Yeah, quite a bit of overlap in fact. It seems to me to be very difficult in discerning a difference between a karrenweg/multi_use and shared_use. The only thing I can parse is that the first may admit motorcycles but the other categorically doesnât? Wouldnât this work better with existing access-tags?
This seems troublesome concerning a point made in (the current version of) the proposal under âKarrenwegâ requirements:
Should not overlap with shared_use, highway=cycleway, mtb_single_track.
as it indeed seems to overlap with shared_useandhighway=cycleway (as it is currently being used).
Sure, but mixing physical characteristics and function or legal designations (as you put it above) in the definitions of the new tags invites huge overlap and edit wars. Except for the truly specialized, and thus very seldomly used categories like mtb_downhill and all the mountaineering stuff, where the function is explicit.
There are, for example, cycleways (with the traffic sign you posted, designating the way as useable for foot traffic and bicycles) that either admit or forbid motorcycles/mopeds (with an extra traffic sign). It would seem strange to me to change the path= category based on this. But I may just be weird.
Iâd say that defining the different path-categories based solely (or almost solely) on physical characteristics would be an easier way forward (thus lumping together the 20 parameters you mentioned into more usable, few, clumps or categories). Assume that no cars can drive on path-categorized ways (because we have highway=track for those). Suggest that people use cycleway and footway whenever possible (for municipally maintained and traffic-signed ways) After that, just use access-tags to describe legalities for mopeds and bicycles on the path-ways (as is used today, as access-tagging can easily incorporate the different local defaults).
Ending on a positive note, I do see value in the kind of lumping together of physical criteria under new path= categories, and in the new tags for the truly specialized (few) ways. But this is just my 2Âą.
I think youâre grossly overstating it here. There will certainly be no more overlap than between tertiary and service roads.
As a matter of fact, in a few countries Iâve visited, Iâve never seen that shared_use sign so I think Karrenweg is more universally applicable.
If it was up to me, I would probably do it the way you describe later in the post.
A mapper adds a highway=path to the map. They may or may not use other tags.
Someone has been asked to create a map based on OSM data. They donât have much time, and donât know much about OSM tagging.
They want to include âplaces you can walk and places that you can cycleâ as well as âplaces you can driveâ, so they include highway=path on their map. Their mental picture of a path is perhaps a footpath through a park, or an urban cycleway.
Someone wants to use their general purpose map to go for a nice walk in the countryside, even though itâs not a âhikingâ map. Maybe its just an app on their phone, and they donât even know that many different sorts of maps are available.
They are very surprised when a suggested route home involves a technically difficult route (that they may not be able to recognise on the ground) for which they are ill-equipped.
Arguably, the blame lies in more than one place for the situation above. The person going for a ânice walkâ clearly did inadequate research. The person creating the map did too - but maybe they were directed by their employers to produce something quickly.
Iâd also suggest that the OSM mapper didnât choose the best tag, and (because of the lack of time and knowledge that the person making the map had) even including other tags such as sac_scale would not necessarily help.
Correct. And I donât see how the division above doesnât help. Paths are categorised as ways for traffic, just like roads. If all the roads were displayed the same on the map, of course the drive back home through the most direct route would take forever, instead of taking the motorway.
If someone decides to follow a Trace and ignore a Wide Path, thatâs their choice. Right now, they donât have that choice.
The proposal does not immediately cancel the path and replace it with an alternative because there is too much opposition and it is probably not practical to do so immediately. But it provides a path where a transition can be made and the usefulness of the new categories can be tested.
Because it has a bunch of superfluous content that doesnât solve a problem. The only âproblemâ clearly articulated here is the tagging of scrambles and climbs as a path.
And the easiest solution would be to use about 1-2 new highway tags for those âsuper difficult waysâ.
For all the things proposed here, we already have more detailed tags.
If you want to re-structure everything below highway=track you need to first define the âsort orderâ. For everything highway=track to highway=motorway itâs clearly importance. Though you canât âsortâ highway=footway, cycleway, bridleway by importance. The current proposed values are kind of random categories and even partly overlapping. Thatâs never going to work out.
In my understanding of the problem, you need something like path=maintained_way for all the maintained/paved ways regardless of usage. path=unofficial_way for easy usable short-cuts and other not maintained ways. path=trail for more difficult to use ways. They might have a mtb:scale or a sac_scale as well as path=scramble. Climbs I would not keep as path.
There is not much overlapping, itâs scaling kind of by difficulty and solves the problem of the super wide range.
