This discussion was done to death elsewhere, and one thing that came out of that was people in the UK saying that there was a widespread understanding here that bicycle=no meant “no cycling AND you’re not allowed to push your bike” and bicycle=dismount meant “no cycling BUT you are allowed to push your bike”. Britain and Ireland are responsible for about 10% of both bicycle and bicycle=dismount internationally (see here, here, here and here), and dismount is not disproportionately used here.
There was a suggestion to clarify what bicycle=dismount meant by also tagging bicycle:pushing=yes/no, but in the UK only one person (me) seems to have tried to do that.
This leaves us in the UK with a lot (about 24k) of highway=footway;footway=sidewalk;bicycle=no. Locally to me, that looks like this. This most charitable way of describing that tagging is “incomplete” (because it does not distinguish whether pushing a bicycle is allowed or not). Given the historical understanding of what bicycle=dismount means in the UK I’d argue that that tagging is simply wrong. It is absolutely incorrect (as suggested by the person who started this thread above said) to suggest that cycling on (for example) this sidewalk is legal, but it is equally incorrect to suggest that pushing a bicycle is prohibited on sidewalks such as that.
Correct me if I’m wrong but I understand that this is an informational page for data consumers, not some sort of default that OSM applies to otherwise untagged objects. When you pull data from OSM you only get what is explicitly tagged on the object.
The problem @Pete_Owens is trying to “fix” is cycling-focused routers advising cyclists to walk their bikes along the pavement, on the assumption that cyclists will not walk but instead would cycle along the pavements. As noted above this seems to be working (e.g. breaking CycleStreets routing, which respects bicycle=no).
It is incredibly clear in both cases that the cyclist needs to dismount and push. In both cases it is a sane routing decision - walking a short distance on a good-quality path to avoid a 2.1km round trip. The mapping practice under discussion breaks this.
I don’t want to derail this discussion, but the view that pushing a bicycle along a Public Footpath without permission is trespassing, is not a fact. It’s an argument put forward by Walking Organisations, while Cycling Organisations argue it is allowed. It’s a complex issue that would ultimately need a court judgement.
Slightly off-topic indeed but I wouldn’t characterise it quite like that. Cycling UK think a bike could be a usual accompaniment. Pretty much everyone else - not just walking organisations, but PROW officers and highway authorities too - thinks it isn’t. Plenty of councils have documents on their websites saying “no, you can’t push a bike on a public footpath”. Until there is case law then yes, the issue remains technically unresolved, but - sadly! - the balance of opinion is against pushing.
(still on pushing a bicycle on a public footpath) it’s perhaps also worth mentioning that there are plenty of public footpaths where people pushing bicycles is permitted, even though there is no explicit legal access (perhaps by explicit signage, perhaps by “general tolerance”)
Because it’s a grey area, I personally wouldn’t tag bicycle=anything on a public footpath unless there was explicit signage. That doesn’t mean that a router in England and Wales should default to sending people with bicycles down public footpaths with no signage - there’s no “right to roam” here (Scotland is of course more enlightened in that regard). We tend not to tag “the law” (although some people sometimes do), and people will expect routers to only display legal routes to them.
(perhaps @mods-uk might consider moving these “pushing a bicycle on a public footpath” replies to a new topic? It’s an important point, but not really related to hoghway=footway; footway=sidewalk).
I can’t agree with that because I know local PROW officers and council highway officers that are of the view that a bicycle is a “Usual Accompaniment”. To be fair, Highway Officers views change depending on the direction of wind, who they are talking to and political pressure.
The “Usual Accompaniment” definition comes from legal Pavement (footway) case law. If a bicycle is not a usual accompaniment then this view must apply for pavements (footways). This would mean pushing a bicycle on a pavement (footway) would be crime.
I am not aware of any prosecutions, or challenges, for pushing a bicycle along a pavement (footway). In practice it is the opposite and Society and Appropriate Authorities (and routers) clearly expect and plan for bicycles to be pushed along pavements. You have extensive cycling infrastructure that can only be accessed by pushing a bicycle along pavement (eg cycle parking). It would be interesting to ask the Highway Authorities that publish advice against pushing on Public Footpaths, to also give a view on pushing on a pavement (footway)
Ultimately this argument is a matter for a court. Personally I just can’t see the view of The Ramblers being accepted. I think there is too much historical evidence of bicycles being pushed on pavements by pedestrians.
Fortunately it’s not a significant issue for OSM UK.
I knew this would happen as soon as I saw the split topic title
I don’t think that there is any doubt that (in the absence of unusual, specially signed, circumstances) pushing a bike on a highway=footway; footway=sidewalk is enyirely legal (that was what the other thread was about).
This one is about pushing a bike on a highway=footway; designation=public_footpath.
I’m not aware of any case law on that (and am certainly not a lawyer). I can think of councils that have documented “we don’t think it is a legal right to push a bike on a public footpath”. Is there anyone (apart from lobby groups like Cycling UK) in the other camp?
There is the case law in Crank v Brooks [1980] RTR 441 where it was held someone walking and pushing a bicycle on a zebra crossing is a foot passenger, with the judgement text and discussion on the CyclingUK forum found via Google.
Of course all these arguments do not apply to viewers in Scotland, where the Roads Scotland Act 1984 S129(5)(b) makes clear that it is legal to propel a pedal cycle which is not being ridden on a foot[way/path].
The subjects are linked because the Public Footpath arguement is based on case law for English legislative Footways (pavements). The argument given MUST first apply for Footway (pavements) before it can apply to Public Footpaths.
The legal arguement used for opposing pushing a bike on a Public Footpath ROW is that a bicycle is not a “Usual Accompaniment of a pedestrian” test. But, this test was created for deciding what can be used by a pedestrian on a English legislative footway (pavement).
