The eastern boundary is completely missing when you look at: osm.org/relation/7237830

I'm not sure about iD, but in JOSM they are reported as errors.

I see these errors because I operate a web site that uses boundary data. When they break, I see that they are broken.

If you are able to get on Slack chat (https://slack.openstreetmap.us/) I would be happy to help you more.

Looks good! I added the missing "outer" boundary relation roles. The boundary for Punda (osm.org/relation/7237830) is still broken.

This boundary also is missing a section: osm.org/relation/7237830

I trust that you will choose the correct location of the boundary! As long as it forms a complete loop.

Hi, if you load osm.org/relation/9507096 into a web browser, you should see that a piece of the boundary is missing.

Hi, it looks like the boundary relation for Berg Altena (9507096) was left in a broken state?

Hi, this relation appears to violate the guidelines at osm.wiki/Relations/Relations_are_not_Categories

Hello, in this change, you had the place node assigned to a role of "outer" rather than "label" or "admin_centre". I've changed it to "admin_centre" but feel free to change that if not correct.

Okay, should be all set for Rhode Island town/city boundaries!

Are you local?

Yes, I see what you are describing. admin_level is clearly not required on these boundary relation ways. Fortunately this is an easy fix.

Let me investigate and get back to you. Thanks for alerting me.

This looks great! Keep up the good work.

Hello, yes, on further review, my edit was probably a step too far as protect_class=22 "cultural area" as probably a better characterization of these sites versus merely "historic". I have restored the original tagging and apologies for the churn.

Yes, I understand your preference for protect_class=24 in this case.

My bad, I researched this area on Wikipedia and it seemed like aboriginal_lands alone was a sufficient fit (which is also supported by renderers). I understand if you wish to revert the change.

I attached the place nodes that had the same name as the surrounding boundary, ignoring subdivisions that didn't have a place node with a matching name. If that was not a correct assumption, my apologies and this changeset can easily be reverted.

Hi, it looks like this change broke the town border at osm.org/relation/9157671

I have added this information.

Neat... but why not just "name" rather than "official_name"?

No worries, thanks.