@Nakaner

I agree entirely, I was not trying to argue against what has been said here. I was only trying to explain I made my edit in good faith that it was supported with evidence I saw. I was wrong, and created trouble for people. I should have been more thorough when making the edit.

We try to be careful as possible editing in Europe.

Sorry, I saw the Mapillary street level imagery and thought that would be reliable evidence to make this edit on. It said 2017, i thought it would be recent enough, I was wrong. Thank you for the link about the construction, sorry for the trouble.

https://www.mapillary.com/app/?pKey=ZgwiAYkD3J59n9xSHJOPxQ&focus=photo&lat=49.0057953284537&lng=8.409497403738328&z=17

abrensch, my apologies, I must not have been zoomed in enough to notice that segment (way/98309825) was in fact connected to (way/250807036). I thought it was connected to (way/250807031#map=19/49.00563/8.41037). I did not notice I was creating a hanging segment, it looked like it was attached to the rest of the network at Kriegsstraße. I corrected the segment in changeset: changeset/63593044#map=18/49.00558/8.40691. Sorry for any trouble this segment caused you, I meant no harm, i overlooked a small detail, i'll be more careful.

Sorry, I miss understood the use of that road attribution, thank you for following up with my edit.

Agreed, I wish I had some, OSM has the GPS traces layer but it is not clear enough to reflect the massive development, ton of buildings added.

http://www2.uwe.ac.uk/services/Marketing/students/pdf/Accommodation%20information/Wallscourt_Park.pdf

ndm, thank you for the follow up, when the gates were added, the evidence used for adding them was street level imagery: old gate - (https://www.bing.com/maps?cp=51.447401~-2.592546&lvl=17&dir=104.101&style=x&v=2&sV=1) you can see some type of barrier if you move over a few frames you can see the second old gate. according to the meta data on the imagery it is 2012, as you mentioned, likely out of date.Thank you for following up with the edit. It was made in good faith, but the imagery was not recent enough to accurately map the area. I appreciate you correcting the area.

ndm, I agree that "Track" suits this case very well, however I was referencing this discussion https://github.com/openstreetmap/iD/issues/3038 where "unmaintained track road" is suppose to replace the "track" tag. Other similar roads (way/148760936#map=19/51.45730/-2.77046) in the area are attributed as "unmaintained track road". I was only trying to be consistent with what I was noticing. I'm partial to making it "unmaintained track road", but if you feel like "track" will suffice, i'll change it to "track". Thank you for your help, I appreciate the time you took to help me out with this edit.

Hello ndm,

The newly added service roads with gates is a more accurate representation than the parking aisles that were drawn over the parking spots were. However, in doing the edit in unintentionally altered one of the building polygons that were attached to the deleted parking lot as seen in https://osmcha.mapbox.com/changesets/62057092/. The building has been restored to its correct shape since.

Thank you for input, please respond back if you have any suggestions.

Hello trigpoint,

Agreed, marking this added segment as a parking aisle is more accurate, it should be updated.
In terms of adding access tags, these features fall within: way/75845092#map=18/52.79177/-1.57295, which looks like a business park that is public, this doesn't look like an enclosed facility. The businesses here may be expecting traffic, making this road segment motor_vehicle=private would be the most appropriate.

Please respond back if you have suggestions on what access tags you think are missing.

Hello ndm,

Yes, according the the intended use of "Unmaintained Track Road" as discussed in https://github.com/openstreetmap/iD/issues/3038, and in, https://github.com/openstreetmap/iD/issues/4092. This should be an unmaintained track road since it looks like this road leads to farm. The road from the satellite imagery shows this road to be unpaved as well. unmaintained track road would be appropriate according to osm community norms.

Please respond back if you have suggestions on what you think the road should be classified as.

Yeah, sorry for the map noise, they should be back to how they were before I made the changes. I acted too soon before doing all the research. Thanks for the follow up.

Toni,

-Wow, what an explanation. Thanks, so clear now, I see what happened. I see why it is tagged the way so, I will revert back my changes. I think we need to look at how we treat motor_vehicle=private tag internally.

-For more context, when we make deliveries, if a route is motor_vehicle=private, we do not want to drop this road from the map, sometimes someone orders a package and the only way to their facility is through a private route, we will route them through it if it is the only way to their facility.

-We drop a road if it is access=no. You're right, 265849909 is successfully dropped from the map and drivers will not be routed to there. but 265849859, 265849958, and 265849921 were still part of our map (they had access=yes). What this looks like is half a round about. You have a round about where the only way in is through an out bound one way road. There is no way into the round about because 265849909 is not considered. This confuses our routing engine.

-This is why motor_vehicle=private was not enough, and why I made 265849859, 265849958, and 265849921 access=no. So that all segments will be not considered for routing. Better to have no roundabout than half and confuse the routing. There is a round about and no way to get to it, it loses its mind.

-Solution: problem needs to be fixed on our end, osm should go back to the way it was before i changed it.

Toni, this has been a pleasure, i'm grateful for you taking the time to explain the situation. Let me know if you'd like to change it back or don't mind if i do it. Please let me know your thoughts.

Best
Sam

Hello Toni,

I was following the openstreemap wiki:

access=no

"a road tagged access=no and psv=yes generally withdraws any legal right of way but permits public service vehicles. A routing program in psv mode may use this way."

The Navi correctly avoids Schindlerplatz becuase it has "access=no and psv=yes", your tagging. But the roundabout is still detected because it lacks either of these two.

"The access=no tag indicates that the object is not to be used by the general public, with stronger interdiction than the access=private tag."

Which is why i added access=no to all the round about segments, for our navi this enough to make it so routing does not try to navigate through here, or detect the roundabout.

Please let me know what you think, i'd really like your opinion. It could be nice to have round about as "access=no and psv=yes", but if this breaks something you have, then i'll change everything back.

Question: in Germany is the moto_vehicle=private a widely used tag?

Thanks for your time
Sam