Excellent point. Not sure why I tagged it like that

Any reason to use turn restrictions instead of making the road a one-way?

I copied these shapes from a chapel hill open data kml file and thought I'd removed the gis tags from them all but clearly missed those two. Thanks for fixing them!

OK, I see what you're doing, and I agree with it. I had experiences in the past with people deleting 10s of miles of abandoned railroads that were still obvious, and thought that was what was going on here. The segments you are removing are obviously completely gone, and should be removed.

You can't tell if these are gone without also looking at terrain data to see if any trace of the changed terrain exists. Abandoned railroads are very useful to have in OSM, even if the tracks are removed because it's not about the tracks, but about the corridor that still exists and still feels like a railroad corridor. I use this data a lot and don't like it getting deleted for no apparent reason.

An example of this: splitting osm.org/go/ZS_ONQvzL into two forests requires splitting the left way twice, removing the middle bit from the woods relation, deleting it, connecting the lower left piece up to the grass, then splitting the grass twice, removing the middle bit of the grass from the woods relation, removing the upper part from the woods relation, connecting the upper part back to the grass, creating a new relation for the new wood area, and adding all of the tags needed.

If you had done this mapping without multipolygons, I could have done it in about 10 seconds by creating some new nodes along the grass, using alt-x (josm) to split the woods into 3 areas, and deleted the middle one.

Because of how you mapped the area, adding that gap takes me 10 times longer. Every little imperfection is a huge pain to fix.

Why does osm.org/relation/13588757 not contain the clearings inside of it?
Also, why does osm.org/relation/13613356 contain the pond as an outer way instead of an inner way?
Why is osm.org/edit?relation=13613356#map=20/35.95392/-79.15280 inside of a forest polygon when it's not forested?
Why does this forest polygon contain a ton of areas that aren't forested? osm.org/relation/13589950

There are tons of problems with the data you're uploading that mostly stem from the way you're creating it. If you: 1) Only create a multipolygon for an area if it would have inner ways, 2) Only include a single outer way in each landcover polygon, and 3) Keep polygons smaller, then these issues would be much easier to fix. As it is, I have to untangle a web of multipolygons before I can do anything to fix issues like this: osm.org/#map=19/35.95070/-79.15885

Thanks

I love the way the new woods look on the map, but trying to maintain/fix up woods is extremely difficult due to the way you've mapped them. Could you try to avoid using so many multipolygons in the future, and just map woods as closed ways? Multipolygons are much more difficult to maintain than closed ways, and are much more error prone/breakable, so they are not worth reducing overlapping ways for.
Thanks!

I'm like 99% sure this is wrong. Would you mind me deleting this node?
I've lived near here for 15 years, but haven't ever heard anyone call this area Bolin.

It's a government owned parcel that isn't protected from development in the future.

Hmm... usually access=private is only used to mark that it's private access, not private ownership. If there are no trespassing signs, or the general public is not allowed to use a road, add access=private. Otherwise, I'd avoid the tag. Maybe try ownership=private?

Oh, no, that's definitely a mistake. I think it is supposed to be 4:(3) to indicate a merge. I'll fix this. Thanks for working on tools to find these sort of things! Also, thanks for letting me know about this case.
Leif

Det var et fejl. Jeg har slettede græsområdet nu.

Det må har været en fejl. Kan du linke til græsområdet du har set?

Ja, men online scriptet kan ikke klare store changesets. Hvis der er tusindvis af ændringer, virker den ikke.

Nu er stien tilbage

Det har du faktisk ret i. Jeg prøver at "revert" den her changeset hvis jeg kan uden JOSM.

Mapbox updates every 4-6 months usually. Some of its maps update every year. I'm hoping they improve this soon.

Nej, men normalt skiller jeg kun veje hvor man ikke længere kan skifte mellem vejene hvor man kører (der er ikke noget physikelt mellem banerne).
Mere accurate data kan tilføjes hvis man kun skiller veje hvor det er umuligt at køre fra en til den anden, som for eksempel turn:lanes som kan bruges af OsmAnd og Magic Earth til at hjelpe folk med a vide hvilken bane de skal bruge.
Tak for kommentaren!
Leif

Thanks for your contributions! Another other local mapper and I are currently doing a public transport import from all GTFS sources in the Triangle area, and finished the bus stops a few days ago. We will be moving on to adding missing routes soon. We should have all routes added by the end of July with PTv2.
We are currently working on cleaning up existing routes, so this changeset is highly appreciated!