see also recent discussion here:
Note that in Netherlands you have some cycleways in form and partially in function being actually close to roads.
So I understand why they tag this way, even if I am not a fan (and took several extra hours of work, not sure how it ended with StreetComplete support).
Mechanical retagging is the easiest part - but I expect that there will be no consensus for it. Even if it would be a good idea (and I am not entirely sure would it be).
(not sure is it visible from my comments in Does this cycle path have a sidewalk? (ask only in Netherlands) · Issue #3785 · streetcomplete/StreetComplete · GitHub but my reaction was “oh my God, another tagging schema” - but according to what I know now I would not pursue bot retagging attempt and deprecation)
One of the cardinal rules of OSM is that you don’t get to decide for other people what’s worth mapping and what isn’t.
I mean, personally I find all these buildings a damn inconvenient nuisance that just make the .osm.pbf slower to download, but I don’t propose banning them. (Partly because I suspect I’d be refused entrance to France if I tried.)
This seems the preferred tagging to me when cycleway IS the main way, like on all Dutch cycleways that use sign G11. The traffic sign in your picture doesn’t exist in the Netherlands.
I’d discourage method 2 because of the redundant tagging of bicycle=designated on a cycleway and method 1 on ways that are officially cycleways. Additionally, changing the sidewalk tag to segregated, would loose information about location of the sidewalk (left, right, both).
Differences exist between different countries, which also causes tagging differences in Openstreetmap. That’s why we have local communities and local knowledge is important. Especially in a cycle country like the Netherlands.
From a Dutch perspective I’d ask: why would you tag a sidewalk of a street differently from a sidewalk of a cycleway?
That is legally so, but the on-the-ground appearance of these ways in the Netherlands is no different from those in the rest of the world.
That’s true, but I wonder when that would make any difference to a map user. But if there is a use case for it, it could be mapped with something like footway=right
, etc.
Absolutely. I am Dutch so I know the local situation. But I’ve lived in several other countries, so I also know that the Dutch situation is not unique.
To align with international practice. I think we should aim for tagging practices that are the same world wide, as far as is practically possible. Here I see no reason why data users like StreetComplete should adapt to the Dutch practice. It would be much easier, seen on a worldwide level, if the Dutch would adapt to the international practice.
So basically the established cycleway tagging has to be changed to make it easier for StreetComplete? And details that you don’t think are important can just be ignored?
StreetComplete is an editor that takes away much of the tagging work from the mapper. But by taking this task upon itself, it also takes the responsibility to map correctly according to local tagging guidelines upon itself.
You also seem to ignore the legal side of road appearance completely. But if something looks the same in different countries, it doesn’t mean that it is the same.
Exactly the things that sets them apart and make it easy and worthwile to tag these sidewalks as sidewalk=yes. Segregated is more suitable for line-in-the-middle segregation, which is much more frequent outside Nederland.
I expect the practice will be adopted elsewhere once cycleways with true sidewalks (kerbs, pedestrian pavement, traffic signs) will be used systematically.
That is exactly what I was going to say. First we tag it as well as we see fit, then output-oriented data users can profit from it. QA tools could support it - or not, as they please. Of course, I’d rather they did.
Re-tagging for the tools is not my preference.
cycleway:surface
and footway:surface
make NO sense at all on a highway=cycleway
It is a cycleway and it has a surface. Not a cycleway:surface
If there is a sidewalk next to a cycleway tag it as sidewalk=
If this logic continues are we going to get: residential:surface
and footway:surface
on highway=residential
next?
If in any other country cycleways have proper sidewalks in the same way streets have, they could do a lot worse than to tag these as highway=cycleway
plus sidewalk=right|left
. There is a fundamental difference between a shared cycleway/footway and a cycleway with a sidewalk. I’ve explained this difference a while back in forum posts referenced in the related issues.
If you dislike those cycleways, the advised solution is to just not visit the Netherlands, the country is full of them! (But leave the tags alone please.)
