“highway=path”: “if a way is tagged highway=footway and mtbiking is
possible change to highway=path foot=designated”
In this specific case, I completed the previously incomplete mapping
of some paths. I added the new paths, and changed the tagging on the
existing ones to match the Wiki as well as possible. The paths in
question are most similar to the photo on the “Tags” Wiki page under
“highway=path” but about 0.5m width and are non-official but are very
well used for both mountainbikers and hikers (about 50/50 usage).
So I followed the Wiki advice and changed some tags from
“highway=footway” to “highway=path”. (I did not add “foot=designated”
as they are not designated for pedestrians.) Shortly after making this
change, another used changed the tags back to “highway=footway”.
It would appear that the Wiki is not clear on the tagging of such
paths. Can someone clarify? (I’m happy to add appropriate
clarifications to the Wiki, but only after I know the general
consensus on this.)
(May I also ask what the appropriate resolution would be in the case
where there is a difference of opinion about such tagging on
particular paths? It is important as the paths are rendered
differently in this case, at least by the renderers I have looked at.)
If I may dare to summarize the information contained in the thread you mentioned: Opinions differ, but the majority opinion and that of the Wiki pages is that “highway=footway” should only be used for a way that is designated for pedestrians. (The exact meaning of “designated for pedestrians” depends upon the country.)
The problem (as I see it) for most users is if the same way may be rendered very differently depending upon who tagged it, based on their personal opinion. For example, that OCM does not render “highway=path”, although this tag implies general use including bikes; yet it renders “highway=footway” although this tag implies (unless otherwise specified) that bikes are not allowed. This significantly limits the use of OCM for cyclists who want to ride on narrow tracks in the forest, equally for hikers who want to walk on such tracks.
I’d rather not see this thread as just replicating the arguments of the thread Alex quoted. But I believe it is very important to many users for there to be a clear definition of how to tag narrow, unpaved, undesignated, high-usage ways and have these rendered in a sensible way by the commonly used rendering tools. How can this be done?
I also agree that highway=path should be use in such case (just a path, not built or designated for pedestrians). And I also don’t like OCM not rendering that. That is one of the reasons I have set up own renderer for cycle maps of the local area.
I think you have well read and understood the wiki. I personaly do tag 95% of hiking ways as path because moutainbikers are not more nor less in the right to ride on those I tag. Nor they are “exclusively” for pedestrians. (wiki definition of footway)
But there have been (and still is !) terrible battle against dark and bright forces (don’t know who is who) :
Some extremists do refuse that path exists (open cycle map’s admins are some of them)
Some do insists that footway/cycleway be deleted
The problem we face is the lack of alternatives to open cycle map, the force is unbalanced, let’s hope a Jedi will return equilibrium in the galaxy…
PS: However I must admit I understand their point, not at refusing path, but at refusing to render it because what they want is a map for cyclist, not for moutainbikers. If people start to tag cycleway as path, distinction between a mtb trail and perfectly smooth cycleway will become problematic.
A continually changing playing field is always going to be a problem with OSM though. We need to turn it from a problem into a strength. Due to the nature of the data it’s relatively easy to see how tags are being used. That allows for at least a semi-automatic way of adding rules to render different types of tagging.
Certainly though I think that more services offering competing maps will help to rectify the problem.
So according to current guidelines, a narrow, unpaved, unclassified way through a forest is a “highway=path”. (“Current guidelines” are always open to debate and changes, but I did not want this thread to be the place for this debate.)
The remaining problems:
OCM does not render paths: None of the Wiki examples of path show them suitable for road bikes, so there should be no confusion to road bikers if paths are rendered in a similar way to footways. I’ll write to Andy of OCM about this. Can anyone point us to a place where Andy has stated his reason for not rendering path?
If one correctly tags a way with path, but another user incorrectly (according to current guidelines) changes the tag to footway, obviously the first step is to contact the user. What does one do if the user refuses to accept this and changes it back again? (The user will have a valid reason, such as wanting to see their favourite mountainbiking or hiking track on OCM on their iPhone. But his footway tag will prevent anyone’s renderer from distinguishing between somewhere bikes are banned or normally unsuitable and somewhere not.)
On the same page, the fifth example is (and several others are) “suitable for road bikes”.
Since the deprecation of footway and cycleway was not “approved in vote”, a footway that looks like a footway and is used like a footway is still a footway, except apparently in (at least) Germany where a footway can’t be a footway unless it’s signed with the blue pedestrian sign.
In my opinion a footway should be a way mainly for “pedestrians”, and I exclude people hiking or walking from the definition of pedestrian. I a footway allows cyclists, but they can’t go fast because to many pedestrians or obstacles, then I’m in favour of leaving it a footway. If however a way is suitable for cycling and pedestrians it should be a path. OCM rendering is a major hinderance to proper tagging in my opinion. They should really render path.
I think on the other hand we really would need a bicycle suitability tag, just like we have mtb:scale. This tag should be used to identify who can use such a way (and also make it clear if a way is so crowded you can’t cycle fast, or if a street is really unpleasant to cycle along because cars go very fast and don’t expect cyclist (even though cycling is allowed) and similar cases. If OCM doesn’t wan’t difficult pathes, whell then they should simply exclude any path that has sac_scale or mtb:scale attached to it, or tracktype=grade3 and higher.
