Since its inception, the highway=path documentation shows a grade 5 forestry track in the abstract. But the examples section starts with pictures of traffic signs as known from municipally maintained infrastructure in countries signed to the Vienna Convention traffic code.
This is giving current and potential consumers of openstreetmap data a wrong impression of what highway=path mostly is mapping: Two numbers from the ohsome dashboard (please share how this can be learned from overpass.)
There are 5 million km of path mapped as of today. A mere 150 thousand km, that is 3%, are both foot=designated and bicycle=designated, i.e. so called shared-use paths as depicted on top of the samples section.
I propose to move those three rows down to the bottom of the examples table in order to make the documentation less confusing.
This seems like a reasonable change. Adding clarification that these are signs in use would be helpful, particularly for those of us that don’t live in those countries. They do look a bit like diagrams.
Also, the example listed at the end that is not part of the table should be incorporated into the table for uniformity.
I disagree with that change. highway=path is globally used for such ways. Just because Austria use it in a different way, the wiki should not change. You might want to add to those examples that it’s not to be used in Austria.
Vieleicht sollte man lieber das Eingangsbild durch eine Montage ersetzen, die das Spektrum von highway=path deutlicher macht. Ich halte nichts von dem Versuch die Definition dieses Tags schleichend zu verändern.
I prefer a systematic approach to show the spectrum.
If this system or similar is implemented, actual usage may be quite different from current practice; only then it would probably be practical to order the list according to actual usage, with outlyers at the end. Personally I think the shared_use path will move to the top of the list, right after the path with no extra tags.
The numbers say 3% of highway=path is used to map such ways. And it is not clear from the numbers, that all of those are indeed ways sporting the Vienna Convention traffic signs. The Vienna Convention has a number of signatories. but there are prominent other traffic codes too.
Why should the examples section start with something, that is clearly insignificant (3% of global use of path), i.e. ways signposted with any of these three blue rounds and another one wholly peculiar to Germany?
Because it’s still a lot and is tagged all over the world? That’s not insignificant in my book.
Ordering by current usage is also an option, however, should path=… be widely adopted, I couldn’t tell to which subcategory al the bare highway=path’s will go.
The traffic sign images were introduced in 2014. Together with the Hexenstieg picture. I’d say, this picture matches a much higher percentage of what is mapped “path” than 3% of signed with some signs. Even sac_scale, which is clearly a niche tag has a better ratio. Why should the Hexenstieg picture not be on top? As it shows a shared use path, the path, not some signs.
BTW: For traffic signs, there is the “traffic_sign” key, that can be applied to ways too. The examples section now makes one think, that highway=path was specifically created to map a certain traffic sign
Did you know? P.A.T.H. stands for “pedestrian and tricyclehighway”. According to the unwritten rules of Vienna Convention signs, the tricycle is shown in profile, so you don’t see the third wheel.
It is important to me that the wiki continues to make it very clear that highway=path is suitable for all types of multi-purpose paths that are not used by multi-lane vehicles.
The picture is just an example and probably fits just as well or badly as the cover picture (which probably shows a forest track)
I don’t see the resorting the pictures really improves the wiki article. But its o.k. if I’m outvoted
to make the documentation even more confusing - in my opinion.
The only thing that would make sense would be moving the forth row with the explicit tags at the first position - and leave the example(s) without additional tags at the last position.
If it is a track - and that’s really possible, then rather a grade2 one. But not a grade5 for sure.
The “and therefore” part of that doesn’t follow from the “where you don’t” part of that.
In England and Wales, designated isn’t useful other than as a synonym for yes to describe legal access (something muddied by its use on permissive highway=path).
They’re not, because (a) almost no-one reads the wiki (they look at editor presets etc.) and (b) the examples are the wrong way round. Leaving aside legal access (which highway=path just can’t do), the decision on segregated is not connected to the one on surface, or the one on smoothness. The decision on mofa=yes is similarly unconnected to the others.
None of them follow from one traffic sign - they are independent decisions.
I’d be curious to learn, how the mappers in those areas are helped by getting presented icons of traffic signs that they do not have.
they could see that they don’t have the concept because they don’t recognize the sign (or assume that they have the concept but their sign looks different) and actually from experience if the concept isn’t there it doesn’t mean the tag won’t be used, e.g. people tag landuse=village_green in Germany or highway=living_street in Italy, which seems to confirm SomeoneElse:
I queried ohsome for the length of all paths and then that of paths marked with foot=designated and bicycle=designated. The share of designated was ~3%.
I will do just that. And
too.
So true, it hurts. Makes me want to just remove all four traffic sign rows. Cycleway has a picture one may want to use to replace them, it is in row “Combined cycle- and footway”. Any takers?
UPDATE: I guess, now I understand. The bare path, no attributes, should come first! That makes sense!
I don’t think that ordering the examples by the length of existing instances is necessarily the right approach. That assumes that everything presently in existance in OSM is tagged according to the intentions of the wiki, which is not true.
Another way to say the same thing perhaps, the references on that wiki page cite some valid criticisms of the highway=path tag, as it is commonly used without additional tags like surface=*.
I do think there’s scope to re-order the examples, and improve the images that accompany them.
I’m confused by Austria’s routing treatment of highway=path. The second example in the wiki highway=path,foot=designated,bicycle=designated,segregated=no says "Signposted foot and bicycle path. " is the most common type of “bike path” we have in my area. Very limited highway=cycleway here.
