[Tagging] Usefulness of bicycle=dismount on ways

Philip Barnes phil at trigpoint.me.uk
Fri Oct 11 13:38:10 UTC 2013


On Fri, 2013-10-11 at 11:53 +0100, Robert Whittaker (OSM lists) wrote:
> On 10 October 2013 15:28, fly <lowflight66 at googlemail.com> wrote:
> > +1 for a separate tag and deprecating  bicycle=dismount
> 
> To make the case for this clearer, consider the following. There are
> four combinations of access for bicycles and cyclists, depending on
> whether you are allowed to cycle and/or allowed to
> push a bike:
> 
> (a) Cycling and pushing both allowed
> (b) Cycling allowed, but pushing not allowed
> (c) Cycling not allowed, but pushing is allowed
> (d) Neither cycling nor pushing allowed
> 
> I beleive all of these combinations are possible in real life. In the
> UK (a) would be a normal cycleway that's shared with pedestrians, (b)
> could occur on a cycleway that's only for cyclists (i.e. no
> pedestrians allowed), (c) would be the case of (e.g.) a narrow bridge
> on a cycle route, where "dismount" signs are shown, or a typical
> pedestrian shopping street with "no cycling" signs, and (d) would be
> an area/route explicitly signed as e.g. "no bicycles not even pushed"
> (Oxford University Parks used to be like this until a couple of years
> ago).
> 
> Clearly if you are travelling with a bike you would want to
> distinguish between at least (a)/(b) vs. (c) vs. (d), to determine
> where you can go with your bike and at what pace.
> 
> Currently the tagging used is bicycle=yes/no/dismount. The problem
> with this is that while bicyle=dismount unambiguously indicates (c),
> people have used bicycle=no for both (c) and (d) -- interpreting it as
> either "no cycling" or "no bicycles". Also (although less importantly)
> using bicycle=yes offers no way to explicitly distinguish between
> cases (a) and (b).
> 
> I would therefore propose a new access tag be introduced to capture
> information about whether pushing a bike is allowed. I'll call this
> bicycle_pushed for now, but the actual name is something that can  be
> discussed and agreed upon later.
> 
> With this tag and the existing bicycle=* access tag (whose values are
> now taken, as I believe was originally intended, to apply to 'cycling'
> rather than 'bicycles'), it is now possible to unambiguously
> distingiush between the four cases above:
> 
> (a) bicycle=yes + bicycle_pushed=yes
> (b) bicycle=yes + bicycle_pushed=no
> (c) bicycle=no + bicycle_pushed=yes
> (d) bicycle=no + bicycle_pushed=no
> 
> bicycle=dismount is then deprecated, and the same information captured
> by using bicycle=no + bicycle_pushed=yes (i.e. no cycling, but you can
> push your bike).
> 
> For actual tagging use, It might be worth considering that whether, in
> the absense of a bicycle_pushed tag, the presense of foot=yes implies
> you can push a bicycle on that route -- which is probably a sensible
> default in most of the world. Although we would have to think
> carefully about how to handle the case of people who have previously
> tagged bicycle=no to indicate case (d).
> 
+1

b can also cover roads where pedestrians are prohibited, but cyclists
are allowed. A real life example I can think of is the A483 between
Chirk and Wrexham.

Phil (trigpoint)




More information about the Tagging mailing list