Proposal talk:Continuous crossings

From OpenStreetMap Wiki
Jump to navigation Jump to search


Resolved: Bkil confirmed to me they have no further comments or questions. Osmuser63783 (talk) 08:27, 28 November 2023 (UTC)

Although it is much more rare over here, we usually add traffic_calming=hump or traffic_calming=table on the crossing instead of this. -Bkil (talk) 22:57, 9 November 2023 (UTC)

I do agree that most continuous sidewalks should also get a traffic_calming=* but the main justification is psychological: A pure traffic_calming=hump breaks the surface continuity (pavers to asphalt and then back, makes it feel like one is crossing a street) wheras a continuous sidewalk doesn't (pavers all the way through, makes one feel like one still is on a sidewalk), the latter which can otherwise only be properly implemented with highway=crossing ways (not nodes) which is overkill in most situations. There is a reason why counterexamples have been included. --ManuelB701 (talk) 07:05, 10 November 2023 (UTC)


Resolved: Bkil confirmed to me they have no further comments or questions. Osmuser63783 (talk) 08:27, 28 November 2023 (UTC)

We usually use kerb=no in such situations Bkil (talk) 08:48, 10 November 2023 (UTC)

I agree a continuous crossing very often does not have a kerb, but I have also seen examples with a flush kerb. (This is already in the proposal) Likewise, kerb=no may mean that a crossing is continuous, but there are also examples of crossings without a kerb that aren't continuous, for example in rural settings when there isn't a sidewalk (picture). Therefore, the two tags are orthogonal (neither tag implies the other). Osmuser63783 (talk) 23:29, 12 November 2023 (UTC)


Resolved: Bkil confirmed to me they have no further comments or questions. Osmuser63783 (talk) 08:27, 28 November 2023 (UTC)

surface=asphalt added to a crossing on a road also marked with this tag conveys the semantics of not being a difference in pavement. Bkil (talk) 08:49, 10 November 2023 (UTC)

That’s true, but the surface tag can only do so much work. In this French example both the footway and the road have surface=asphalt so the surface tag doesn’t say anything about which of the two is continuous and which one is interrupted. In this German counterexample both the footway and the crossing have surface=paving_stones but the crossing is not continuous. Osmuser63783 (talk) 22:30, 13 November 2023 (UTC)
This is a bit of a nitpicking but the counterexample is in Poland, not Germany (notice the yellow give way sign, not to mention the text on the sign below). :) --ManuelB701 (talk) 09:05, 15 November 2023 (UTC)


Resolved: Bkil confirmed to me they have no further comments or questions. Osmuser63783 (talk) 08:27, 28 November 2023 (UTC)

We sometimes add colour=* to a crossing that is visually distinctive (pink, magenta, yellow or white usually) from the road. Bkil (talk) 08:51, 10 November 2023 (UTC)

See the rightmost counterexample: A continuous sidewalk is specifically a crossing which doesn't break the sidewalk. A change in surface means a discontinuous sidewalk. Relatedly (since this came up in your previous question), a break in surface means a crossing is not a continuous sidewalk either. Edit: Reading it again, it can also work in favour of a continuous sidewalk. At the same time, a lot of sidewalks don't have any colour=* either, be it part of a highway or mapped separately, and colour=* thus isn't really helpful if not overkill for just this purpose. --ManuelB701 (talk) 12:09, 13 November 2023 (UTC)


Resolved: Bkil confirmed to me they have no further comments or questions. Osmuser63783 (talk) 08:27, 28 November 2023 (UTC)

Why is the presence of tactile_paving=* on both sides not considered an interruption according to your definition? Bkil (talk) 08:54, 10 November 2023 (UTC)

In Germany, tactile paving would be an indicator that pedestrian crossing doesn't have priority, but in other countries, it's apparently different. I guess every country needs to find their own definition of what to consider continuous and what not. Nadjita (talk) 18:04, 10 November 2023 (UTC)
The concept of tactile paving is not necessarily about priority. It helps one with visual impairment to know which direction to go further to reach the other side on the shortest path. It may also help caution the pedestrian to listen around before they walk, although if they are fortunate enough to have a guide dog, it will already be capable of signalling that. Lacking this (and lacking changes in surface, kerb, traffic calming, a nearby wall or a dog), a visually impaired parson "might" start to wander off the middle of the road if it "feels" like a pedestrian footway, and that could be dangerous as well. Bkil (talk) 18:12, 10 November 2023 (UTC)
In the UK, whether or not continuous sidewalks should have tactile paving is still an open question, but when tactile paving is present that doesn't mean it's no longer considered a continuous sidewalk (see the linked article for an example). Therefore the proposal provides for continuous crossings both with and without tactile paving, and emphasises that it is the overall visual impression that matters. As Nadjita mentioned, the situation may be slightly different in different countries e.g. when continuous crossings have special legal status. Osmuser63783 (talk) 23:24, 12 November 2023 (UTC)

What is this tag good for?

