Hi,

"Pinfold Close" was from survey. It was added using Vespucci so probably input as I was looking at it on the ground. I pass by that location regularly, so when I get a chance, I'll check it again. I'm fairly sure the name is there, but perhaps it doesn't refer to the house.

I completely agree it should not have been moved to the name tag. I see the same user has moved lots of other building_name tags to name. I've often used building_name for names inscribed on houses, when they are unlikely to be used as part of the address. As these names are sometimes of historical interest, I thought they were worth recording, but it's hard to decide the best tag to use.

Will

I stayed at this hotel at the weekend and noticed it wasn't tagged properly. I didn't realise the removal of the tags was part of a wider problem. Glad you're fixing it - thanks.

Definitely a mistake. Looking at Bing Streetside photos, the layby alongside is asphalt, so I probably intended to tag that.

Hi,

Is the road through Calke village to the Severn Trent car park really only accessible to customers? When I've travelled that way I've assumed the road is probably an adopted highway or at least permissive access.

You have also added horse=no which seems unlikely because it provides the only access to a public bridleway. Therefore I wonder whether the access tagging is a mistake.

This is the road I am referring to:
osm.org/way/25786854/

Thanks,
Will (will_p)

Caribbean Plate has closed in the last week or so. Note in window saying they are planning to reopen in Sandiacre. At present the Caribbean Plate sign is still up and no indication of Hotdog Haus opening imminently.

I don't want to remove anything myself. I will send a message to the Data Working Group and they can decide what action, if any, needs to be taken.

Hi Alex,

I think I see where you are getting the address data from now, but it raises a lot of questions. In particular, I can't see how the address search on the street register page could be suitably licensed for use in OSM.

This is the page I am looking at: https://geoserver.nottinghamcity.gov.uk/streetregister/
Before you can access that page you must agree to some restrictive terms and conditions. Furthermore there is no mention made of any data released under an open licence. The terms and conditions state you are only granted a revocable licence to view the data for non-commercial purposes. Next it says copying or otherwise making the data available is not permitted at all. It indicates that some of the data comes from the Ordnance Survey under the Public Sector Geospatial Agreement.

Some of the layers shown on that page have been released elsewhere under the Open Government Licence, specifically:
Adopted Highways: https://www.opendatanottingham.org.uk/dataset.aspx?id=177
Public Rights of Way: https://www.opendatanottingham.org.uk/dataset.aspx?id=74
The addresses, however, must surely come from a closed dataset. This is particularly clear with the addresses in Arnold. The city council is not responsible for addresses in Arnold. Therefore the addresses must come from a third party, perhaps Addressbase, which is mentioned on the terms and conditions page.

It is also possible you got some of the addresses from the background imagery, which appears to be derived from Ordnance Survey Mastermap or a similar product. Again this is not suitable for use in OSM.

Am I missing something? What leads you to think the address data on that page can be copied into OSM?

Thanks,
Will

Hi Alex,

In this changeset you have updated several shops in Arnold and added addresses to shops already mapped. Please could you confirm the source used. It doesn't appear on be from survey and the NCC Street Register mentioned in your changeset sources only relates to adopted highways in the city council area, so does not appear relevant.

Thanks,
Will

I don't recall my reasoning for this change. The changeset was made using Vespucci so would have most probably been done from observation on the ground. I assume the Smart branding was most prominent and so I added that as the name. Mercedes-Benz remained in the brand tag. Perhaps the name should be 'Smart / Mercedes-Benz of Nottingham' instead?

I don't remember changing the name tag here but looking at the history I clearly did. I think my intention was to make the access tags on the allotment tracks consistent, in doing so I guess I must have accidentally copied the East Avenue name tag across, which was obviously a bit careless. I will restore the correct name.

Accidentally uploaded with wrong changeset description, should be: added missing street name (Fairham Brook Lane) from survey.

Your initial statement is inaccurate. Nothing here is imported. I have always been very careful in tagging imported data. Also, I wouldn't say anything is really 'guessed'. There is certainly abstraction/simplification as with any mapping. When I added these barriers it was common practice to draw right around the property to save time, but that doesn't change the fact the intention was to represent physical features (walls/fences/hedges).

To reiterate, you have deleted a number of clearly visible barriers. If you had improved the data, perhaps by deleting the bits that pass through buildings, I would have no disagreement.

Any reason why you have deleted lots of barriers between residential properties? Most are still clearly visible on the aerial imagery. The aerial imagery in this area is now poorer quality than when I added these barriers, so it would be preferable to revert this, rather than trace back less accurately.