Having a good key that defines how nice a way is to cycle along (yes I know we are more or less able to define it using a myriad of keys, but first this is really difficult to remember, and second if a renderer needs to check 10 attributes, and then decide how nice it is to cycle along the renderer will make many mistakes in classifying. Better to have a specialised cyclability key therefore) on the other hand would make the footway/cycleway/path disaster much less significant - even though I would still prefer to see multiuse ways tagged as path. Especially in the mountains I see no reason at all for a way to be tagged as footway. What about people with mules? They definitely can go nearly anything tagged footway in the mountains right now, but they would not wan’t to use a footway in a city with the mule trotting along. Similar a mountainbiker would like all pathes in the mountains, but not a footway in a city!
Quite right. I just suggest that, if OCM renders footway, it should render path, and one possible rendering is for it to be the same as footway. (The rendering could be made more complicated if you want to differentiate snowmobile paths or whatever else.)
The paths I am personally worried about are in Germany. Unmarked forest paths in the area I am looking at are being changed from path to footway, against the Wiki guidelines and general opinion in this forum (not just this topic, and granted, there are some differing opinions.)
Most people agree with you. But how do you deal with the situation where correctly tagged ways you’ve entered are a few days later changed to have incorrect tags? And how do you explain to a user of OCM that some forest paths are rendered, and some not, though they are physically and legally identical and in the same area?
It looks like there won’t be one accepted opinion about meaning of “path”. Could it be better to go forward and start defining more accurate alternatives like footpath, cyclepath, forestpath, citypath?
I like that idea, but I’m afraid we’ll run out of names one day
How about having a central point to store and enumerate proposals – somehow like the RFC collection in the internet. We could have a reserved tag “proposal” optional in any data base entry, which tells the reader and changer of the data, according to which proposal number the tags are to be interpreted. If someone changing the data does not want to handle the data according to this proposal, he/she shall remove the proposal tag. It would make translation to future proposals easier and it would not stop anyone from introducing new ways of expressing things. We should however require, that the contents of the proposal is clearly specified, such that the contents can be automatically be transformed to other compatible proposals as far as possible.
As values for “highway”? I don’t think these alternatives leave enough space for partial information. They would be appropriate for me, when I go to do OSMing. But we should be able to enter the information into the map, if someone just contributes a track of the last bicycle tour of his bowling group.
I’ve been somewhat confused by this whole issue. My own opinion is that highway=path makes sense for the unsigned paths through the forest. Aside from other arguments, the red dots rendered over forest look ugly. Paths between roads in urban areas tend to be well signed but in general both bikes and pedestrians are designated as allowed. I’ve been using footway if it’s narrow (~2 metres) otherwise cycleway (with appropriate foot=yes, etc).
In my area, track seems to be used for some things I can barely get my bike down. I don’t think it should be used for anything you wouldn’t be willing to drive your own 4WD car down.
Some things are difficult. For paths where rocks or tree roots make it near impossible to travel on a bike or even a horse, it almost seems that footway makes more sense. I still think they should be labeled has path, just with bicycle=no. Then there are official signed footpaths through national parks or country parks which on the ground look no different to the random paths in the local woos.
Another difficulty I’ve found it with paths that aren’t consistent. One nearby track starts out tarmac and very gradually reduces via gravel to unmade path to climbing over fallen trees following a still clear path to being completely nonexistent. There’s also old tracks that are essentially just firebreaks now but occasionally the forestry people come along and use them for a few months and they start to look like tracks again.
This is part of the problem; that a lot of people do things according to their opinion, instead of how the guidelines say. This leads to difficulties in accurate use of the data due to fundamental inconsistencies.
Why did you leave out the last part of the definitions?
“highway=cycleway: For designated cycleways ; i.e., mainly/exclusively for bicycles. If pedestrians are banned, add foot=no.”
“highway=footway: For designated footpaths; i.e., mainly/exclusively for pedestrians.”
Word “designated” in that is a remnant of the original definitions from 2006 (still valid!) and not used in the same meaning as in access=designated. A way “mainly or exclusively” for pedestrians is still a footway; changing that was rejected when highway=path was introduced. Likewise a way “mainly or exclusively for bicycles” is a cycleway even if it can be tagged as highway=path;bicycle=designated;foot=no/yes (or even foot=designated for the signposted combined-foot-and-cycling-ways).
The sign doesn’t matter for the users but can be recorded and emphasized by foot/bicycle=designated. If the signs forbid other expected users, they should then be tagged as foot=no or similar. Only when the values footway or cycleway are “inappropriate or inadequate” ( http://wiki.openstreetmap.org/wiki/Tag:highway=path ), you would need to use path.
Was just a short quote. I gave the link so you could read the rest of the page.
I agree with your post. My point to Oliver Kiddle was to follow the tagging guidelines; not make stuff up (such as footway<2m, cycleway>2m), but use footway for footways, cycleway for cycleways, and path for something that’s non-specific.
should highway=cycleway / footway considered as being obsolete because
it can replaced by path + foot/bicycle=designated or not?
I would prefer the “old” style. Then I can see just by checking the
highway tag if a way is preferred/designated for pedestrians or
cyclers without checking all the possible combinations (foot=… / bicycle=… ).
In the editors (JOSM/potlatch) I can also see the kind of way just by looking
at the color of the way.
For shared-use-ways I can see no problem in tagging it as hw=cycleway; foot=designated
(or the other way round, if the way is mostly used by pedestrians).
So in my opinion path should be used only for all kind of trails : small paths in the forest,
mountainbike trails and so on.