Since a continuous crossing may be marked with paint or not, and also may be signalized or not, automatically choosing a default crossing= value when the user specifies crossing:continuous=yes doesn’t really make sense. Probably better to add a continuous attribute to the existing presets.
2 Likes
Thanks all for your input!
I’m now creating a proposal at Proposal:Continuous crossings - OpenStreetMap Wiki
Let me know what you think. If the people here are happy with it I’ll move it to “Proposed” stage soon and send out announcements.
3 Likes
That’s indeed a very nice summary of the discussion, including some great example images 
My only beef is segregated=yes on the crossing node. That key isn’t defined on nodes.
Other than that, we should maybe define some defaults that come with crossing:continuous=yes, which are very likely
kerb=no
tactile_paving=no
crossing:island=no
But this might just be over here? If it’s country-specific, then maybe not 
Looking a little closer, I’m not sure that your example from Edinburgh would be considered a continuous sidewalk from a pedestrian point of view, because it has a kerb and tactile paving. At least in my jurisdiction, the absence of a kerb and tactile paving would be requirements to consider this a continuous sidewalk, but maybe the situation in Scotland is different. We should be very sure about this when giving examples, that’s all I’m trying to say.
In the UK, it’s common for tactile paving to be present. The type seen in the image has the meaning of “Hazard”, which I feel is fair because it is still a crossing, and there could be a long motor vehicle crossing.
It’s also common for the pedestrian way to dip down to the carriage way level. This should not happen, and I suggest it’s a cost issue, and at some point will be changed.
For each location & country I assume there will local peculiarities.
I didn’t know that. This usage is documented on the page for Key:crossing, but not on the page for Key:segregated. Anyway, I’ve removed it from the proposal.
If you are consuming OSM data and crossing:continuous=yes has been tagged, but these fields have not, then these are probably sensible assumptions to make in many parts of the world. I’d rather not make them part of the proposal, because I want the vote to be about whether people agree with introducing the new tag, not about whether they agree with the default assumptions. Researching how continuous crossing designs differ around the world would be a big task - and maybe one that, in future, will be aided by OSM data 
(Having said that I really cannot imagine a continuous crossing with an island 
Maybe someone will prove me wrong!)
1 Like
I’ve moved the proposal from “draft” to “proposed” status and created a new topic here: [RFC] Feature Proposal - Continuous crossings
1 Like
Great work with the proposal, @osmuser63783! I took the liberty to make a few copyediting changes to (hopefully) make the text a bit clearer. I took care to preserve all the information in the proposal with the changes I made, but please take a look and check if you agree with them.
1 Like
Great work in having this discussion and drafting the proposal. I have a photo I could contribute for this if helpful?
On the proposal page, it says the most likely distinguishing feature is a change of surface but I think it’s a kerb across the street and none across the path?
I think there’s a couple of loose ends it would be good to tie up here although I don’t think they need to be part of the proposal.
It’s not totally clear to me from existing wiki text why the examples with contrasting colour paving (even though the same from pedestrians viewpoint) don’t count as ‘uncontrolled’ or crossing:markings=surface? Is it because the markings aren’t ‘road’ markings? Yet we’re saying give way marks also don’t make it uncontrolled?
How should the streetComplete text be changed to help guide users that yes, this is still a crossing?
I think another example could be added to the yes, it’s a crossing, such as “the sidepath continues uninterrupted” but that’s maybe a bit wordy?
Also, I’ll just say that I wouldn’t have gone for ‘crossing:continuous=yes’ because I don’t think it’s clear without reading the documentation as it’s not the crossing that’s continuous but the path. However, I’m not sure I have a better suggestion.
At least in Greater London, tactile paving will usually be present for a crossing with a table before the junction. For a “Copenhagen crossing” continuous sidewalk, there will not be any tactile paving.
1 Like
Hazard type tactile paving is used in Devon, UK. Would be helpful if the UK had, or least practised a national rule on this.
1 Like
It would help, but I can’t see the current administration or the next one doing anything useful for pedestrians, lest it be seen as part of the war on cars.
