[Tagging] Reviving the path discussion - the increasing importance of trails in OSM

Fri May 22 10:52:16 UTC 2020

Ok, so I realize there will not really be any other way to distinguish an
urban, paved path from a small forest path, other than by other attributes
than highway=path itself. Path=mtb is nice for paths specifically created
for MTB and nothing else. But I don't see an easily verifiable way of doing
the same for other forest/mountain/meadow paths.

So we're stuck with other attributes, which mappers should be encouraged to
always use together with highway=path. Like there should never be a
highway=path without a surface tag. Currently only 21% of highway=path has
a surface tag, which contributes to the problem we're discussing.

Then there is width, which is only tagged on 3.5% of highway=path. I was
discussing width of paths in another forum. For a forest path, would you
say width is measured as the actual tread on the ground only? For a runner
and MTB cyclist that would make sense, but for a hiker with a big backpack
a width of 0.3 m may mean they think it's not possible to walk there.

See these three paths for example:
1. https://ibb.co/TkJ2V1g
2. https://ibb.co/Cmtp6LK
3. https://ibb.co/qgjW5dz

/Daniel

Den tors 21 maj 2020 kl 23:09 skrev Volker Schmidt <voschix at gmail.com>:

>
> I am not a fan of the confusing use of highway=path  for foot-cycleways
> and narrow mountain hiking ways, but that is a fact in OSM, and we need to
> live with that.
>
> However I would like to underline that highway=cycleway or highway=path +
> foot=designated + bicycle=designated do not necessarily imply the
> suitability of the way for normal bicycles.. These tags only tell you about
> he legal access of the way. Surface, smoothness, and width (or est_width),
> together with the elevation profile (data that is not in OSM) are also
> needed for bicycle routing..
> For hiking paths you have in addition SAC-scale and MTB-scale.
>
> Examples of unpaved cycleways in my city:
> https://www.mapillary.com/map/im/Ezjn-npOmRSQ-dHkMztzl
> <https://www.mapillary.com/map/im/Ezjn-npOmRSQ-dHkMztzlQ>  (cycleway)
> https://www.mapillary.com/map/im/DWHevDzL7i9eQDYSNbvCJcg
> <https://www.mapillary.com/map/im/DWHevDzL7i9eQDYSNbCJcg> (foot-cycleway)
> https://www.mapillary.com/map/im/lAnBsThrDjTxjhvfXhB0Yg (cycle lane)
>
> The problem is guessing by routers in case of incomplete tagging. Just to
> get myself an idea I checked:
> My city shows 1533 ways tagged as cycleways and foot-cycleways, of which
> 91.7% with surface, 54.7% with smoothness, 52.1% with width
> (This excludes all cycle lanes and a few cycleways that are not present as
> separate ways in OSM)
>
> Basically we have the instruments - let's use them instead of inventing
> new tags.
>
>
>
> On Thu, 21 May 2020 at 16:15, Adam Franco <adamfranco at gmail.com> wrote:
>
>> For those who missed it, a related discussion was just had on this list
>> about differentiating mountain-biking trails from cycleways.
>> See the resulting proposal for path=mtb
>> https://wiki.openstreetmap.org/wiki/Proposed_features/Tag:path%3Dmtb and
>> threads from April in Tagging:
>> https://lists.openstreetmap.org/pipermail/tagging/2020-April/051864.html
>>
>> On Thu, May 21, 2020 at 8:51 AM Andy Townsend <ajt1047 at gmail.com> wrote:
>>
>>>
>>> On 21/05/2020 10:50, Mateusz Konieczny via Tagging wrote:
>>>
>>>
>>>
>>> Similarly anyone creating
>>> highway=footway + danger="you will be shot" + "access=no" + foot=yes"
>>> should probably switch to pickpocketing, telemarketing or other less
>>> harmful activity.
>>>
>>> While "danger" isn't a much used tag (and I'm sure wasn't a serious
>>> suggestion here - https://taginfo.openstreetmap.org/keys/danger#values
>>> ), sometimes "foot=yes" is correct and other tags need to be taken into
>>> account.  I've used the area around
>>> https://www.openstreetmap.org/way/431056034 as an example of that
>>> before.  Here "foot=yes" is correct - there is a legal right of access.  "
>>> sac_scale
>>> <https://wiki.openstreetmap.org/wiki/Key:sac%0D%20%20%20%20%20%20%20%20scale?uselang=en-GB>=demanding_alpine_hiking"
>>> also makes sense here I think.
>>>
>>> I take Frederik's reference to Andy Allan's point about "a
>>> multi-billion-dollar-revenue organisation that were rendering anything with
>>> a highway tag the same as their most minor road style" but frankly there's
>>> simply no solution to that - presumably "highway=dangerouspath" (to make up
>>> a nonsensical value) or
>>> https://taginfo.openstreetmap.org/tags/highway=via%20ferrata would
>>> still get shown as a "road".
>>>
>>> Map styles need to be clear about what they're showing and what they're
>>> not showing and people using maps need to be able to read maps so that they
>>> understand what they're being told.  This isn't really a tagging issue,
>>> unless OSM mappers aren't using appropriate other tags when they should
>>> (sac_scale, trail_visibility, surface, etc.)
>>>
>>> Best Regards,
>>>
>>> Andy
>>>
>>>
>>> _______________________________________________
>>> Tagging mailing list
>>> Tagging at openstreetmap.org
>>> https://lists.openstreetmap.org/listinfo/tagging
>>>
>> _______________________________________________
>> Tagging mailing list
>> Tagging at openstreetmap.org
>> https://lists.openstreetmap.org/listinfo/tagging
>>
> _______________________________________________
> Tagging mailing list
> Tagging at openstreetmap.org
> https://lists.openstreetmap.org/listinfo/tagging
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/tagging/attachments/20200522/d236a638/attachment.htm>