oneway=yes + lanes=1 + parking:right=lane means you can get through. oneway=yes + lanes=1 + parking:right=traffic_lane means you can’t and would be invalid.
We must just assume we’re talking regular cars/trucks here, since lanes are designed built to a width for these vehicles. The rules don’t really apply to bicycles who can fit past parked cars in the same lane, at the risk of being doored, so still usually safer to exit the lane like cars would have to.
There is no parking:<side>=traffic_lane yet. I’m arguing about the current meaning of parking:<side>=lane to always mean traffic lane.
As seen on this picture:
parking=lane takes away space for cars to drive. That’s how its defined. Whether there’s enough left for another car to drive on that same lane, is a matter of width.
So which one is it? When parking is allowed in a lane some of the time and then it blocks the lane for through traffic, but driving in it is allowed and possible when there are no parked cars - does that lane count in lanes or no? Example photos of such “4 lane” streets have been posted by Andrew and me in this thread. @Nadjita @Supaplex030 can you discuss directly?
It came to my mind, that the documented staggered key is already in use for situations with “loose, scattered parking” like the examples we are discussing here. E.g. StreetComplete is suggesting it for single sides as well. So we already have a tagging that is able to decribe the situation in a precise and evaluable way:
lanes=4
parking:both=lane
parking:both:orientation=parallel
parking:both:staggered=yes
Regarding the lane count and taking this tagging into account, I think it is indeed valid to assume 4 instead of 2 lanes, as “staggered” parking allows to evaluate the limited traffic capacity of one lane per direction.
(P.S. I reworked the paragraph in the wiki created by @aharvey, which imho wasn’t consensual, as the street parking scheme is a widely discussed and well-established scheme into which one should not simply insert a new main value without broad agreement)
No staggered is documented as “On some streets, parking is theoretically possible on both sides, but they are too narrow so that in practice parking is only possible on one side without obstructing traffic.” That’s completely different to my example, where you can fit as many cars as you like along both outer lanes as traffic can still flow in the two inner lanes unobstructed.
To be honest you’ve done the exact same thing, suggesting this should be staggered without a consensus or discussion.
I added these examples to the wiki since I feel they aren’t well documented, and at the least the wiki should acknowledge that it’s disputed what to do in that scenario.
In my view what you’ve changed the wiki to is more misleading and incorrect, my examples were to highlight the confusion between lanes=* and parking=lane since sometimes the parking is inside the lane count and other times outside. Instead now we somehow mixed in staggered which this is not.
Sorry, I didn’t express myself precisely enough. I did not mean that this use of staggered is documented, but that the key staggered is de facto already used in this way (“in use”). Especially via StreetComplete and was discussed in this way during development of the quest. Unfortunately, this use has not yet found its way into the Wiki documentation.
Therefore, I would say that I have rather documented this in use tagging (but not “invented” any new values).
However, there is also the question of how parking lanes are related to traffic lanes and @Nadjita has started another thread and poll on this issue.
Yes, and StreetComplete labels it “on street on alternating sides”. I think how it’s currently at https://wiki.openstreetmap.org/wiki/Street_parking#Staggered_parking_on_narrow_roads is good and makes sense.
Unfortunately my example is definitely not “staggered”. I could go back another time and find the parking full on both sides, not alternating. It’s not an example of “Loose, scattered parking impeding the traffic flow”, it’s an example of parking within a traffic lane included in the lanes=* count.
