Thanks for your reply. I’m very doubtful, and have removed it in changeset/173536594. It doesn’t fit with any of the local naming etymology, and nobody on https://ukcaving.com/board/index.php?threads/a-gragareth-mystery.32857/page-2 had any idea either (and those people like figuring out names).

Are you sure that’s a real name and not a deliberate mistake in the OS map data by OS? I’ve never heard this area called by that name (I am local).

Fixed in changeset/173387985 using Bing Streetside imagery and the parks’ websites.

No answer (to any of my comments on your changesets) ⇒ reverted as changeset/173384993

Happy to discuss this further if you’d like to reply :)

Fixed various unlikely paths copied from Strava Heatmap in the following changesets:

changeset/173377845
changeset/173377985
changeset/173378293

If you disagree with these changes I’m happy to discuss further :)

Reverted as changeset/173377461 due to no reply (to any of my comments on your changesets)

Filing an issue on GitHub is a good idea. I believe the StreetComplete developers are quite responsive, so hopefully it should give good results.

Knowing every house and having enough friends round the village to ask them their postcodes sounds like a good approach and shouldn’t be legally problematic at all.

The problems potentially arise when people mass-copy postcodes into OSM, or state that they’ve copied postcodes from the Royal Mail database. At that point, if the Royal Mail were to notice they’d have good grounds for asserting their database rights. At best, that would mean volunteer time gets consumed in erasing the unlicensed data from OSM. I believe they have taken action against people other than OSM copying their database before (they make money by licensing it to businesses), so it’s not a baseless threat. It is unlikely though.

---

Thanks for testing StreetComplete. It’s a shame it isn’t fully compatible with the format from the wiki.

I think my inclination would still be to use a combination of the first and third formats I suggested. That would be compatible with the wiki (most of its advice around addr:place is to avoid using addr:place and addr:street together), and would keep StreetComplete happy.

While you may remember to not answer the street/place StreetComplete quests, other users in future will not, so going with addr:city with no addr:street is likely to get messed up the next time a well-meaning user who hasn’t read this discussion does some StreetCompleting in Casterton.

Agreed that adding all the addresses in Casterton is a great goal, and should help with deliveries. Please note that the data in the Royal Mail address database is not licensed for use in OpenStreetMap, so you unfortunately cannot (and must not) use it to fill OSM in.

There’s some information about this on osm.wiki/Postal_codes_in_the_United_Kingdom

The best information which we _can_ use is the postcode centroids data, which is available in the ‘Postcode centroids’ layer in ID. It’s what Royal Mail have chosen to release freely. For cul-de-sacs and short residential roads, it’s fairly easy to associate the centroid with the street. For longer streets, or where the postcodes differ for odd/even numbers, it’s a bit harder.

You might find that if you use the format you specified, but omit addr:street (for the houses with no associated street name), then StreetComplete will start to ask you about the street names for the houses again. That’s where my confusion before arose. StreetComplete expects either addr:street and addr:city to be set, or no addr:street and addr:place instead.

This (and the fact that post towns are outdated) is why I suggest going with a combination of the first and third tagging schemes I listed (the first for addresses which have a street, the third for the others).

You could try your tagging scheme and see if StreetComplete starts to ask about street names again, though, if you wanted.

---

I (and other editors who’ve been around a while) check all the edits in the north west of England. Mostly it’s to prevent vandalism, but also helps with catching tagging mistakes or typos while the edit is still fresh. With relatively new users (such as yourself) it also means we get a chance to provide some guidance on some of the trickier aspects of editing OSM, like this! :)

Heya, I think there’s some confusion here.

There are two ways to tag UK addresses in OSM. The most common way is to put the settlement (Casterton) in addr:city and the street (Fern Croft) in addr:street.

Another way is to add the post town, although aiui Royal Mail doesn’t really use these since 2004, and they are irrelevant to most people. There is, however, an accepted tagging scheme for them, which would be addr:city=Carnforth addr:suburb=Casterton addr:street=Fern Croft.

See osm.wiki/Addresses_in_the_United_Kingdom

There is a third tagging scheme for addresses which _don’t_ have a street name (which applies to many of the addresses in Casterton). I had to brush up on this one, hence me incorrectly undoing some of your addressing changes along those lines a few weeks ago. Sorry about that. If an address doesn’t have a street name, use addr:place=Casterton, and no addr:street or addr:city tags.

Putting a comma in addr:street is not one of the valid tagging schemes.

I hope that makes sense. I realise it’s a lot to read (sorry!) but addresses are inherently complex, and the tagging scheme on OSM needs to work globally.

Happy to discuss this further if you’ve got any questions :)

You also appear to have removed footpaths from OSM which appear on the Public Rights of Way overlay as public footpaths. For example: way/540698481

These should be mapped on OSM, even if they don’t appear on Strava Heatmap as they don’t get much use.

Please fix this, and re-check your recent changesets for it. You seem to be rushing through changes and creating a lot of work for other people. Thanks.

Hiya, did you mean for way/1435505356 to be a wall connecting to the middle of the road?

Why have you removed a lot of the public footpath network around North Stainmore/Argill? The Public Rights of Way overlay shows them as public footpaths, so even if they’re not regularly used (and don’t show up on Strava Heatmap), then should be on OSM.

Are you sure things like way/1441615066 are actual paths, and not just one Strava user’s dog walking route?

Can you please split your changesets up. This one also spans most of northern England. Big changesets spanning a large area are hard for other volunteers to review.

See osm.wiki/Changeset#Geographical_size_of_changesets

Please also reply to comments on your changesets, or people will lose trust in your work.

Thanks.

Hiya, are you sure this is right? https://www.rosegrovecaravanpark.com/ makes it look like Rose Grove still exists, and https://bridlingtoncaravans.com/sandy-bay-caravan-park.html has a different address for Sandy Bay.

Looking at the aerial imagery, it looks like way/182054674 could actually be two caravan parks. Is one of them actually Rose Grove or Sandy Bay?

Where are you getting your information from? You should always add a source to your changesets as it helps people check each others’ work. See source=*

Thanks

Hiya, what source are you using? In its previous location it was on the high point according to OpenTopoMap.

Thanks :)

Hiya, thanks for continuing to try and sort this out. I don’t think the changes in this changeset are correct from a tagging point of view. The tagging as it was before looked correct.

Possibly the problem now is the apps people are using to navigate, or their interpretation of the map. Unfortunately neither of those can be sorted by editing OpenStreetMap! Your next step would be to find out which apps people are using which are telling them to take this path, and get in touch with the app authors to say they’re giving incorrect directions.

Hiya, this edit spans most of the north of England, making it very hard for others to review.

Please be mindful of this and submit changes in smaller, more geographically focussed changesets in future. This should probably have been 7 different changesets.

See also osm.wiki/Changeset#Geographical_size_of_changesets

Thanks :)