferably this month, so debate can continue without downtime. You do not need to subscribe to the mailing list, but you can easily create a topic here in this community.
The table confuses me: Somewhere I read, YDS once had class four the top, meaning serious climbs. Your account of class two reads a bit like SAC T3: there may be something on the route that may make you use hands for balance, but anywhere, nothing of note, could be mapped as a node barrier=step on a path. Unlike SAC T4, where you are certain, that you will have to use your hands for balance at least several times. I find it curious, that you say for comfort; This came up in RfC, and I used it to subtitle the T6 picture, a bit playful, mostly meant as a joke on the video linked in RfC showing a monk in Asia walking hands-free supposedly UIAA III terrain; still, I liked the term and I find it quite fitting. Where falls likely fatal, use of hands a good choice.
Honestly, highway=demanding_path is a terrible idea. Within five minutes someone will have walked across a field with a bunch of mud and some vaguely threatening-looking cows and decided that it’s “demanding”. If we have learned anything from the smoothness= debacle I would hope it’s to avoid tags like that.
A scramble is a scramble is a scramble. You can mither around the edges about the National Alpinism Council of Kansas and their 300-point scale or whatever and that’s the sort of thing that OSM people enjoy, but ultimately 90% of the time a scramble means “you have to use your hands to make progress”. It’s an easily understood top-level tag and you can fine-tune it with details all you like. That’s the sort of tag that works well in OSM, from mapper to editor to consumer to end-user.
Yeah, so? I’d actually want an above ankle-deep mud and vaguely threatening cows to be mapped with highway=demanding_path (or whatever value it happens to be).
Actually, for my use case I’d prefer highway=grandma_hostile_path, but I don’t really care about name.
(It could be highway=demanding_path or highway=scramble or highway=Q123456, as long as it would never lead poor grandma there - nor even show her that this way exists!)
What would be the primary purpose, as others have noticed, is for that tag to stop being rendering and routed everywhere by default. Only after intentional work by mounteering-app-developer (or whoever) should it become usable (and then hopefully rendered differently - but even if not, it would still fulfiled its primary purpose).
The purpose being, when grandma does open www.osm.org or their granda-routing-app, she will only see nice walk in the park paths, and never be presented with other type of path: “yeah, any ironman finalist has almost 90% chance to pass this, but it is still highway=path. See, it has also ironman_scale=90% and smoothness=impassable and sac_scale=T99. Neener-neener-neener!”
I absolutely agree with that part! Exactly what I am talking about!
Although, there are no mountains here, so you might be wondering why a I care a lot about highway=scramble or similar. I’d be marking ways with halfway-up-to-your-knee-mud and be-prepared-to-be-mightly-scratched-to-get-through-those-thorny-bushes. Hey, you have to use your hands or you’d lose an eye, so it is kind-of-scramble, right?
Yes, I currently map them with: highway=path + surface=mud + informal=yes + foot=discouraged + smoothess=impassable + trail_visibility=horrible + sac_scale=demanding_mountain_hiking + obstacle=vegetation + note=do not go here, and you know what? www.osm.org and mostly everything else still navigates my grandma there.
That is my use case. Now, before I get beaten to death (appropriately using both foot and hands ) with “yeah but we don’t care about that but only about scramble!”, do note that it does not preclude that! Instead, highway=grandma_hostile_path (or highway=demanding_path or similar) only widens the possible usages.
Extra tag demanding_path=scramble would when used in combination with highway=demanding_path would mark scramble exactly as well as highway=scramble!
But it would also leave open the possibility of having demanding_path=jungle instead, or demanding_path=knee_dep_mud, or demanding_path=corrida, or demanding_path=active_volcano. (thus widening the circle of interested users even if they don’t particularly care about mountains and stuff)
The elephant in the room is that what we really want to stop being rendered / routed are paths that are significantly hazardous aka you are going to kill yourself if you don’t know what you are doing / are not properly prepared over what an average person would expect is necessary.
Simply not being able to use the way because it is “grandma hostile” (me thinks “couch potato hostile” would be better, but that’s a different topic) isn’t really in scope and is already covered by attribute tags (surface, smoothness etc)…
And yes
there are lots of scrambles that are perfectly safe, but it seems to be a sensible difficulty cut off in light of the perceived people killing themselves because of OSM (we’ve talked ourselves in to that that is a thing)
we still need to rely on data consumers to interpret attribute tags in a sensible manner. Creating lots of different path types for every conceivable hazard along the way wouldn’t seem to scale.
I think it is not practical, hence not wise, to rely on combinations of attributes to determine what a feature is. As soon as you hear yourself say something like “oh, that thing T is easily recognizable because it has attribute a and dimension max d and category c1, c2 or c3 in system S” you better think of a single main key value T or a single secondary tag with value T. Where T is what people generally call it.
Example 1: highway=steps
Example 2: bridge=yes
and there are also many ways that aren’t “safe”, although you do not need to or even can use your hands (e.g. you risk of falling off a cliff, on a modestly wide path that runs along it, I posted some example image about such situation in German hill areas).
Generally, I would not consider a scramble more dangerous than a path where you do not use your hands to improve balance. The danger in the mountains comes often from situations where you risk to fall down a significant height (we ignore weather related dangers here), which is not directly related to scramble.
Currently, we use “path” for anything up to T4, while scramble would take the usage away from path but extend it up to T6 (if I understand the proposal correctly, which is illustrated with a T6 example).
In actual mountains, it will typically take you some hiking to arrive at such places, so grandma would not have to be protected because she doesn’t arrive there anyway. But in some hilly areas, you can find similar situations as in the mountain regions, but below 1000 meters, and there it seems reasonable to have a first level distinction between almost climbing and using your hands, and walking with a stroller or a wheelchair, e.g. this is 460m above sealevel:
This is wrong: Path currently is used for anything up T6, completely in line with what the documentation says. The proposed tag scramble is not meant to rebase all of that, but instead to highlight the scrambling sections therein. Precision surgery, a.k.a micro-mapping.
From the wiki, I see that T5 and T6 are listed as possible combinations for path on the highway=path page.
These are not so many in absolute terms, 4499 T5 ways and 2249 T6 ways (total uses of sac_scale, not all are in combination with highway=path, e.g. there is via_ferrata, but from an unrepresentative check on overpass turbo, many/most of them are highway=path), out of 700.000 uses of sac_scale: sac_scale=difficult_alpine_hiking | Tags | OpenStreetMap Taginfo
. But yes, point taken.
) with “yeah but we don’t care about that but only about scramble!”, do note that it does not preclude that! Instead,