[Tagging] Usefulness of bicycle=dismount on ways

Robert Whittaker (OSM lists) robert.whittaker+osm at gmail.com
Wed Oct 9 07:45:57 UTC 2013

On 7 October 2013 17:09, fly <lowflight66 at googlemail.com> wrote:
> I wonder if it is useful to tag bicycle=dismount on ways.
> At least in Germany there is no official traffic sign despite of the
> existence of some.

I don't think the issue here is really whether there is a need within
instances of "no cycling" to distinguish between "no bicycles at all"
and "bicycles can be pushed". It seems from the posts below that there
are plenty of situations where both cases apply, and it's clearly
important information to know if you're planning cycling routes. We
therefore do need a way to distinguish between the two cases.

The big problem that I see is (especially in areas where the default
position is "no cycling" = "bikes can be pushed") that people have
used bicycle=no on ways where cycling is banned but it's fine to push
a bike. In other words the bicycle=* key has been used to express
access rights for cycling, not for bicycles. As a result (at least on
some areas) data users are forced to interpret bicycle=no as "no
cycling, but bikes can be pushed" as a best guess at what the mapper
meant. Thus bicycle=dismount actually add no further information,
except that you can be more certain that pushing bike is allowed.

If bicycle=* is currently widely used to express access rights for
cycling, then I'd suggest we leave it like that, as it does the job
pretty well. Rather than trying to add additional values to this key
to capture access rigths for pushed/wheeled bicycles (e.g.
bicycle=no_and_not_even_pushed), I'd suggest that we define an
additional key: Something along the lines of bicycle:pushed=*.
bicycle=* then tells you if you can ride a bike (as it does
currently), while bicycle:pushed=* tells you if you can push/wheel it.

Any cases of bicycle=dismount could be easily converted to bicycle=no,
bicycle:pushed=yes. The only issue is cases of bicycle=no which have
been used to mean "no cycling and no pushing either". Perhaps it will
be necessary to look at national defaults to handle this (i.e. what
value of bicycle:pushed should be assumed if bicycle=no and there's no
bicycle:pushed=* tag present).


Robert Whittaker

More information about the Tagging mailing list