Many of my local ones look like this:
Thanks! Is that a real street or a (shared) driveway? I’m asking because the issue of driveways came up and people agreed that in that case the node should just be left bare. For the proposal / Wiki page I looked for pictures where it’s very clear that it’s an actual street, so if we add the picture to the proposal the street type should be mentioned.
Good point. The proposal already mentions that there isn’t usually a kerb across the path (though I’ve seen flush kerbs). I’ll add that there will be a (lowered) kerb across the road.
When I looked for pictures, a rule of thumb I used was that it’s definitely crossing:markings=surface if there is a clear surface change for both the driver and the pedestrian. But the line between that tag and crossing=unmarked isn’t always so clear, see @Minh_Nguyen’s comments earlier.
I’ve already seen an example where someone encountered a continuous crossing and answered “No, but crossing is possible” (which puts crossing=informal on the node - not at all intended). If the proposal is successful I’ll suggest to the SC developers that they add an option like “The path just continues without interruption”.
1 Like
Oh, now we know they are “officially” called “blended crossings” in the UK. Interesting, thanks for the read!
It’s tagged as a residential highway and I think that’s fair. It’s named and has houses on either side, even though it has block paving. I can photo another very plain example
https://maps.app.goo.gl/uWB9n2ypwQakprmq9 which is like a link road to a residential road (so is also tagged residential) which had me wondering because I hadn’t tagged it as a crossing but then the StreetComplete quest had me wondering how best to handle it…
Yes, and if I didn’t know the tag it would apply I think that would have been my instinct.
Seems in New Zealand they are using that term to mean what we would call a crossing with a traffic calming table https://www.nzta.govt.nz/assets/resources/pedestrian-planning-guide/docs/pedestrian-planning-guide.pdf
It’s not the “official” term for continuous footways in the UK. It appears to be a term put forward in London. Don’t think there is specific official term, and there could be variance for the different countries that make up the UK.
1 Like
Correct, it’s what some London Borough Councils use, but the Waltham Forest page is useful as an illustrative reference for their appearance and position of road markings.
I can’t see the Department for Transport in England doing anything to standardise them any time soon, as it might be seen as part of “the war on cars”.
1 Like
I’ve been asked to clarify if the tag can be applies to nodes, ways or both.
This is a tricky one because there are different mapping styles for crossings. Some mappers put crossing tags (such as crossing:markings) on the highway=crossing node and the way only gets highway=footway, footway=crossing and typical way tags (e.g. surface) (example). Other mappers duplicate the crossing tags on the way. Yet other mappers do not tag the node at all (example way with untagged node).
Therefore I suggest the following:
The tag can be used on nodes (with
highway=crossing) or on ways (withfootway=crossing), but there should be no need to use it on both for a single crossing. When ahighway=crossingnode is present, it is considered sufficient to just tag the node withcrossing:continuous=*.
I appreciate that you’ve considered the issue of these different crossing representations, but I wonder if preferring the tag on the node only would add to the confusion. As it is, many mappers simply try to keep the tags consistent between the two elements. This is pretty easy to do in iD, for example by selecting both and clicking on each “Multiple values” label in the raw tag field. But mappers are already getting confused because of the rule that tactile_paving means something different on the crossing node versus the crossing way. This doesn’t give me much confidence about a rule that crossing:continuous should go on one but not the other.
Additionally, the practice of mapping the crossing as both a node and a way simultaneously is mainly for backwards compatibility with data consumers that aren’t capable of converting between these two representations. If the two representations aren’t equivalent, then this compatibility shim is less effective and we may not be able to deprecate it for even longer.
1 Like
I see your point and I’m inclined to agree. To be honest I am just trying to introduce a new crossing tag here, I’m not trying to settle the question about whether crossing tags, in general, should go on nodes or on ways. These are two separate issues that I don’t really want to conflate: A vote for or against the proposal shouldn’t be seen as a vote for or against putting crossing tags on nodes or ways.
The phrasing I suggested is the result of some discussion between me and a fellow mapper who suggested that my proposal “should take a clear stance on double tagging by discouraging it”… which I am reluctant to do for the reasons above!
2 Likes