I’m not planning to redefine highway=path
, but would like to move things that aren’t actually paths (for example scrambles, climbing-related ways, “pathless” paths in general) to their own tags. To me, they should’ve never been tagged as path anyway. So that what remains, I would then sub-tag with path=*
I think we have something in common – I also see paths everywhere when out rambling.
The originator of the video called it a route. Only somebody familiar with the area can tell if he or she followed what is mapped as a path there in openstreetmap.
Can you provide some examples of what you see as potential values for sub-tag path=*
?
I haven’t come to a long list yet, because I find it really hard to think of categories. The things that I would really like to see in these categories are:
shared-use
paths (pedestrians + bicycles only)mtb
pathsconnecting
paths[1]recreational
paths (everything up to sac_scale T1, but not higher)
All the other path types people have mentioned, like motorcycleway, I don’t know enough about to classify them at all. That’s why I think some brainstorming group to get this started would be a good thing. Also, I’m a coder, I spend 90% of my time thinking about the name of a function or variable, and only 10% actually writing code. I’m terrible at this.
And by this, I mean any path meant for getting from A to B, much like unclassified vs residential roads. These have to be in a condition that allows travel in a speed similar to a road of that country ↩︎
Mountain bike trails really shouldn’t use the same primary tag as regular (shared) cycleways. This is pretty much the same case as scrambles, just for cyclists. Breaking backwards compatibility is a feature here.
This is a very good observation. The OSM schema is a bit car-centric, and there’s significant opposition (for whatever reason) to treating anything pedestrian- or cycle-related in the same manner.
It’d be nice if we could create a good tag for combined footway/cycleway, since this concept exists around the world.
I would think that this depends. A lot of easier MTB trails I know share ways with hiking trails. But it’s a gray area.
In my jurisdiction, these shared paths are cycling infrastructure. That corresponds well with the common practice of using cycle for the highway tag.
I had read that as MTB downhill trails that are single-track single-use.
This will get even more pronounced in my area, as tourism agencies are in need of more summer entertainment and exactly this kind of shared-use is coming up in talks right now. Opening all forestry tracks for MTB though is no longer actively pushed by any major player.
I see that under cycling and rules for cyclists, it’s considered one of many types of cycling infrastructure, which of course it is, but does that mean it’s regarded as being primarily cycling infrastructure? If so, that’s very interesting.
I can imagine a definition of types of pedestrian infrastructure would include pavements, footways, pedestrian crossings, etc., and could also potentially include foot- and cycleways. I couldn’t find any definition at all for pedestrian infrastructure in Austria.
Also, the NO tagging schema says to use highway=cycleway
+ foot=designated
for foot- and cycleways. Nothing intrinsically wrong with that, but it does place foot- and cycleways in the cycleway hierarchy. In NO, there’s no grounds to place it there, and some are arguing that recent court verdicts must be interpreted such that foot- and cycleways legally must be treated as footways…
The classification is literally from the legal code. On the shared spaces, cyclists have to yield to pedestrians though (they must not endanger them.) Mind you, cyclestreets are not cycling-infrastructure here
I doubt that shared spaces for hiking and mountainbiking will ever become cycling infrastructure. Will openstreetmap need two shared-path-tags?
That is my personal take on this. I mapped some such way (no segregation) even like that. It got corrected the next day.
I’m not quite sure if I would put it that way. Cyclists and pedestrians are treated equally in the definition.
11a.
Geh- und Radweg: ein für den Fußgänger- und Fahrradverkehr bestimmter und als solcher gekennzeichneter Weg;
“Foot- and cycleway: a way designated and marked for use by foot and cycle traffic.”
They are both pedestrian and cycle infrastructure at the same time.
That doesn’t make highway=cycleway
+ foot=designated
any less valid though. This tagging combination is used more often than highway=path
+ foot=designated
+ bicycle=designated
in Austria.
highway=cycleway
+ foot=designated
: 12559
highway=footway
+ bicycle=designated
: 1060
highway=path
+ bicycle=designated
+ foot=designated
: 9817
I’ve used highway=footway
+ bicycle=designated
before when the ‘cycleway’ was nothing more than a blue bicycle sign on a normal sidewalk without any real improvements for cyclists. It was clear that the reason it was put there was mostly just for the benefit of drivers on the road so they wouldn’t be ‘bothered’ by the cyclists.
Off-topic to German-themed side-topic: depending on the situation, but a lot of people are afraid to share space with cars so are quite happy for a legalized sidewalk. I usually share the way with cars wanting to be faster but I question my life-choices a lot when somebody overtakes me with a side distance of 20 centimeters (the world would be a much better place if world was motorcar=no :-D).
On topic I think your tagging makes a lot of sense to differentiate purpose-build cyclke-infrastructure and pure legalization.
