When I was researching the international systems to try and find patterns and common ground I do remember that some European countries choose to not map paths for T6. Switzerland does, but Austria doesn’t? If I’m wrong @Hungerburg will correct me heh.
Australia does something similar with it’s most rugged rated paths (those I assume are not traditionally technical, but rough and low visibility with some scrambling here or there).
You can make a poll in the forum! It’s under the cog all the way on the right.
The SAC revised its scale this year. T6 now can contain long and T5 can include short sections of unroped UIAA II scrambling. Under the OSM sac_scale in my reading UIAA II sections are exclusive to T6.
The definition of UIAA II is breathing its age. It is, when climbers start to use three points of contact and beginners should be roped in. It is more about safety than difficulty. In my scramble proposal I put the caption “use of hands for comfort” below the T6 picture.
Never mind that modern routes cannot even be made with three points of contact, as that will not let you advance in places. A leap of faith being the extreme, no points of contact. In the scramble proposal I posted a video of someone doing this free solo: https://www.youtube.com/watch?v=8e0yXMa708Y
So, I’d wager, not an English native speaker, foot_scale=no in analogy to trail_visibility=no, there should be others, as value F6 seems fine to indicate Here be Dragons, as long as the documentation clearly states, that this is only for paths across non walking terrain even with hands like in F5. I am against calling this value climbing. The climbing community has its own tagging that is not based on highway=path. No need to interfere.
Yet, path is so broad in its meaning, we just cannot map all of it into a single scale. This brings up another problem, that @osmuser63783 pointed out: People want to add value to the data, i.e. tag stuff. Unfortunately, there cannot be a default value for foot_scale, i.e. casual_walking? As much as I’d like that, and certainly, most of what is mapped path is casual_walking – the outliers prove us wrong. Never mind that consumers for ever used that default.
UPDATE: One more coincidence for value NO; People mapping new paths, foot_scale=no – much like trail_visibility=no – might be a hint to re-consider mapping a path at all?
Curiously, it is the other way around: The Swiss do not map T6 paths, in case you meant Swisstopo I remember having read that the SAC persuaded them to map T4 routes some ten years ago. In my vicinity, Austria, there are a few routes containing UIAA II (according to local mountain guides assessment) in the national map, e.g Rumerspitze Westgrat, that is shown as a marked trail, which it isn’t (Obviously from below videos, memory did not serve me well.)
I might be misreading this, so apologies if so, but:
The discussion started off in 2022 with the suggestion that paths unsuitable for regular walking could be tagged as highway=scramble, which would mean that unthinking data consumers (like most of them) would not mislead people into walking off mountains.
This proposal has now been dropped in favour of adding an additional tag to existing paths, such as foot_scale:scramble:grade:sac_scale=t5.
Do you have a estimated date for when Mapbox, Facebook, Strava etc. are likely to start parsing this tag?
I didn’t think this discussion here was really a continuation of the highway=scramble discussion: It was never about moving some objects that are currently tagged as paths out of path and into something else.
As I understand it, the rationale for highway=scramble is that data consumers see highway=path and don’t know if it’s a smooth cycle track, a scramble, or a mountaineering route. A version of the problem you described in this ten year old (!) diary entry.
In contrast, the idea for foot_scale started as a side remark (bottom of this post) that sac_scale doesn’t allow us to differentiate between technique and exposure, so (some) mappers found it hard to apply consistently. It also mixes in other things such as slope and footwear requirements. So foot_scale intentionally excludes all that, and just focuses on technique.
Other complaints are that
The range of paths covered by T1 is too wide, so foot_scale has more values at the lower end
Paths are tagged as hiking (T1) solely because they aren’t in the mountains and the next level up (T2) is called mountain_hiking; also, paths are tagged as hiking to indicate that they are “hiking trails”, you go there in your spare time, for fun, ignoring the difficulty of the path. We’ve been debating the names of the values to avoid that
The root cause of all this is probably that the scale was invented in Switzerland, for the sort of paths and terrain encountered in Switzerland, and never intended to cover paths across the globe.
@Reza_Babri you have been mapping in Iran. Have you encountered any difficulties when applying sac_scale?
As far as I understand, foot_scale is intended to be more fine-grained and more globally applicable than sac_scale. But it’s not going to solve the path problem: data consumers that ignore sac_scale are unlikely to support it. I hope one day we get highway=scramble or similar, I just didn’t think this was it.
As a aside, I don’t see why we should replace sac scale with foot_scale:scramble:grade:sac_scale. It’s just the same thing with a more complicated key name.
I do hope, you are reading this wrong 
I at least have no interest in creating the one and only scale that encompasses any imaginable scale on earth as a subclass behind five colons. I alse have not the least interest in replacing sac_scale. Going this way, any proposal is doomed to fail.
sac_scale has its uses in niches. But it will never be a key that meets the needs of a wider class of map data users. Something that openstreetmap needs, and sac_scale needs that too, so it does not get misapplied due to factors already mentioned.
One of the naysayers argument in the scramble vote was, if the national map has a path there, openstreetmap must have a path there too. Here two videos of where our national map has a path.
First one https://www.youtube.com/watch?v=X0O0lWsajN0 starting the trip from Hungerburg 
UIAA II foto at 6:45 and a bit of UIAA I at 6:30. Not much impressive, but this is canonical grading.
Second one uncut https://www.youtube.com/watch?v=v0QqhTJI5OA : As smart people do, up the eastern ridge in order to end the trip with a scree run, video linked some post above. At 29:00 on the summit. Then down the western arete with the short UIAA II sections in between mostly UIAA I and some walking terrain where even a path emerges at times.