I’d suggest it’s wishful thinking by walking groups to believe a judge would decide pushing a bicycle on a pavement fails the test, and therefore pushing a bicycle on a pavement is a criminal offence.
Are any walking groups (or anyone else, with the possible exception of the mapper in the other thread) actually claiming that pushing a bicycle on a pavement by the side of a road is not allowed? I’m not aware of any. It’s an entirely separate issue to “is it allowed to push a bike on a highway=footway; designation=public_footpath in England and Wales, in the absence of any signage expressing the landowner’s permission or prohibition?”.
The key decision there seems to be that someone pushing a bike with both feet on the ground across a pedestrian crossing is a “foot passenger”, and a motorist had to slow down. That seems reasonable, but it is a different claim to the one above that “the view that pushing a bicycle along a Public Footpath without permission is trespassing, is not a fact”.
I am not aware of any case law on the latter (which I guess would mean it is neither “true” nor “untrue” until such case law exists, although as noted previously there is a fairly widespread expectation that a decision would go one way). There’s a whole rabbit-warren of similar-but-not-quite-the-same cases (e.g. here - no, you can’t legally use a public footpath to observe racehorse training on private land!).
Further, there can be no basis for distinguishing highways on publicly owned land and privately owned land. The nature of the public’s right of use of the highway cannot depend upon whether the owner of the sub-soil is a private landowner or a public authority. Any fear, however, that the rights of private landowners might be prejudiced by the right as defined are unfounded. The law of trespass will continue to protect private landowners against unreasonably large, unreasonably prolonged or unreasonably obstructive assemblies upon these highways.
Finally, I regard the conclusion at which I have arrived as desirable, because it promotes the harmonious development of two separate but related chapters in the common law. It is neither desirable in theory nor acceptable in practice for commonplace activities on the public highway not to count as breaches of the criminal law of wilful obstruction of the highway, yet to count as trespasses (even if intrinsically unlikely to be acted against in the civil law), and therefore form the basis for a finding of trespassory assembly for the purposes of the Public Order Act. A system of law sanctioning these discordant outcomes would not command respect.
The two are linked. Case Law for pavements use created the "Usual Accompaniment” test. This test for pavement use is also considred applicable for Public Footpath access rights. A single test that applies for pavments and Public Footpaths.
I’d personally like to bring this to an end (for me at least). Interesing topic, but I think it has little “on the ground” impact" for OSM. And as a member of Rambler and LDWA it feels a bit weird to be taking this position
EDIT: And while maybe it may be treated as distinct in UK, CycleStreet seems to be covering:
UK and Ireland, much of northern/western Europe, and various cities around the world.
Thanks for flagging up that out-of-date text. Now updated.
As to the main point, we have always interpreted bicycle=no since 2008 as meaning no cycling, no pushing.
So yes, indeed, bicycle=no does not (in our view)mean the same thing as bicycle=dismount. Which is why there are different tags.
As discussed on another thread, we would like to see:
A proper tag for pushing a bicycle not permitted (e.g. shopping centres, public footpaths, etc.)
A tag for advisory Cyclists Dismount sign, i.e. the UK-specific civil insurance advisory signage, which is (confusingly) nothing to do with bicycle=dismount
Once these are done, dismount could then become a synonym for no, as the current situation is far too confusing. Until that point, our position remains that no and dismount are different tags as described above.
We strongly recommend that people do not put any bicycle=… tag on a general pavement (sidewalk), as that is redundant information. Routers should be handling that implicitly, just as people shouldn’t be cluttering up the database with bicycle=no on every highway=motorway, or oneway=yes to every junction=roundabout.
I want to defend explicit tagging in at least some cases.
For example, is cyclestreets or any other router supporting “cycling is legal in Poland on sidewalk wider than 2m, along roads with speed limit higher than 50 km/h”?
this implication does not necessarily follows, as far as I know CycleStreets is also not a king of OSM tagging
and this is just not really relevant, and as far as I know inaccurate
actual reason is that bicycle=dismount was invented when bicycle=no was already widely used
to be more specific, from my old notes, bicycle=dismount reached 100 uses worldwide on ways on day 2009-08-27. On that date bicycle=no was used 46 681 times on ways, with usage increased by 163 just on that day (source, removed since then from that page)
are you really sure that at that point bicycle=no was actually used in meaning “not even pushing bicycle allowed”?
and “we have two values for that” quite often does not imply “there is difference between these values”
you probably meant oneway=yes
campaign to remove oneway=yes on roundabouts and rely on implication saved little effort, added extra effort and made oneway processing slightly more obnoxious. Not convinced it was a win.
Agreed, that should be explicitly tagged as bicycle=yes, because (a) it is a pretty clear exception to the norm from around the world, and (b) people have to have gone out and measured the width presumably to be sure!
PS Have corrected the oneway tag, thanks, and made clear the difference is our view.
campaign to remove oneway=yes on roundabouts and rely on implication saved little effort, added extra effort and made oneway processing slightly more obnoxious. Not convinced it was a win.
The main benefit of not adding redundant info like this is that it makes it easier for mappers to tell when something really is an exception case. If every highway was also tagged with surface, oneway, motor_vehicle is yes, you just end up with huge amounts of clutter. It also pointlessly increases the database size. When you are dealing with gigabytes of data, widespread clutter like this bulks things up.
… or, as in a current case in the UK, because one mapper insists on setting “no” when everyone else mapping there thinks that it is wrong.
I agree that it makes no sense to “set default, non-exceptional values”, but alas this current debacle has made the correct (default) value worth recording where someone has explicitly set an incorrect value previously.