I agree this might be the simplest solution. No need for new tags. The highway=path iD preset could be labeled as âOutdoor Trailâ and support any difficulty level.
Personally, Iâm not in favor of excluding âdifficultâ trails from highway=path, as difficulty is quite subjective. Specialized outdoor focused app could instead follow guidelines that highlight and prioritize highway=path with tags like sac_scale, mtb:scale, and others. Iâve noticed that Strava already does this, and itâs incredibly useful.
I was also surprised to see that the iD preset for shared-use âCycle and Foot Pathâ already uses highway=cycleway + bicycle=designated + foot=designated, which shows things are already moving in that direction:
Finally, this means that ways used exclusively or primarily by motorcycles in urban areas would need a different tag than highway=path, but that can be figured out separately.
it is not intended for vote yet, but for feedback. there are numerous parts where feedback is explicitly solicited (all those TBDs).
@Tolstoi21 âshared_useâ is akin to cycleways and footways: ways that are legally designated for the combination of these two modes, usually (always?) accompanied by a sign. So âshared_useâ is a very specific type of path meant and designated for cyclists and pedestrians equally, it is not default. Maybe a bit misleading name. But that is what these are called as per wikipedia (linked from the proposal).
Re âmtb_downhillâ: Is it always downhill? Are not some of these kind of a bit up-and down a bit two? Still the bikes go fast and one does not want to be there, but I am not sure if including downhill is a bit too narrow.
The other widely understood problem is that path is too wide in its meaning. That urban paved ways where buses could go if it were legal are also tagged as a âpathâ feels wrong to many people.
However, based on discussion, I am kind of leaning back to my original position that the distinction should be about built/unbuilt ways, as @aighes suggests.
The division from Karrenweg through Wide Path, Narrow Path, Trace, and Off-Trail is clearly sorted by importance and path quality.
There is a section âSpecial road typesâ, which is a bit hard to sort by importance. And it matches in meaning the Specialized/Specific section for paths, above.
Although, I would put more focus towards path being unbuilt, as the built ones already have appropriate highway=* tags.
Like I said, methinks we should strongly advice people to tag ways that have the blue roundel traffic signs with highway=cycleway (and foot+bicycle=designated & segregated=yes/no depending on the specific traffic sign) instead of =path. Wikipedia may call them whatever, but local laws usually have more stringent meanings and precise nomenclature for explicitly traffic-signed ways. @_MisterY said that they have never been to a country that has those signs, but here in the North (I can only speak for Finland, Sweden and Estonia), they are ubiquitous.
Speaking more generally of traffic-signed pathways, in Finland, we also have very many ways with only âmotor vehicles forbiddenâ traffic signs like this. Since the FtR laws here allow walking and bicycling almost everywhere, these form a widespread network very fittingly made for the highway=path tag (and, usually, would probably fit the path=maintained_way that @aighes proposed, or something along those lines). Iâve no idea how prevalent those traffic signs and such paths are around the world, though.
Interesting! In Finland, we tag roads with that traffic sign as =service (or =track) only if that sign has with it a further sign that explicitly permits service vehicles (âHuoltoajo sallittuâ in Finnish or âServicekörning tillĂ„tenâ in Swedish, as Finland is a bilingual country, roughly: âService vehicles permittedâ, here). In Finland, if the traffic sign does not have that permissive additional sign, driving along the road with any motor vehicle is forbidden. And there are lots of those around. That rules out =service and =track here. But your mielage might very well vary.
Is it? Take the promenades on the National Mall in Washington, DC. They are around 40ft/12m wide so a bus certainly could fit, but that is not one of the intended use cases. These are gravel, not paved, but I think this may be the sort of thing @supsup is describing. Despite the width, I think most mappers would agree that highway=service isnât right for the National Mall walkways.
For those cases, I think highway=footway plus surface tagging is well-suited. Or perhaps, highway=pedestrian. We certainly donât need yet another highway tag for these cases.
This preset has been around for years, but more recently it has been suppressed in favor of a highway=pathâbased preset in a few European countries:
The voices clamoring for those regional overrides were expressing a desire to align with long-held regional tagging conventions, which are being called into question here.
One such problem in Austria was solved by the decease of the mapper, who furiously fought for a UIAA VII climb mapped as a path, because of national importance.
As the author of the not approved scramble proposal: During RfC this was turned into a micro-mapping tool that I until now not wanted to revive. I am not that much into micromapping. Meanwhile we have highway=ladder and micro-mapping that will break routing and create gaps, which I was told is a no-go. I do not see any chance of getting something approved that has such consequences unless it comes dressed as a sheep.