UIAA II section not much exposed, so not really challenging. Route in few places exposed a bit. We here say, Es pfeift - It gets airy. But not too much and not in the difficult places. Tagged sac_scale T5 (E) and T6 (W) in OSM. On site, the Western side is called an “Alpine Route” - decidedly telling guidepost readers that there is not path there.
foot_scale=no, in my opinion. highway=scramble in my opinion.
Regarding applications of foot_scale for cycle routing:
Does that seem plausible?
UPDATE: A day and some thinking later, I am not certain at all, foot_scale needs a fourth value, e.g. “assisted walking” - as that might encompass e.g this, https://i.ds.at/CwHRdQ/c:3808:2537:fp:0.500:0.500/rs:fill:1600:0/plain/lido-images/2023/11/20/7a872ce5-0fa4-4b33-a3e2-527bd5111ba1.jpeg - a location that @osmuser63783 mentioned; Note that the caravan is following a fixed rope. The terrain there certainly is no more demanding than Rumer Spitz, but some extra hazard=suffocation|freezetodeath or obstacle=lotsofpreparation|bigfinancialsupportnecessary might apply only there.
foot_scale=no and highway=mountaineering in my opinion.
@erutan @ezekielf @osmuser63783 - I created RfC: New Key foot_scale=* to start formalities. It is all about walking, thats what feet are good at, a pedestrian scale, so to say.
It allows for YDS, BMC, etc as well - agreed it would make no sense if it was only for SAC. It doesn’t have to be in foot_scale, but it’s the exact same syntax that climbing uses to allow for local scales/ Climbing - OpenStreetMap Wiki
I think it’s useful since we are collapsing a decent amount of difficulty into the highest value. Probably 0.01% of people will care about it, which is fine, but for that 0.01% knowing if foot_scale=scramble is T4 or T6, BMC 1 or 3, YDS 3 or 4 is probably helpful.
We could theoretically put the lower grades into climbing (YDS3, SAC T4, BMC 1) but IMO the scale is meant for technical climbing and it will be rendered as such. We could also just create a scramble_grade:yds/bmc/sac=* as well, but keeping it in foot_scale seems simpler.
It’s a bit of a secondary concern, but we need somewhere to stick scramble ratings since they’re in between most foot traffic and proper climbing.
Experienced MTB people will be able to bike down a lot of surefooted terrain, but overall I think that maps well.
I wasn’t under the impression that Facebook had to support tags for them to be implemented on OSM. Scramble ratings as described above are not the main point of this thread.
I also don’t think that rendering support for subvalues really matters. For 99.99% of people just seeing foot_scale=scramble is good enough. What that actually means in YDS, BMC, or SAC is useful metadata for minority of people, multiple mapping clients let you see raw OSM values for a way or node.
So if this tag is added to OSM do you see sac_scale still existing globally as the only way to describe T4 to T6 terrain? I think that’s sub-par for people in the US, UK, and other places that are more familiar with other systems.
Yes. In your words: Local systems rule, where walking ends. Their benchmarks are too much different. Incommensurable at that.
Scales for routes where individual research is a good idea anyhow and years of practice a highly suggested ingredient never can be condensed in a numeric value teachable from an OSM editor pane.
I guess I will have create another foot_scale RfC, looks like I forgot to ask contributors to stay concise in their expression.
For the avoidance of doubt, when you see a “sac_scale” key in OSM, this is NOT the relevant documentation. This is.
So wouldn’t that be a no to keeping sac_scale as a global default for terrain that falls into T4-T6?
What do you propose for values in that range?
That still doesn’t make the distinction between T4 to T6 terrain as intuitive to me as YDS 3 to 4, or BMC 1-3 to the average scrambler from the UK.
I’m unsure of why technical climbing defaults to local scales, but scrambling shouldn’t. Someone in the Swiss Alps doesn’t have to tag a route as YDS 5.10a, just as someone in the Sierra Nevada wouldn’t have to tag an equally challenging route as UIAA 6−.
Perhaps. In theory at least. Living in an area with lots of incoming tourists: From talks I learned, some of them heavily rely heavily on e.g. hikr.org, a third party, that mentions the SAC™ scale a lot in tour descriptions. Never got told about what the Wiki says.
… which is of course fine, because it’s not referring to to OSM data!
Yes, it only ends up there. Regardless what the Wiki writes.
In the End, it wont matter much, gradings should remain within margin of error.
It’s unfortunate that the climbing page in the Wiki says that the YDS scale is to be tagged using climbing:grade:yds=*, if the YDS scale also covers walking and scrambling. But that’s from a proposal from 2008 and it has been barely used since. As a solution, instead of inventing scrambling:grade:yds=* for the scrambling values of the same scale, why not put everything into yds_scale?
foot_scale can become the main tag that we encourage everyone to use, and it is designed to be as globally applicable as possible. Paths in the Alps can additionally get sac_scale especially if one wants to make the distinction between T4-T6. Paths in the Sierra Nevada can get yds_scale etc. No need to invent more complicated names
My assumption is that it’s only for YDS => 5 and using the climbing tag for YDS 1-4 wouldn’t be appropriate. It’s been done here by people to indicate scrambling, but I feel like renderers are expecting climbing to be for technical climbing.
That was an option I proposed on the foot scale thread yesterday.
I don’t see any problem having yds/bmc/sac_scale exist alongside a foot_scale or even a climbing tag (as SAC overlaps into both along with YDS). If climbing:grade:* isn’t used much then that’d be a good reason to just simplify things by adding more high level scales. Renderers could then do some simple if/then logic to prioritize whatever one they wanted and then have fallbacks.