Regarding ambiguity - one of the ambiguities I observe, when is it a path and when is it a footway. I observed that on OSM when a local mapper retagged a footway as a path and made corresponding changeset comments. I observed that on this forum here, when people make the difference depend on the height or the form of the kerbs between a carriageway and a pavement.
In the survey below, most people see highway=path
as a pathway of unknown classification.
This means the only way to tell them apart from “known” classifications is by creating new highway tags. I don’t see how saying only a bare highway=path without additional/specific tags is considered unknown will work.
If highway=path
is supposed to function like highway=road
, then ideally, routers and renderers would just ignore them, and mappers would get a hint to reclassify them.
You only asked us how to tag “a pathway of unknown classification”, not whether highway=path
represents such a pathway exclusively. In other words, this is “a pathway of unspecified classification”: maybe we don’t know the classification, or maybe we know it’s unclassifiable, whether because of a gap in our tagging scheme, or out a desire for fairness among all modes of transportation, or because the path is too informal to split hairs about.
To some extent, this parallels the unsolved problem of highway=service
: is it a major service road that clearly doesn’t fit any service=*
subcategory, or is it a service road we just haven’t been able to subclassify yet? There have been multiple attempts to define new service=*
values, but they always seem to get hung up on a lack of words.
In the absence of new tags, the simple solution of choosing the best-fitting highway=*way
tag may suffice in many cases. It’s like how we tell people to tag a hybrid store or mixed land use area as the thing that best describes its “general character” instead of comprehensively ranking everything that fits.
Yes, in hindsight, this ambiguity stems from the original highway=path
proposal using the same tag for two different purposes at the same time, skunking it right out of the gate – or, seen another way, from using something as messy as access tags for iterative refinement of a primary tag. This was well before OSM data was ever used for automated routing on a global scale, back when navigation-related tagging discussions were dominated by a very geographically limited subset of the community. We’ve learned a lot since then, and a similar proposal probably wouldn’t fly anymore.
Many data consumers are designed for laypeople, so punitive mapper feedback could easily go unnoticed or result in frustration without any fixes. Shattering the pedestrian network outright isn’t really an option either. However, if a car routing profile does recognize highway=road
, it should heavily penalize it, using it only as a last resort and showing a warning to the user. Likewise, a pedestrian routing profile should probably avoid a bare highway=path
in favor of a nice highway=footway
running parallel in close proximity. This is how some routing profiles already behave, so I’m unclear on whether there’s actually a problem.
Isn’t the vast majority of ways tagged as highway=path
actually outdoor trails, which, according to the wiki, wouldn’t need reclassification?
Well, one argument could be that a ‘path’ (and thus a path
) does form a coherent top-level classification in itself. The main point is that further classifications of path
s would presumably depend mostly on the physical characteristics of said paths. Hence we would not necessarily need a (very) complex set of new top-level highway tags, since we could just add descriptive tags, like width=
, surface=
, smoothness=
, etc. to the existing path
s and thus give them a further classification directly.
I’d say that in this scheme a “bare” highway=path
without any other tags, precisely and fairly understandably classifies an “unknown” path. Routers absolutely should consider them nigh impassable and route around (or very heavily penalize) such “bare” paths.
Except that is highly unlikely to have been the intention of most mappers who have mapped highway=path
up to now. Any time I mapped highway=path, it was simply to map a path with no visible designation for a particular mode of transport (but clearly not for 4 wheel vehicles) - which I understood to be the approved approach to that kind of path. I fully intended it to be routable, just as much as highway=footway
. These days I would try to add a surface=
tag at a minimum, but I wasn’t always very aware of secondary tags so I probably added a lot of bare highway=path
in earlier days. It was never intended in any way as equivalent to highway=road
, and I don’t see how we can now act as if it was.
I agree with @Minh_Nguyen that it probably makes sense to penalise bare highway=path
relative to more clearly-defined nearby alternatives. But considering them impassable, or close to it, seems to be going too far.
I agree, I think you have expressed this the wrong way round here. The poll tells us that most respondents see a pathway of unknown classification as taggable as highway=path
, but that’s a very different thing from seeing highway=path
as generally being of unknown classification.

I’d say that in this scheme a “bare”
highway=path
without any other tags, precisely and fairly understandably classifies an “unknown” path. Routers absolutely should consider them nigh impassable and route around (or very heavily penalize) such “bare” paths.
If highway=path without any other tags were considered impassabe, I would have to re-tag many perfectly passable paths in all the hiking/foot routes I maintain in NL, which is only a fraction of all the routes. Few, if any, NL mappers will do that.

If highway=path without any other tags were considered impassabe, I would have to re-tag many perfectly passable paths in all the hiking/foot routes
indeed, highway=path is passable, or you should not tag it like this.