Resolved: Bkil confirmed to me they have no further comments or questions. Osmuser63783 (talk) 08:27, 28 November 2023 (UTC)

I tried to raise a concern that all semantics I could think of can already be clearly represented in fine detail using their own tags. What is it that makes marking a "continuous crossing" so valuable to you? Note that this term itself is not known in our country, so I may be missing some of the connotation of the word that it implies at your locality. Bkil (talk) 08:33, 13 November 2023 (UTC)

Great question! The reason I went looking for a tag was because in the UK, local councils have started using this design for sidewalks across residential roads that are sometimes fairly busy (in the past this design was only used for the entrances to driveways and similar). This has led to quite a bit of conflict and debate (examples 1, 2, 3, 4]). They are known by various names such as "continuous footways" and "Copenhagen crossings" and considered a distinct type of crossing. They are not unique to the UK: For example, in French, they are known as a trottoir traversant, in the Netherlands as an uitritconstructie, and in German-speaking Switzerland as a Trottoirüberfahrt, though my impression is that often these are technical terms known only to urban planners and traffic designers. In the community forum discussion it became clear that we do not have a widely established tag for them, and people have been discussing for a while how best to tag them. Some of the aspects and details can be expressed by existing tags but no other tag or combination is semantically equivalent. For example, if a crossing has traffic_calming=table and kerb=no (227 combinations) then there is a fairly good chance that it will be a continuous crossing, but as the examples in the proposal show, continuous crossings do not always qualify as a traffic_calming=table, and they may have a flush kerb instead of no kerb at all. In the discussion it was also pointed out that the design is not limited to sidewalks (here is an example where a highway=pedestrian crosses a road), so we decided to propose a tag for the more general case. Osmuser63783 (talk) 22:27, 13 November 2023 (UTC)

Further discussion

Most of the discussion during the proposal stage happened on the community forum, in 1 and 2. Osmuser63783 (talk) 08:27, 28 November 2023 (UTC)

Ambiguity with raised continuous same-surface crossing

How is a raised continuous crossing like this supposed to be tagged where both surfaces are continuous asphalt? This fits all the listed criteria for being a continuous crossing - it's basically the "ideal" case with no kerbs or even minor surface changes. But the counterexample specifically says speed tables are not included "because the road surface continues". Cars would have a speed table here but it's part of "their road" and would hardly even notice the bump. So this is neither (or both) "vehicles have to cross a section of footway or cycleway" nor "pedestrians or cyclists having to cross the road", at least from descriptions and examples given. HellMap (talk) 19:01, 2 December 2023 (UTC)

What you brought up is simply a case of the second counterexample in that the surface is broken by a separate surface colour and thus is better tagged as crossing:markings=surface instead of crossing:continuous=yes. --ManuelB701 (talk) 14:45, 5 December 2023 (UTC)
Still I think this explanation is insufficient. Imagine the same crossing but without any paint or markings, just plain grey asphalt and no curbs (I know a few in my hometown but it is in Ukraine so I wouldn't be able to post a link to Mapillary). I think that either option would be good whether to consider it a continuous sidewalk or to call it a regular crossing - however this needs to be clearly documented with an example so that such crossings could be tagged uniformly anywhere on the OpenStreetMap. --VileGecko (talk) 17:23, 10 December 2023 (UTC)
The example I gave has two parallel crossings - one for cycleway and one for footway. Footway one has no clear break or surface change. Cycleway in this instance has only paint. But is that really the deciding factor for this? Is the proposal really saying one is continuous and one isn't? I would not even call "paint" a "clear break" - this isn't even different surface and will have rubbed off in a couple years. This is all way too subjective, so this needs to be documented and explained properly. HellMap (talk) 21:25, 11 December 2023 (UTC)