I’ve reverted this change as changeset/121595585, and made a further change as changeset/121595616 to ensure the road is reachable — I assume this was causing the warning you were originally trying to fix.

You can use http://revert.osmz.ru/ to revert any changeset — just input the changeset number and a message listing the changeset number and why it’s being reverted.

Once it’s reverted (the tool normally takes a couple of minutes), it sounds like that byway will need a barrier node adding just south of the give way. You’ll probably want to tag it with:
barrier=bollard
motor_vehicle=no
bicycle=yes
foot=yes

or something like that, depending on whether it’s actually a bollard or some other kind of barrier :)

Heya. Are you sure about this one? The satellite imagery shows dashed white lines at the end of the service roads, which are legally equivalent to a ‘give way’ signpost.

The main reason for putting give way nodes into OSM is because they allow routing software to know where the priority is at each road junction. This allows for correct “turn left in 100m” style navigation announcements.

Hi, welcome to OpenStreetMap and thanks for your edits!

It looks like the MBP are doing some coordinated editing on OSM? I’ve seen some local edits from a Jo@MBP too.

If so, that’s great! Please message me if there’s anything I can help with (I’m local).

No problem, there wasn’t much wrong and it’s a learning process!

If you haven’t found it already, if you hold Alt while dragging a node around, it won’t automatically snap to other lines or nodes.

Similarly if you press Q while a building is selected, the editor will automatically square the building’s corners, which can help sometimes (although perhaps not in Orton since a lot of the buildings aren’t really square!)

It’s not an entirely clear-cut issue, but given recent discussions on the mailing list (https://lists.openstreetmap.org/pipermail/talk-gb/2022-February/028604.html, but it’s a long thread), I think the consensus is to map things like this using addr:parentstreet.

You can find some more documentation about it here: osm.wiki/Addresses_in_the_United_Kingdom#addr:substreet,_addr:street_&_addr:parentstreet

The gist of the argument is that the house name isn’t “1 Beech View” because that’s a number and a name. It’s not “Beech View” because then two buildings would exist with the same house name. So it has to be treated as a kind of street. But there’s no road called “Beech View”, so addr:parentstreet needs to be used to say which road the buildings are on.

I’ve taken a look, and it looks great to me. Thanks for taking the time to update it :)

(For anyone else reading this afterwards, the followup changeset is changeset/121556807)

Great! Before you start moving a lot of buildings around, you may also want to turn on the OSMUK Cadastral Parcels layer in the Background Settings of the editor.

It shows cadastral parcels from the Land Registry, and is the only source of ground truth for making sure that satellite imagery is properly aligned before drawing things against it. Unfortunately satellite imagery varies in its offsets by up to 5m across the country (and between updates to the imagery), which means that drawing buildings accurately against it doesn’t necessarily mean they’re in the right place!

I think the imagery offsets in Orton for Bing aerial imagery against the Cadastral Parcels are 0.55,-0.87.

See https://osmuk.org/cadastral-parcels/ for a bit more information

Sorry for the dump of information — but I thought it best to let you know about this before you move a number of buildings :)

Hi, welcome to OpenStreetMap, and thanks for your recent edits around Orton. It looks like you’re adding a lot of useful local knowledge :)

One pointer: if houses are joined together in a terrace in real life, you should join the nodes of the houses together in OpenStreetMap so that their areas are joined. This correctly represents the terrace, and allows consumers of OpenStreetMap data to infer that the houses are terraced and which wall is the party wall.

So to give a local example, the southmost nodes of Arthura Cottage should be joined to the northmost nodes of Stone End, and the southmost nodes of Stone End should be joined to the northmost nodes of Westmorland House. (Assuming I’ve remembered correctly that those three houses are terraced, from when I last visited.)

Thanks! If you have any questions about OpenStreetMap things, just drop me (or any other local mapper) a message :)

OK, no problem. I’ve gone ahead and restored the track in changeset/121536251, and then connected it to the main road and fixed its access tagging in changeset/121536369

So I think this is all resolved now. Thanks for your help :)

Ah, I didn’t realise the tagging came from a way which was already there. What you did makes sense in that context. Thanks for the clarification :)

I’ve removed the golfcart tagging in changeset/121536141, so I think this is all resolved now

Thanks for your help! :)

And Portsmouth too by the look of it.

Please try and keep changesets limited to one city or geographical area, as it helps others review edits.

See osm.wiki/Changeset#Geographical_size_of_changesets

Thanks!

Hi, why have you changed this to a footway? From the satellite imagery it’s almost certainly a service road so that people can access the caravans. I can see a car parked half way round it.

Hi, I don’t think removing the tagging entirely was the correct thing to do here. It looks like this area is for tagging the highway extent, so it should be tagged as area:highway=yes. I’ve fixed that in changeset/121416900

Hi, why did you tag this as a golf cart path when it’s not on a golf course?

Hi, why have you removed this? The satellite imagery shows a track is present. Even if it’s faint higher up, it looks a lot like a typical green lane walled track near the A595.

Although the thing about the Ruskin library in particular might just be a parallax error on the satellite imagery: often the imagery is actually from planes or drones, and parallax is noticeable on taller buildings.

Regarding your other tweets about building positions on OSM recently, are you aware of the OSMUK Cadastral Parcels layer? If you do, please ignore the rest of this comment :)

If not, see https://osmuk.org/cadastral-parcels/ before you think about starting to realign buildings.

Basically the satellite imagery is not guaranteed to be perfectly aligned to any source of ground truth. In particular, the alignment can change when the imagery is updated, or across different parts of the country, by up to about 5m.

Previously there wasn’t a solution for this in OSM apart from taking accurate GPS traces and aligning to those. Now there’s the OSMUK Cadastral Parcels layer, which uses surveyed property boundary data from the government as a source of ground truth.

So before you start realigning buildings, please enable the cadastral parcels layer and align your satellite imagery to it.

That’s a little tricky in the uni as it’s one big land parcel, but it’s possible to align using the houses on Five Ashes Lane. It’s generally best to check alignment in a few places, as not all parcels are perfectly surveyed.

Sorry for the long-winded comment (which you probably knew all about already). I just wanted to double-check you were aware of it before you potentially embarked on a big spree of realigning buildings! :)

I’ve made part of it a bridge in changeset/121388109. Does that look more correct? Feel free to edit it further if I’ve misinterpreted the photo :)

Do the steps cross over the top of each other (and the footpath) with a bridge?