I would think that segregated=yes is redundant, because sidewalk=left already indicates a separate designated space for pedestrians.
I would not bother with lanes on cycleways. Cycleways are (by worldwide default) bidirectional and each direction if indicated, has one lane. The ‘separation’ is usually a dashed line. I would reserve direction indications and lane tags for exceptions only.
I certainly hope that StreetComplete will not start asking “Does this cycleway have a dashed line in the middle?” and set lane tags on all cycleways…
I’ll note that in this location, cycleways are by default oneway and require an additional sign (not present at the time this picture was taken) to indicate if they’re not.
The lack of the legally required sign does not change the ground truth here, as there are other ways to determine this particular cycleway is bidirectional.
Regarding lanes, I think that most of the time just lane_markings=yes says enough about what the lines are. But whichever tags you choose, I think having that tagged is worthwhile, as it is meaningful information about the cycleway.
I did run past some interesting new infrastructure this morning, but didn’t think to stop an take a photo. Basically it was an area at the corner of an intersection, with straight lanes and turning lanes for bikes interspersed with traffic islands for pedestrians, all with just paint. I’ll try get a photo next week, should be fun to work out the tags and ways to represent that
In that case, it is necessary to add oneway=yes to cycleways. That is the way to tell OSM-applications what the local situation is. In Nederland, cycleways along non-urban roads are by national default oneway, so we have to add oneway=no. We can’t expect data users to know and apply our legislation and national defaults.
I would argue that data consumers need to have some knowledge of the default consequences of things being what they are. Which is why there is a whole section dedicated to that on the wiki.
That’s interesting. I guess it would be hard for a data consumer to know which road qualifies for this, even if the legal rule was encoded. How DO you know?
the problem is that many ideas about defaults on wiki are personal ideas and preferences of whoever wrote that and are altogether ignored by data consumers and not known by mappers
even if some subset of roads/cycleways/etc is oneway by default I would not really expect any data consumers to guess it
Yeah, if the cycleway is actually one way. In practice the vast majority are twoway, but our legislators thought it was a great idea to make every municipality add additional signs to them
If you see a cycleway sign, you already know you can ride in that direction; the only additional information the twoway sign provides is that someone might legally be coming the other way. This is… absolutely useless in most parts of the country.
In terms of OSM tagging, here a mapper on the ground could see cycleway signs at both ends of the way, so I guess it’s ”oneway, both ways”
I fully agree.
This might be an interesting example:
I live and map in the Netherlands too and have never heard of this.
Could you point me to where I can find this in the Dutch traffic code? RVV 1990
I only know the “keep right” rule in article 3 that implicitly makes usage of cycleway on the right side of the carriageway mandatory -even with no oneway signs present- when a carriageway has cycleways on both sides (such as here) and if both cycleways are of the mandatory type (sign G11 instead of G13) and depending on the way the traffic signs and road markings are placed. So it always takes some interpreting by the road user / mapper.
So as far as I know its conditional on the presence of 2 cycleways and I know no distinction in rules between within or outside the built up area (traffic_zone).
I read it somewhere in an explanatory text. Maybe it’s different, like you explained, but the point I made is the same: if national defaults, rules or customs differ from the world wide default, tag the attributes explicitly, so apps and non-locals know what’s the local situation without having to know countryspecific or local legislation.
Sounds like Germany where sign 1000-31 (up-and-down arrows) is sometimes noted that a cycleway (especially a sidepath) or bicycle-inclusive footways is bi-directional but is by no means mandatory and could lead to confusions if a sign in the other direction has been removed but not the additional sign.
I would expect a data consumer to know that the default is that anyone can use a road. Or that the default is that motor traffic cannot use a cycleway.
Yes, I agree that the tables with national defaults in the wiki should not be used to decide what to that and what not to tag.
But the same goes for assumed “world wide defaults” , whatever you, I or someone else may assume those to be:
the simple fact that some value is mentioned is some wiki default (either global or national) should not be used as a criterium for what to tag and what not to tag. Because only tagging values that are different from this and/or that default in some tabe introduces unnecessary uncertainty for data users en future mappers based upon assumptions and it obscures the difference between a way that is “not evaluated” and “is evaluated and value is similar to the default that I presumed”.
As an example how unfortunate a tagging strategy is where you would only tag values that are an exception to both to a presumed global and national default:
If you look at actual tagged values for highway=primary in the Netherlands:
97,16% is tagged for bicycle=*
96,87% has a negative value (no or use_sidepath)
0,28% has a positive value (yes)
2,84% remains untagged for bicycle=* (131 ways)
For the 2,84% that is untagged, would someone with an “exception only” tagging strategy as personal preference assume that all these ways are actually bicycle=yes, since that is the default and that should not be tagged?
Or would it be possible that these 131 ways just aren’t evaluated yet and >95% of these ways could probably have a negative value just like the tagged ways?
And when evaluating these 131 roads and encountering rare ways where cycling is actually not forbidden, would someone with exception only strategy really omit to tag bicycle=yes “since this is both the global and national default”?
(or even discourage other mappers to place such tags or -even worse- remove these tags as some mappers have done in the past? )
There can be good reasons to omit some tags, but only looking at contents of default tables as a rule is not among them.
though I would not expect data consumer to know that you can cycle on sidewalks in Poland that are wider than 2m and along road with speed limit higher than 50 km/h. And I would expect mappers to apply bicycle=yes in such cases.
In the same way I would not expect data consumer to know that in Nederland, cycleways along non-urban roads are by national default oneway. I would still expect oneway=yes to be tagged on them in such case.
(in both cases “along” part makes it worse in multiple way and more worthwile to tag values explicitly)
bicycle=no differs from the world wide default, so does use_sidepath. This shows that tagging this exception for the primary roads is very complete, suggesting that a systematic update workflow has been applied.
Furthermore that the national default for primary roads is not to allow bicycles and to redirect bicycles to a side path. I think the country specific OSM default for primary should probably be bicycle=no. All the more reason to be explicit both ways in this case.
In such a case it makes sense to tag explicit no’s, as an exception to the world wide default, and explicit yes’s as an exception to the overwhelming national customs. Still, if the tag is missing, I would expect data users to allow cycling (or apply country specific rules, which I do not expect to happen any time soon). Routers might apply a small penalty for not having an explicit tag. but that doesn’t change the default access value.
Anyway, this was just an example. The point is, you always have to tag the exceptions (or further refinements, such as use_sidewalk to refine no) to the default explicitly. That said, there can be valid reasons to also tag the default itself explicitly.
It seems I misunderstood that! In practice it’s true, but there is no formal difference between urban and non-urban roads. The rule is much simpler (from the view of the legislator), and even more unknown (by road users) than I imagined… Which stresses the point even more: you can’t expect data users to know and apply these country specific rules. We have to translate the local rules into unambiguous world wide OSM tags.
