To be fair, it is complicated dealing with combinations when there are a lot of them. A map style of mine has around 6 legal categories for “paths and tracks”, 2 width-derived values and 3 trail_visibility-derived ones also displayed, in addition to the familiar private, destination etc. overlays. On the page there are a limited number of variables to adjust (colour, pattern, width. etc.) - there aren’t an infinite number of variables available in the finished product.
Not enough to make it happen, though - that PR is still open! I’ve supported that in an OSM-Carto derived style since 2017
It “seems simpler” to those of use familiar with the foibles of OSM data, but some things simply make no sense (“what do you mean a closed way leisure=track can be either a ring or an area and there’s no way to tell which”?). Some kid out of school who has just been pointed at the MapBox documentation is going to struggle, and no-one who is allotted a small period of time to create “a map” for some random website is going to have time to understand the nuance of path.
We don’t have to continually make OSM data harder to consume. We don’t have to use the same tag value for urban cycleways that allow foot and bicycle access as for the Hillary Step. The fact that “some people have always done it that way” is not an excuse.
Which is the issue. Pathless is not a path yet it is a path. In a way it is an oxymoron, a paradox, something that is something it is not. No wonder there is no easy common name for it. It is very easy to say pathless sounds unnatural. But do we have anything better? I am not sure whether something being common is a good yardstick. Something being easy to understood is more important. And is “pathless” something you and other Brtitish speakers do not understand?
It’s only unnatural because it’s being shoehorned into the highway=* key. Why does it have to be in that key? Some ferries are considered highways; is it a mistake to map them as route=ferry instead of highway=ferry?
If a niche routing profile needs to route over one of these inferences, could it not recognize a second key besides highway=*?
I always lived under the impression, that English has 200 something terms, while German has only 150 something terms to describe the world.
Now, in German it is readily understood when a guide says: From the saddle go down the terrain until you reach the some and such path. There is no path from the saddle to the path, but you certainly will find your way. In German that would read “Weglos zum Weg absteigen”. “Weglos” (pathless) is kind of a promise, the promise though not that there is a path, but that the terrain is passable on foot. How can this be expressed in English then?
Yeah, a nice simple fix for a lot of things would be for routers to realise that you can drive into one side of a car park / walk into a public square / paddle your canoe along a river into a lake, & then continue out the other side!
I do not think it should go to highway=. I would be quite ok with it being path=pathless. As I said above, I think highway=path should be devided by subtags (with the eventual possibility of it being deprecated in some distatn future if the subtags get used a lot).
If one spends enough time gardening Wikidata items, one eventually realizes that it’s German and English that both have an inordinate number of names for things. But whereas English has too many names for the same concept, German has too many named concepts. I say “too many” rather ironically – neither language has as many names for things as OSM English! But the path to a translation in English is not necessarily straightforward. Sometimes it requires looking at something from a different vantage point.
In fairness, emergency=suction_point does sound like the kind of thing that an English-speaking firefighter would seek out in their line of work. I had to search high and low before finally discovering that it’s called a drafting site, fire site, filling site, or fill site in industry jargon. Rather than naming it after the equipment (a suction hose), English-speaking firefighters apparently name it after what they need to do (draw water, fill the tank, don’t care how) or why they need to do it (fire!).
In this thread, you might’ve noticed me equivocating about which tag to use, if any. I guess a layperson might be satisfied by something like a “pathless” or “indefinite” route segment, but it’s hard for me to say for sure because I’m not entirely sure we’re all on the same page about what it is we’re trying to tag. One thing I am certain about is that we’ll eventually have to name a preset, and preset names definitely are nouns or noun phrases, not adjectives on their own.
Ultimately it’s a spectrum with shades of gray, with the boundary between them impossible to separate definitively. If people regularly walk (or hike… or lightly scramble) across the ground, path is good enough for me.
Suppose you’re hiking through the woods on a single-track dirt path. You keep following the trail. Then it goes across a stretch of bare granite rock (this is common in my area), from cairn to cairn before returning back to single-track dirt path.
Are we saying that we shouldn’t tag =path for the bit where the trail goes over the rock?
That seems so outrageously pedantic, that I don’t even know where to begin to respond. highway=path + surface=bare_rock or whatever. A path is a place where people walk over, let’s not make things complicated.
The route goes over a class 4 climb? highway=path. After all, you’re putting your feet on the ground, right?
The world itself is a bit messy, and making a map of it too. OSM is lowering the barrier for the cartography process in having the data open, but but can’t solve everything.
Not to say we don’t have to deal with our shortcomings and try to simplify the tagging model, but taking into account already existing secondary tags into a few rendering classes on a general purpose map does not happen often enough, not even on our own. There is no need to display each and every nuance.
I think not being able to show the way to do it with documented rendering examples is a major issue.
Roadbook is from point to point, when they go from the track/path then it is off_piste off_track, then always a C cap is mentioned the bearing and the length, when it is not a straight line it is C MOY moyen average bearing. The next point is drawn as a tulip, giving hint what to expect.
Could this be a relation with a role from point to point?
Pathless is this the same as off_piste, off_path, off_track?
Knowing that you must walk, ride, drive on a bearing.
I miss the bearing aspect in this kind off route.
See a roadbook lexicon how they express all kinds of situations.
Roadbook, osm location, Bolivia, not much of these sandy roads are drawn in.
This is a dead end track now (64), this goes over in a off_piste, off_track route, a more dune vegatation section with a CAP C292 degree.
To the next point is 0,44 kilometer then you must go CAP C14 degree.
Dashed line is off_piste off_track, this was made for single_tracked en double_tracked.
The corosion and the water problems in that part of Bolovia.
To get a idea of how to navigate through a terrain with a roadbook.
I was asking myself, how should a off_piste relation look like in OSM.
And can we learn from this type of navigation.
It’s just a bit south of the nearby, more aptly named 
