Following an earlier discussion on the community forum I propose a new key, crossing:continuous, with two values, yes and no, where yes is for a situation like this:
The link to the proposal can be found here:
https://wiki.openstreetmap.org/wiki/Proposal:Continuous_crossings
Please let me know here, on the tagging mailing list, or on the Wiki talk page if you have any comments.
Could someone who is on the tagging mailing list please cross-post this request for comments? Thanks.
(I have tried to subscribe but it’s not working, I am not getting a confirmation email. That’s odd because I have been able to subscribe to the talk list with no issues!)
It’s funny that you opened up this discussion considered I did cross some continuous sidewalks myself yesterday (Google Earth).
I myself see no problems in the draft, especially because it differenciates between continuous sidewalks with traffic calming, continuous sidewalks without traffic calming and traffic calming without continuous sidewalks (and for completion, neither traffic calming nor a continuous sidewalk).
I like that you are working on this, sometimes I see people splitting the crossing highway=* up to change the surface tags, which is not handy for many reasons. However, I think that there are some things that seem a bit confusing to me.
My main problem with this proposal is that it is not always clear what highway=* crosses the other highway. Since that depends on the perspective of hierarchy. Let me explain what I mean:
You’re definition:
A footway, cycleway or similar continues uninterrupted across a road.
Assumes that cycleways are not roads. However, a cycleway can be a path, street, or a road.
I would say that in a situation where two roads/paths or streets meet that are of a different level we are talking about a crossing, and if they are the same level than it is a junction. The one that is of a lower level is crossing the higher-level highway=*.
So, for example here a highway=tertiary crosses a highway=cycleway, both are roads:
The surface of the lower classification (highway=tertiary) is not continuous so this would then be crossing:continuous=no?
Or would you say that a cycleway should be considered to always be of a lower classification than any vehicular road, so than this would be crossing:continuous=yes?
I would like it if you would think about how to make it clear what highway=* crosses another highway=* only then you can specify which part is continuous.
And rephrase some sentences to be clearer and more open. Like the definition.
I made some suggestions to about this here:
Why is highway=tertiary a lower classification than highway=cycleway?
I think the important factor is the speed on each road.
Road users of a highway=tertiary generally go faster than those on highway=cycleway
So when the faster road users are slowed down by moving up to the level of the slower road, we can speak of a continuous crossing.
I do agree that this should be specified in the proposal.
An issue with this is that the road might be reclassified at some point, but the tag is not (automatically) changed. Then the crossing ends up in an invalid state. The crossing:continuous can have the same problem - where it flips priority. However that requires the less important road to become more important than the other, which seems more rare.
The highway=cycleway connects traffic across the whole neighborhood and is therefore a secondary cycleway. While the vehicular road connects to only a small part of the neighborhood and is therefore a tertiary vehicular road.
The speed cannot be used to determine the hierarchy of a cycle way, since there are only two speeds for cycleways in the Netherlands.
The speed can be an indirect indicator of road hierarchy, but it is never the reason of a specific hierarchy, there are primary and secondary vehicular roads with speed limits of 30kmh or even 15kmh while many unclassified country roads have a speed limit of 60 or even 80kmh.
Thanks for your comments. The topic of junctions vs crossings came up in the other thread, have a look here (and the posts immediately before and after). Does this make sense to you?
It is fine if you do not want to include junctions. However, I still think that there is a paradox here. If you take a usual continuous sidewalk the street is crossing the sidewalk, and the surface of the street is not continuous so that would be crossing:continuous=no
But then if the sidewalk is not continuous the sidewalk is crossing the street and the crossing is also crossing:continuous=no.
You can say that highway=x is or is not continuous, but it depends on your own perspective if something is or is not crossing something else.
I would say that there is never a case where crossing:continuous=yes applies since if the cycleway/footway is continuous than the other highway is crossing the cycleway/footway.
Do you understand my problem?
I like the suggestion from @ezekielf that would indeed clarify the crossing better.
The definition of crossing:continuous=yes is:
“A footway, cycleway or similar continues uninterrupted across a road.”
If you take a usual continuous sidewalk you have a highway=footway and a road (e.g. residential). The footway continues uninterrupted across the road, so it’s crossing:continuous=yes.
