I’ve just started getting involved with OSM recently.
Background: I’m mostly interested in cleaning up access=, designation=, prow_ref=, surface= for bridleways and footpaths in the countryside around north east Leeds. In principle there is some nice gravel biking, but it was hard to see what could and could not be accessed or ridden. At least one footpath nearby was incorrectly marked as a bridleway on OSM and cyclists seemed to keep removing signs asking them not to cycle because they thought it was a bridleway… Komoot always suggests routes along it and the Komoot guidance says routing issues like that should be altered on OSM. So I did.
What I’ve been doing: I’ve been adding or correcting prow_ref= and designation= from the Leeds City Council definitive map (OGL v3). For designated footpaths over farmland and estates I have been setting foot=designated, bicycle=no, horse=no and for designated bridleways foot=designated, horse=designated, bicycle=yes. I don’t usually fiddle with access on public highways. For adjacent paths/tracks on private land I have been setting access to private (for example paths marked on golf courses where a PRoW footpath goes through the middle).
Why I’m confused: Reading some of the PRoW threads on this forum, it seems there is a lot of guidance not to assume bicycle=no and not to be setting paths to private unless there is clear signage forbidding cycling or asserting private access. However, I don’t understand the logic. I think most land in England is private and there is no right of access (no right to roam). Therefore surely the lack of any signage implies private access. Therefore PRoW’s only grant a specific form of access and all other forms are by default private unless something more permissive is granted by signs on the ground. i.e. rather than looking from signs asserting private access, we should be mostly care about signs giving permissive access.
I feel like I’ve got the wrong end of the stick somehow, maybe from a misunderstanding of what we are and are not supposed to include in OSM. I feel like the edits I’ve been making are useful for routing and people who want to use maps for navigating PRoWs, but don’t want to keep doing the wrong thing if I’ve got it wrong.
I mean that I set the PRoW to foot=designated, horse=designated, bicycle=yes (it’s a bridle way) but someone has mapped all of the paths and golf buggy tracks on the golf course, so I set those to private.
In the case of the golf course there are plenty of signs asserting the private access, but over some estates and farmland this is not always the case.
My understanding is that while bicycles are allowed to use bridleways by law, the law only requires the bridleway be designed and maintained for pedestrians and horse riders. So it’s not really designated cycling infrastructure. There are many bridleways that are very difficult or impossible to bike. So a bike path or cycleway would be bicycle=designated because it has been designed and designated for bicycles to use, whereas a bridleway is bicycle=yes because it is legally allowed but was not created for bicycles to use.
But I also wasn’t so sure about the difference between these two access types so I just tried to be consistent with my edits. I’d be interested to know more about the correct choice.
I think there is a problem in assuming access =private due to the meaning it has in OSM. For many access =private means that any journey along that way needs specific permission. A way were your presence may be challenged.
The following goes off from your point, but I think explains part of the issue in assuming access = private.
It’s common in the UK to have a sign to state a road is private, but the intent is to stop cars parking or people having a random walk. Residents would still expect people who have a reason to go to their house to access the road. They would not expect a visiting friend to phone first and arrange permission to access.
There may be sign saying “PRIVATE” but what is meant is access = destination.
The example I think of is a school.
There will commonly be a path leading from the front gate to the reception door. I’d tag that as foot=destination to show you shouldn’t be on that path unless you have business at the school.
There will be several paths that lead away and go through the school grounds. I’d map those as foot=private to inform that you simply should not be on those paths unless you have clear permission, and you can expect to be challenged.
The same principles can apply to visiting a Farm (house).
For a Golf Course.
I map the car park, and paths to Club House, as =destination to show you must have a Golf Club reason to be there.
I’d map the paths/footways/tracks on the course as =customers to show it’s for the golfers.
I’d map paths that are fenced off or lead to green keepers machine yard as =private to show you can only be there with clear permission, and expect to be challenged.
That might not directly engage with your point, but I believe an issue could be the OSM definition of access =private is very strict is often wrongly tagged.
There is a double meaning and real division, or lack clarity on the Designated tag
Some link it to “built and designed for”, others a clear legal “designated for”
I fall in the designated tag means there exists a legal “designated for”. Useful for our PRoW;'s
I always map PRoW Bridleways as bicycle=designated, because a PRoW Bridleway is a route which 3 different modes have a legislated designated right to travel along. Those being Foot, Horse, and Pedal Cycle. I consider the maintenance issue, and and especially the Give Way issue not relevant.
I see. I thought I had based my tagging on the tooltip that appears in the OSM online editor. It says:
access=yes - Access allowed by law; a right of way
access=designated - Access allowed according to signs or specific local laws
But I don’t think that’s very helpful actually and if anything makes a case of tagging as yes for all 3 rather than designated. Hence why I’m quite confused.
Can you give an example of when you would use access=yes? Would you simply never use it on a PRoW?
Apart from actively avoiding misleading tagging (e.g. tagging public footpaths as highway=bridleway), the most important thing IMO for public rights of way is to tag designation=.
=yes vs =designated is a bit of a red herring. As this thread demonstrates, =designated is a value that was invented for Continental (specifically DACH) use; there is no consensus on what it means in the UK; and so (as far as I’m aware) no data consumers make any useful distinction. They might in mainland Europe, but not here.
designation=, on the other hand, records a real fact, and enables data consumers to adjust their behaviour based on that (and several do!).
The theory, as I understand it, is that it’s possible that a public footpath might have permissive bike access even in the absence of signs. So unless there is an actual “no cycling” sign, then you don’t know that it’s bicycle=no.
