I think the point is that there are at least some trails here in Europe that are (legally or organizationally) bike only, no (though I would expect that to count at most in the low thousands of km at the moment)? Whereas in the rest of the world, they pretty much non-existant. However, this is a conjecture, I do not know what the situation is.
For these, highway=path, leisure=track, foot=no seems the sensible way to go IMO.
For these, highway=path, leisure=track, foot=no seems the sensible way to go IMO.
IMHO this would be a significant change in application for leisure=track, which currently is defined as: “ A dedicated track for running, cycling and other non-motorised racing such as horses, greyhounds, typically not a part of the normal network of ways and paths.”
and represented with this picture:
https://upload.wikimedia.org/wikipedia/commons/thumb/d/d8/Skyttis_athletics_tracks.jpg/640px-Skyttis_athletics_tracks.jpg
4 Likes
I agree, I would expect any leasure=* entity to be managed and not part of a general-public network.
1 Like
BTW: I just checked the legal situation and legally, pedestrans are allowed there, I think. it is just discouraged, as it is somewhat dangerous. So foot=no would be (is) wrong from the legalistic definition that access should be used only for legal matters (which the wiki says and i think is widely disregarded based on my experience :-D). Probably even onewayness is not legally enforcible.
This is a special type of a path. The most straightforward (adn umabigious) way to make it would b esomeething like path)way=single_track-mtb. i still have not found a decisive argument agains. (though please note I am a hiker, not biker - but still as a hiker i want to know about these becasue in general I do NOT want* to be on them - there is a risk some well meaning cyclist might crash into me. in theory one is supposed to ride so that they can stop anytime, but we all know how well theories work when they are supposed to be upheld by mistaking humans).
*and using foot=no is not really helping because I do not always respect unjust (from my point of view) laws. i do not want to be killed for that.
2 Likes
Seems good enough. These are fairly specific subtypes of path and, yes, can be pretty dangerous. I’ve seen a few in the forest and, with vegetation, I would not want to spend too long there. Sometimes they cross the pedestrian paths and I don’t know how that is intended to work, considering the inclination just before the path.
… practice.
1 Like
I’d say, many parts of the world have their peculiarities. Curiously though, I’d say, in this case, Austria (being in Europe) is not that much special.
Most of what here gets called Mountainbike Route goes over forestry tracks; speaking of Tyrol, which is in Austria. I ride them with my trekking bike mostly fine. BTW: Forestry tracks account for more than half of all of the vehicular traffic infrastructure combined here. Part of those are allowed for cycling by special agreements. Those of course are mixed use.
I guess, what @Road_Runner meant are paths especially, not tracks: While walking the woods generally allowed by law anywhere apart a few narrowly termed exceptions, cycling in the woods generally prohibited by law. No mixed use on paths therefore by default.
To offer to the sportive crowd, tourist resorts and also some municipalities commission track builders to create MTB downhill races with all the contractual and legal whistles checked. Though legally not binding, these are designated as not for walking and one-way-only so-called MTB single-trails. People commonly abide by that “pleas”. Common sense dictates that. They are single use as a matter of fact.
1 Like
mtb-single-track sounds even better. @supsup, you can add it to your list of path (sub-)types.
So, where is the problem, some may ask?
These downhill races commonly get marked “highway=path; bicycle=designated; foot=no”. That makes them cycleways in the strictest sense, same class as those. The only difference being presence of an mtb:scale>0 tag.
All of the renderers available from the openstreetmap website will show them the same as bike paths. Some of the routers available on the openstreetmap website will treat them the same as bike paths.
Some other renderers even paint them as a white line with blue casing. Something that map readers accustomed to local map signatures will naturally interpret as a 2+m wide paved road. I asked the German map style producers to consider mtb:scale. They denied, what is mapped like a cycleway will get shown as a cycleway. This reminds me on the quest on the OSM-Carto issue tracker to show higher sac_scale paths differently. At least, OSM-Carto does not show paths as a white line with reddish casing. Some very prominent renderers do. So browsing my local area from them I see roads running up every other summit, where there are in fact difficult-alpine-hikes, in OSM terms. Shall I be fine and call them crap consumers? That might certainly make me feel more comfortable.