The fact that the surface of the street is not continuous is only further evidence that you are looking at a continuous crossing.
Do you have a suggestion for how to rephrase the definition to make that clearer?
The current proposal is essentially this suggestion, but simplified to have just two values yes and no.
So I suppose in situations where crossing:continuous=yes isn’t clear you could still use, for example, crossing:continuous=cycleway, but for your average continuous sidewalk, crossing:continuous=yes should be enough, and I’m open to rephrasing the definition if this can be made clearer.
I would be happy if you add something in there that allows for crossing:continuous=[highway].
My main problem with the definition is that many cycleways are also roads. so, to me the definition right now reads as:
A road continues uninterrupted across a road.
I would rephrase the definition to:
Indicates what highway continues uninterrupted across another highway.
And then you can add the current definition to the definition for crossing:continuous=yes
And you might then change the “footway, cycleway or similar” to “path”. Since “or similar” is not clear.
Then you make it clear that your specifically referring to cycleways that are paths.
So, then you get something like:
crossing:continuous=yes indicates that a path continues uninterrupted over a road.
I see, thanks.
I was using the word “road” like it’s used elsewhere on the OSM wiki e.g. here, to mean anything from motorway to residential, plus service, living_street etc., including cyclestreet and bicycle_road but definitely not including cycleway.
I didn’t know that some cycleways could also be considered roads. How do you distinguish “cycleways that are paths” from “cycleways that are roads”? Do you still tag them cycleway? Do you still use highway=crossing when they meet a “motor vehicle road”?
Anyway, I can see how if you think of a cycleway as a type of road, it would then be unclear how to tag the crossing. I’ll take your suggestions into account when I next make edits to the proposal.
In the Netherlands we tag everything that has a cycleway sign as highway=cycleway, this includes everything from side paths along vehicular roads, cycleways through nature and bicycle highways.
One difference between a bicycle path and a bicycle road is that a bicycle road connects major destinations in a similar way to how a highway=secondary or highway=primary connects destinations.
It is at least 3 meters wide and is completely separated. They are generally named. And the emergency services can use the cycleway if required.
This is different from cycleways that are not for transportation, they don’t have these properties, and cycleways that run along a vehicular road would also be considered a path even if they are very wide.
A cycleway that is a street (not to be confused with a cyclestreet) is a cycleway that has all the characteristics of a street like lighting, sewer, internet, sidewalks, and houses next to it. But does not allow cars and has a cycleway sign.
Also, in the Netherlands a lot of mopeds, mofas and disability micro cars use the cycleway network, which further complicates things.
Yes because i would still want to add crossing:markings=dots+crossing=marked to the crossing.
And even if destination vehicular traffic is allowed like here:
motor_vehicle=private or motor_vehicle=destination gets added, but it is still a highway=cycleway
The trunk/primary/secondary/tertiary scheme is applied very inconsistently among regions for all sorts of reasons. Maybe a more intuitive analogy would be the difference between highway=footway and highway=pedestrian, except that for some reason we lack a cycling-oriented counterpart to highway=pedestrian?
Seems like it would be clearer to tag the bicycle roads as bicycle_road=yes so their importance can be tagged with the appropriate highway=primary/secondary/etc and highway=cycleway isn’t overloaded for both paths and roads. I suppose this is only tangentially related to continuous crossings though and I don’t mean to derail the thread.
I’ve made some edits based on your suggestions. It should hopefully be clearer now.
I’ve changed the definition to “A path continues uninterrupted across a road.” and added:
In some cases, it may not be obvious which of the two highways is the “path” and which one is the “road” (e.g. when a very wide
highway=cyclewaymeets ahighway=pedestrian). In such cases, the value of thehighway=*key can be put incrossing:continuous=*, to indicate which of the two highways continues without interruption, e.g.crossing:continuous=cycleway. However, the tag is only for pedestrian and cycle crossings. It is not intended for use on road intersections.
Please let me know if that makes sense.
I didn’t go as far as making the definition “a highway continues uninterrupted across another highway” because I was worried that might lead to people to map e.g. the junction of a residential and a tertiary with crossing: continuous=residential.
Are there any more comments, questions or concerns? If not we could take the proposal to a vote soon.
Thanks all for your comments. I’ve started the vote.
