|
Have had a read through previous posts but haven't found specific mention of what is best practice in this particular instance. Round my way there are numerous instances of clearly private dead end residential roads off "public" larger roads that cannot be openly accessed due to a locked gate to everybody - cars or on foot etc etc. Clearly the section of the road behind the gate is access=private and same for the access on the gate. However, as there nearly always has to be a very short "turn in" bit of road to approach the gate, should that very small section of road be marked access=yes as even if the road is "private" as clearly anybody can "access" that bit. I had previously been marking that whole dead end as private but have recently seen several examples where this tiny approach bit has access=yes before the gate. Part of my confusion is the blurred line between a road being private in that not maintained at the public purse and sections of road that have no "barrier" to access even if you are not meant to be there unless you have specific business - like delivery or visiting somebody - in which case you can legitimately ask for the gate to be opened. Generally I feel that marking maybe a few metres of a "private" road up to a gate as access=yes seems overkill and not sure correct anyway. If there was no gate (or indeed always open), the road would still be "private" even if then "accessible" without asking. One such example of this split is Lovekyn Close in Kingston upon Thames, London, UK https://www.openstreetmap.org/way/22494916 and how looks on the ground https://goo.gl/maps/tiiYzparoHAsa6ak9 So the access=yes bit is only a few metres with double yellow lines on both sides so you could not "park" there anyway unless in an emergency of course. Any thoughts most appreciated |
|
As always in OSM: You can keep it simple reducing the reality to a meaningful abstraction, or you can map as many details you can find. Personally, I would not bother with splitting the way into a private and public part. I see limited value in spending the effort and the limited benefits that would create. But there is nothing wrong with splitting. So feel free to do it if you feel like it. |
|
In the Lovekyn Close example: There are sidewalks along QE Road. They're currently not mapped as separate highways, but imagine that they are. Now imagine a bicycle starting from the house https://www.openstreetmap.org/way/103010689. A routing engine, in bicycle mode, might choose to begin navigation on the sidewalk pavement (the closest highway to the house). If Lovekyn Close has public access, that bicycle can immediately merge onto QE. If Lovekyn Close is private, that bicycle will have to remain on the sidewalk pavement until it reaches the intersection at the end of the road. Several hypotheticals here, and very minor difference navigationally, but this is an example of why moving the private tag behind the gate might be a meaningful choice. ...But mappers don't always need such a meaningful reason. Sometimes they'll just want to add the maximum amount of detail. Mappers are generally advised to follow local tagging standards in these matters, but local standards do evolve with time. Personally, I'd probably just tag the whole thing private (and even moreso with the Pyms Stables example.) But I wouldn't be inclined to unsplit the highway and tag it all private, once it's been done. |
Here is another more obviously "private" one I came across today. A rather large locked gate! https://goo.gl/maps/ij7GuE3sK25NtDCY7
This road is actually marked access=yes with no gate on OSM so clearly needs correcting.
As you can see the very short "approach" from the main road is "accessible".
They are often (ab)used for making easy three-point turns, so this is "helpful". Legitimate uses would include dropping someone/something off without going through security, or hailing/alighting a taxi conveniently.