Paths like this do exist: here’s one. (Oxford locals will know that there’s actually an “alternative route for bikes” section slightly higher up which does have signage, but let’s elide that for now.) I suspect they’re fairly rare in rural areas but a little less so in cities. It’s all a bit angels-on-a-pin, but hey, this is OSM and kind of the entire point of the project is counting angels.
But, that said, there’s an easy answer, and that’s to use highway=footway and designation=public_footpath with no bicycle= tag at all. That accurately records the situation and gives data consumers the information they need. Of course, that doesn’t help those who insist on using highway=path, but in that case may God have mercy on your soul.
I’d agree with pretty much all of that, especially:
The one slight caveat is:
That might be correct in cases where something that is legally a public footpath has a landowner that either permits or even encourages horse access - overpass finds a few with explicit permissive tagging, sometimes with more detail, and if you remove the explicit ["horse"="permissive"] from that query you’ll get lots more, including this in @rivenage’s patch. That looks similar because the council PRoW data has it as a public footpath but the name (“Eltofts Bridleway”) suggests also permissive horse access. I guess there may be some signage that confirms on what terms.
On the whole bicycle=no thing, a lot of these tags were added by one user (see previous topics here and here). As noted earlier most of this was to do with preventing people cycling on roadside pavements, or being directed to do so by routers, but there was a bit of a spillover into whether it makes sense to tag bicycle=no on a designation=public_footpath where access would otherwise be assumed to be private and there is no signage. I also wouldn’t add that tag unless it is explicitly signed, and sensible routers shouldn’t send people cycling down a highway=footway anyway.
Actually, the reason I noticed a difference in yes and designated is because cyclosm.org colours bicycle=designated as bold blue, the same as it colours ‘proper’ cycling infrastructure, whereas bicycle=yes is coloured dark green. Hence on their map, one gets the impression that blue lines are going to be cycling infrastructure wheres as green lines are going to be some sort of ‘track’ that you can cycle on.
That’s an interesting point. Again, because I mostly look at cyclosm.org every single path without bicycle=no seemed to show up as a cycling track. Rural drive ways, informal paths across scrubland or ploughed fields. It looked like a kid had scribbled on the map. It was almost impossible to work out where you are actually allowed to cycle. I also found a lot of lines where the highway type was set based on designation, rather than the physical highway type (e.g. highway=bridleway but a survey shows it should be highway=track, designation=public_bridleway).
So it sounds like the sensible thing to do is be a bit more careful with highway= rather than tagging a lot of bicycle=no.
Yes, I need to survey it! I updated the PRoW data, but didn’t want to touch anything else because I have no idea what it looks like on the ground. I’m quickly learning that small focused edits based on surveys are probably the most important contributions I can make.
I, er, wouldn’t use that map then … Looking at some cycleable paths and tracks east of Knaresborough, Cyclosm seems very misleading. It’s understandable that it doesn’t know about the designation tag because it’s clearly not designed for use in England and Wales, but to ignore the other tags here seems bizarre - that is tagged as somewhere that is both legal (bicycle=yes) and not that bad (surface=asphalt; smoothness=intermediate) to ride on.
This is something I remember trying to wrap my head around when starting out editing OSM. I have settled on using *=designated for occasions where the designation=* is also useful to tag (e.g PRoWs, CRoW open access land). As a result, I take the opposite view of this statement (when mapping in England, at least) :
I.e, virtue of being a designation=public_bridleway, access for cyclists is “designated” (in most local authorities) but not necessarily suitable. Suitability can be further described using physical property tags, eg: overgrown=, obstacle=, and others. This allows for informal paths which deviate from the route on the definitive map due to temporary obstacles to be tagged *=yes together with a note=* describing the situation (I’ve gone for “path diverges from designated route” previously).
As for using access=*, I’d try to avoid it entirely when tagging PRoWs as it’s superfluous when the highway=* and designation=* are tagged correctly. See: Access Restrictions, together with UK Access Provisions. This also describes why using bicycle=no on a designation=public_footpath is unneccesary.
Finally, I would echo the above re. cyclosm.org; while potentially useful to identify issues at a glance, remember not to base your tagging decisions on the interpretations of a render designer! This was a helpful point when I first came across it: Tagging for the renderer - counter examples
edit:
Forgot to add another example of this, local to me. Bristol City Council have confirmed by email that cyclists are permitted on this public footpath, and indeed it forms part of several cycling routes. There is no on-the-ground signage which says so, but nor is there signage declaring that bicycle use is only “permissive” as is often the case with the common “this land has not been designated a public right of way” sign.
I believe bicycle access on Public Bridleways is slightly different to the other rights, and you could argue such routes (in general) aren’t really legally designated for cycling. IIRC, it cycling wasn’t originally allowed on Public Bridleways, but then the prohibition on pedal cycle use was removed, with the proviso that there was no obligation on land-owners or highway authorities to make the routes suitable/accessible for bikes.
But there will be quite a few exceptions to this. For example, Forestry England is generally happy for walkers and cyclists to use any tracks over its land, unless explicitly signed otherwise. But you typically don’t have “permissive path” signs everywhere. Much of the land is already Access Land - that automatically gives rights on foot, but not by bicycle. Such routes should be bicycle=permissive in OSM terms. Adding bicycle=no to PRoWs in this case would be wrong.
That was my original understanding, but actually it seems that the law was changed to give bicycles a legal right of way. i.e. it is not the removal of a prohibition but the addition of a right: Countryside Act 1968 So there really is a legal designation for cycling.
Given the good discussion we’ve had here, what is the best course of action for the work I’ve done so far? Should I revert changes where I have set bicycle=no on public footpaths. Or leave as is for now and rely more on surveys in the future?
… Looking at some cycleable paths and tracks