As someone who knows this area well, your "delete some unmappable things" comment doesn't make sense to me. The data here is from survey or careful use of aerial imagery.

I think I last used the property_boundary tag about 9 years ago. It was experimental and deliberately limited to a small area. At that time there were few other examples of micromapping around individual residential properties and I was trying to come up with a satisfactory approach. It was influenced by the micromapping around Sutton Coldfield, which is similar to this, except barrier=fence was used there. I liked the idea of defining the areas of individual properties, but thought completely surrounding properties with barrier=fence was problematic, because in reality there will be a variety of barrier types or no barrier at all. I thought it was better to use a more generic tag which could, where appropriate, be replaced by something more specific later on.

I would have no objection to property_boundary being replaced with more specific tags or deleted if no physical barrier exists. Although I do still like the idea of generic barrier tags (property_boundary and field_boundary), because it is often difficult to determine the correct type from aerial imagery. I prefer to use something general, and where possible make it more specific later, rather than just guessing. Based on previous discussion I suspect I am in the minority in favouring this approach.

Hi,

I appreciate your quick response and the correction of the issues I pointed out.

I have restored recently surveyed stop names so they match what is on the ground. I appreciate there will be different opinions on how to tag stop signs which use the format 'street / landmark'. I have surveyed well over 100 stops including both parts in the name tag and I don't think they should have remotely been changed to another format without any discussion.

In cases where the names were surveyed several years ago, I have left the name you added, although I have spotted some I suspect are wrong.

Recently I have deliberately focussed on surveying stops in Derbyshire rather than Nottinghamshire because the NaPTAN data is much poorer quality there. The data in much more likely to match what is on the ground in Nottinghamshire. Also, I may be wrong, but I don't think the released Derbyshire NaPTAN data has been updated for about two years. There are places near me where the stops were reconfigured a while ago and this is still not in the data.

A couple of other observations:
Please don't delete both_ways=yes from stops. This is useful information indicating the stop on the other side of the road doesn't physically exist.

You appear sometimes to be deleting benches near bus stops. Personally I never separately map benches inside bus shelters, however, I do map benches near bus stops that look likely to be used by people not waiting for a bus and I don't think these should be removed.

Regards,
Will

Hi,

Another question. You have deleted 2 bus stops I recently added from survey:
osm.org/node/7143262191
osm.org/node/7143262188

Please could you confirm you have checked on the ground that these have really gone? At time of survey the Notts & Derby 59 route was shown on the stop and I've just confirmed they are still listed in the timetable for that route.

Regards,
Will

Hi,

I notice you are changing a lot of bus stop names. I've found several examples where these names differ from what I have recently seen on the ground. What source are you using and have you confirmed they have really changed?

For example, I surveyed this one about 5 months ago:
osm.org/node/4033900789/history

The name then was definitely:
Station Road
High Lane West

You have changed it to 'Crossroads'. I note this is the name given in the NaPTAN data, but the Derbyshire NaPTAN data does frequently differ with what is on the ground, and it's what is written on the stop that should be used.

Thanks,
Will

Your reply doesn't really address the specific issue I raised. I agree with quite a few of the changes you have made here. Where the areas contain several pitches, pavilions, car parks, etc. then recreation ground does seem a better choice than pitch. The examples I gave, however, are all single pitches already located within parks or recreation grounds. To me these are clearly pitches and not recreation grounds.

If you don't agree with the shape that the pitches have been drawn, then change the shape, but don't change them to something they aren't. More informal cricket pitches (such as are often found in public recreation grounds) often don't have clear boundaries, so the shapes used to represent them will vary. It's not reasonable to remove them just because they are not round or oval-shaped. It's also normal for these sort of pitches to overlap with football pitches.

Some of the changes made here look incorrect.

In the following three examples, you have changed cricket pitches located inside parks or recreation grounds to leisure=recreation_ground. This results in recreation grounds inside recreation grounds. Surely the common sense way to tag a single pitch inside a recreation ground is leisure=pitch?
osm.org/way/257476065
osm.org/way/29371977
osm.org/way/31288883

Additionally, I don't agree that a BMX track inside a park is a recreation ground. Pitch seems a better choice.
osm.org/way/328564494

Thank you for correcting the mapping. I took a few photos when I was there and will make some updates soon.

In general, please be aware that developers' plans and similar sources are usually subject to copyright, and can't be used in OSM without obtaining permission.