1 Like
Something that map readers accustomed to local map signatures will naturally interpret as a 2+m wide paved road. I asked the German map style producers to consider mtb:scale. They denied, what is mapped like a cycleway will get shown as a cycleway.
arguably something mapped as designated bicycle path with high mtb rating is not tagged as a „cycleway“, is it?
…and that is a difference, which should be sufficient.
If OSM is providing the information (mtb:scale= some high value), and a data consumer isn’t using that information, they are to blame, not OSM.
3 Likes
And that is where you are wrong.
1 Like
Kind of. OSM tagging documentation is very much “beware of the leopard” and I don’t think it’s realistic for data consumers to understand every single ever-changing nuance.
I probably take more notice of tagging intricacies for cycle.travel than anyone else does (except perhaps @SomeoneElse), and even I get caught out regularly because someone has decided to invent a new exciting way of tagging something that has been perfectly well mapped for the last 15 years. Most data consumers, even some very large ones with very large amounts of VC behind them, have a thought process that goes broadly “does that look like a map? yeah, great, it’ll do then”.
Ultimately, as mappers, we all want our mapping to be used. Making tagging simple, understandable, consistent, reliable and well-documented is a big part of making that happen.
4 Likes
My post in that thread can no longer be edited :-(.
I would recommend using the wiki for such documentation
2 Likes
Oh, ok. It is time, anyway, to summarize the discussions so far. Or, rather - to narrow down to one proposal that works for most cases.
I think the big majority agrees on path-subtypes, not necessarily the new highway tags. I guess something like path=* or pathway=* would work.
The values you had before are ok as the start for filtering them down. The more I think about the distinction “by width” during recent hikes, the more I like it.
I’ve listed the examples at Discussion about deleting the highway samples page - #18 by _MisterY. The post could use some polishing but contains the essence of what I find useful for hiking paths.
Perhaps we can have a summary post here and/or start a new discussion on what the sub-types should be. Most have (rightfully) agreed that the route to Manaslu is not a path. The rest are just shades.
For example, I still don’t see scramble as a separate entity. Scrambling routes are the lower part of the UIAA scale. I’ve been tagging many climbing crags and routes recently, so that they are visible on OpenClimbing, and noticed that there is all sorts of tagging applied to them. Some are paths, some are climbing routes. Scrambles can be identified by using (unfortunately) both of these - higher-SAC paths and lower UIAA grade climbing routes. Some people free solo 4+ paths, so that can be considered a scramble, too.
And so on. The point is that we should now move on to see which are obvious sub-types to start with.
And, yes, we should continue with wiki posts for the work-in-progress items. Before it is suggested as an RFC, we could polish the ideas down further in a wiki post.
1 Like
I agree on using the wiki. I think a new wiki page (a draft proposal, but not yet posted for RfC) is the way to go. And accompanying thread serving to discuss that wikipage.
I guess I can live with subdivision of trail into wide and narrow (maybe narrow_trail and wide_trail. I think the proposal should formally say that path requires a subtype and tools should complain when not filled. Traces can probably cover pathless and scramble. We can probably separate the discussion if traces should move out of highway=path until later. I will be travelling from December 9 until end of March, so I will only be able to contribute until then.
1 Like
I’ve just edited it; can you try again?
Historically the wiki hasn’t proved effective for that sort of thing - wiki editors have on a couple of occasions gone off on their own flights of fancy in ways that do not match OSM data. A wiki post here as part of the discussion is actually far better at communicating what’s happening.
1 Like
This is what I had in mind, too. No need to switch to the wiki site until we’re happy with the shape of the proposal and it covers the problematic cases and paves a way to handle others.
1 Like
i do not have the pencil icon anymore there.
i think discourse probably locks down editing for posts after some time (similar like you cannot unlikesomething in time).
if we think we are in the proposal-writing page, makes much more sense to use the wiki. We do not mean to start chaning random wiki pages. But I guess editing here can do too if somebody recreates the structure of a proposal.