This user marks a lot of public rights of way as access=no. Presumably the logic is that this makes all the access tags default to access=no and then the positive access tags (I.E. foot=yes) override that.
1) Is this a valid method of tagging public rights of way?
2) Won’t tagging in this manner cause problems. For example, even if routers may still route along the way, public rights of way will be incorrectly displayed on the map showing access is not permitted, when access is permitted – as it’s a legal right of way.
For example, the below public footpath now displays as greyed out due to the change to access=no, meaning anyone looking at the map will think it’s private and access is not permitted.
3) Other changes seem odd too. For example, the access tags of the below way have been changed to bicycle=private and horse=private , when it’s clearly a public bridleway. It’s not physically blocked as I’ve been along it before, there’s no note=* tag indicating the reason for the change and GCC still show it as a public bridleway.
If it’s being used on a footway, cycleway, bridleway or path which is in use, it’s troll tagging and should be removed. Restricting other transport modes is unnecessary if the correct highway=* type is used.
For private tracks and service roads which are also PRoWs, these might well be access=private + foot=designated. It’s far too common for access=no to be misused as if it were a synonym for access=private.
(about access=no, on a private track that is also designated as a public footpath)
Not necessarily? The wiki page suggests “stronger interdiction than private”, which I’d agree with, but it’s perfectly OK (and described as such on the wiki page) for something to be access=no and e.g. foot=yes. It’d certainly be much more likely for something to be access=private and foot=yes though. However, I can imagine an example where “if you are delivering a parcel, we (the owners of the property) want you to use XYZ which is access=private rather than ABC which is access=no”, and where ABC is also a designated public footpath.
That’s simply not true in England and Wales. Something that would be mapped in OSM as a highway=track might have access for pretty much any transport mode that is basically anything from “not allowed” through “permissive” all the way to “public right of way” Where something is designated as something (public_footpath, byway_open_to_all_traffic, unclassified_highway etc.) then a set of minimum access rights can be derived (but, for example, a designation=public_footpath can also be horse=permissive if that is also allowed).
Where something isn’t designated as anything signage may help, but these cases might still be anything from “no access” up to “completely public access” (as an example, this was “not on any government list but everyone knew it was public” until a group of horsey locals got it properly designated as a bridleway a while ago).
I was only referring to PRoWs mapped as highway types with implicit access restrictions (i.e. footway/cycleway/bridleway) where adding access=no and then restoring the implicit access with foot=yes is pointlessly redundant troll tagging. The second paragraph of my reply covered (inadequately) other highway types where access=private together with other access tags is likely to be correct.
I had a quick look at a sample changeset by @12pjw34 and noticed that a public bridleway over what appears to be a farm track had access=private added (presumably correctly), but also horse=private and bicycle=private (unlikely). Adding access=* to PRoWs may not be the biggest problem with this mapper.