That … isn’t really my experience? … at least based on cycleways in the UK, the Netherlands, Australia, and a few other places. If you showed me pictures of cycleways from those countries with all signage covered up, I’m pretty sure I’d be able to tell the difference, even of ostensibly the same feature (e.g. a segregated cycleway with pedestrians to one side).
Indeed. It is one of the most noticeable things when going abroad and seeing the cycling infrastructure.
@rhhs This is typical in the Netherlands: a street with a sidewalk on the left, turns into a cycleway with the same sidewalk continuing on the left (for cars thus a dead-end street, obviously). (Photo from Google Streetview, because it’s rather dark outside now.)
The street and cycleway are mapped with a line following their centre, and the sidewalk is either mapped separately or with an attribute like sidewalk=left
. It wouldn’t make sense to place the line in the middle of both the sidewalk and the cycleway here.
Here’s an example of a cyclestreet continuing as a cycleway (usually highway=residential
or highway=unclassified
with cyclestreet=yes
).
Please read this thread (which was already linked here) so you know what the mapping consensus is here in The Netherlands. And especially why we use this method, which is quite nicely explained there.
In my opinion this is really quite alot for a tagging combination that is (almost) only used in such a small country as NL.
There are a few cycleways with small sidewalk (separated by kerb) in Slovenia, too.
Number 3 (sidewalks) is different on the ground and therefore it makes totally sense to tag it differently.
If someone wants to unify tagging, they should try to get rid of number 1 or number 2, because they describe the same thing on the ground.
Can someone please enlighten me on the legal difference between tagging 1&2 vs. 3 in the Netherlands? Is it described somewhere that a footway (“voetpad”) that is adjacent to a cycleway (with or without height difference) is legally always a sidewalk (“trottoir”)? And that it is legally relevant that a footway is a sidewalk? I searched the Dutch traffic law (RVV) for “trottoir” and found that it always occurs together with “voetpad”, i.e. there seems to be no legal difference between them as far as traffic rules are concerned. “trottoir” is not defined in that law. Is there another law that makes a legal distinction between footway and sidewalk?
I don’t think that object properties have to have legal implications to be tag-worthy.
For example, not everywhere in the world are there legal implications of hipped-and-gabled roof shape or colour of a bench, yet it’s ok to tag these properties.
Sorry to barge in as I’m neither Dutch nor a significant contributor to cycleways My attention was attracted because the second pattern sometimes causes difficulties for me when I map hiking routes and I find “cycleways” that are obviously plain paths or footways where cycles are allowed.
Reading this thread reminds me of discussions we had about hiking and cycling routes and where an important concept was introduced: the operator’s intent. It seems to me that it is what you are trying to capture when seeking legal references or visual clues. How many different intent patterns are there?
They make sense if it’s tagged with segregated=yes, which is the only time I recall those tags being applied by StreetComplete.
A large part of my commute used to be on ways (correctly IMO) tagged highway=cycleway + segregated=yes + cycleway:surface=asphalt + footway:surface=paving_stones
The legal difference is irrelevant. You can see the biggest difference in the photo I posted or the Google Streetview link. It’s a matter of physical design, not legality. In the Netherlands sidewalks can be found alongside streets and cycleways, and it makes sense to map those with the ways aligned to the centre of the street/cycleway (which excludes any sidewalks).
(Legally a cycleway is signposted with one of the three common signs, and legally, pedestrians must use these unless a sidewalk or nearby roughly parallel footpath is present, in which case use of that path is mandatory.)
I’ve documented this on the wiki too (under Usage on cycleways). If we were to map cycleways with sidewalks as shared cycleway/footway paths, the centre line would shift. That doesn’t make sense when the sidewalks are built the way they are in the Netherlands.
I would say that regardless of the presence of segregated=yes
, the presence of a truthy sidewalk
value means that cycleway:surface
and footway:surface
are incorrect and should be surface
and sidewalk:surface
instead. A way with a sidewalk implies segregated=yes
. Only without a sidewalk (or negating sidewalk
values) and segregated=yes
should it be considered a combined path.