I'm against splitting the way into two diverging ways. The usual argument for splitting due to physical separation makes sense for longer separations as it affects routing which doesn't apply here.
Splitting makes the data more complicated than it needs to be, and doesn't add more value or accuracy compared to simply tagging the traffic island as a node. One with a gap for pedestrians gets tagged as crossing:island=yes, without a crossing maybe traffic_calming=island, or some other tag. Lane tagging becomes more difficult with split ways as the way geometry shifts from the center line to the middle of the lane, so placement tagging is more complicated. Interesting to see those city stats Andrew. What I'd like to hear is from those who do split, is why? Is it just because you're trying to follow the documented rules, or is there a reason for splitting being better? Ideally we'd document the community preferred approach along with the reasons for. On Mon, 15 Nov 2021 at 18:22, Dian Ågesson <[email protected]> wrote: > Hello, > > Quick question, as I'm not sure that there is an established consensus in > Australia for this. > > Where a way leading to a roundabout has a small traffic island, what is > the preferred way to map? I have seen both the "traffic island as a node" > approach (because they aren't really separate carriageways) and the > "splitting ways" approach (because physical separation and more "detailed). > > Specific examples: https://www.openstreetmap.org/way/121203404/history or > https://www.openstreetmap.org/way/603989993 > > Is there a preferred approach, or does it not really matter? If splitting > ways, are u-turns restrictions required? > > > Dian > _______________________________________________ > Talk-au mailing list > [email protected] > https://lists.openstreetmap.org/listinfo/talk-au >
_______________________________________________ Talk-au mailing list [email protected] https://lists.openstreetmap.org/listinfo/talk-au
