[Tagging] Tagging cycleways to distinguish them from combined cycle and footways

[Jens Glad Balchen]

On 03.08.2022 11:42, Jeroen Hoek wrote:
> On 03-08-2022 11:18, Jens Glad Balchen wrote:
>> On 03.08.2022 10:59, Jeroen Hoek wrote:
>>> Have a look at this typical Dutch situation:
>>>
>>> https://www.google.com/maps/@53.1426041,5.7397237,3a,75y,228.33h,88.04t/data=!3m6!1e1!3m4!1sAhezo74J5tsgo0u0mM-oNQ!2e0!7i16384!8i8192 
>>>
>>>
>>> On the right, a road, on the left, a compulsory cycleway. Dutch law 
>>> says that pedestrians must use the cycleway here (and they may walk 
>>> on its hard surface), and may not use the shoulder of the road 
>>> because the (much safer) cycleway is there. So the road gets 
>>> foot=use_sidepath, and the cycleway has no foot value (because the 
>>> Dutch default access tags have foot=yes for highway=cycleway), or 
>>> foot=yes.
>>>
>>> So you are saying that in Norway, in a case like this, pedestrians 
>>> can choose to use the shoulder of both the road and the cycleway?
>>
>> Yes.
>>
>> There are no compulsory roads of any kind in Norway, and the only 
>> mandate is the one for pedestrians that I quoted. Everything else is 
>> based on explicit prohibition.
>
> Ideally, navigation tools should be able to choose the best option 
> based on tags present without denying access to road users who can 
> legally use it. In terms of tagging you can choose to simply set 
> foot=yes on both road and cycleway (either directly, or implied via 
> the default access values for Norway), and then let navigation aids 
> apply their own logic. Often, when faced with equally long paths where 
> one is highway=primary and the other is highway=cycleway, navigation 
> aids will penalize the primary road for pedestrians, and choose the 
> cycleway. Similarly, when faced with a parallel highway=footway and 
> highway=cycleway, pedestrian navigation tools usually choose the 
> footway if it is not much longer, even if you are allowed to walk on 
> both.
>
> You can, potentially, further aid navigation tools by mapping 
> shoulder=* and/or verge=* (as appropriate) as well (in particular 
> where sidewalk=no). This can help navigation tools further penalize 
> roads that lack a hard shoulder.

The objective with this discussion is not how to tag for routing or 
navigation. That seems to already be accomplished by the existing 
tagging standard, and it seems to have been an overarching goal for a 
long time.

My objective as a map-maker is not to make a map for navigation, but to 
make a map of what our roads actually are. The purpose is to let road 
planners, city planners, and politicians see not just what the city and 
zoning plans tell us that we should have had, but what we actually have. 
When the city plan says there is a cycle network in a particular 
location, and my map shows that there are no cycleways, no combined 
cycleway and footways, and no cycle lanes, it will be apparent that 
there is something missing. To accomplish that, I need to be able to 
accurately determine what kind of road I am dealing with.

The problem, again, is that there seems to be an opinion among some OSM 
contributors -- at least as regards Norway -- that any highway=cycleway 
is a combined cycleway and footway unless specficially tagged otherwise, 
and that I as a map-maker cannot use highway=cycleway to conclude that a 
road is a signed cycleway. This is not very clear from the documented 
tagging standard for Norway 
<https://wiki.openstreetmap.org/wiki/No:Map_Features>.

The proposed way to circumvent "any highway=cycleway is a combined 
cycleway and footway unless specficially tagged otherwise" is to tag it 
with foot=no/discouraged, contrary to what the law states. This is the 
crux of the issue, that we are forced to tag contrary to the legal 
status just to be able to circumvent the (IMO) faulty and illogical 
initial stance.

Jens
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/tagging/attachments/20220803/9984dca8/attachment.htm>