Can I get your opinion on petrol stations as well, most are just https://wiki.openstreetmap.org/wiki/Tag:amenity=fuel but some are like this. Having a separate https://wiki.openstreetmap.org/wiki/Tag:shop=convenience just seems wrong.

• Leave them separate?
• Would it be best to add https://wiki.openstreetmap.org/wiki/Tag:amenity=fuel to the shop?
• Change the building to https://wiki.openstreetmap.org/wiki/Tag:building:part=roof and https://wiki.openstreetmap.org/wiki/Tag:building:part=yes with an outline of building:yes with the shop and fuel tags?

Or something else.

I’m confidant that https://wiki.openstreetmap.org/wiki/Tag:shop=household is not just wrong, but invalid, it’s not listed on the wiki. So I was changing them to https://wiki.openstreetmap.org/wiki/Tag:shop=houseware which was the closest in terms of text, but also most popular. I have no argument against https://wiki.openstreetmap.org/wiki/Tag:shop=department_store my edits of them weren’t calculated, I either did them on auto pilot or accidentally copy the tag, all of this was manually so unfortunately its not as perfect as a program.

I did notice the variation in shop types, hence I was using the name tag to differentiate from supermarkets. I too have never been to an Asda Living store so I was going to stay out of that discussion. But if there is consensus I’m happy to do the bulk edit, I have all the node and way ids.

Just had a closer look at Asda Living stores, and at the beginning I wasn’t adding brand data to them. OSM actually has 30 out of 33. Only missing 3! (-9%) That’s amazing :) Notes for missing stores: 1 2 3

@GinaroZ thanks I’ll take a look. How do you know which imagery is older? I sometimes worry I’m readding a building to OSM that somebody else has deleted.

Just done some calculations. Asda website has 580 Superstore, Supermarket or Supercentre stores. While OSM has 577 https://wiki.openstreetmap.org/wiki/Tag:brand:wikidata=Q297410 https://wiki.openstreetmap.org/wiki/Tag:shop=supermarket. -3 or -0.5%. Pretty damn close!

Asda website has 319 petrol stations (19 stand alone and 300 at other stores). OSM has 229 https://wiki.openstreetmap.org/wiki/Tag:brand:wikidata=Q297410 https://wiki.openstreetmap.org/wiki/Tag:amenity=fuel. -90 or -28%. Less good, but I think a lot of them are in OSM but not tagged with brand data.

Asda website has 33 Asda Living stores. OSM has 21 https://wiki.openstreetmap.org/wiki/Tag:brand:wikidata=Q297410 https://wiki.openstreetmap.org/wiki/Tag:name=Asda Living. -12 or -36%.

Almost, I think I have notes open for each remaining unmapped store. 1 2 3 4 5 6 7

I’ve actually not checked how many Asdas there are in OSM, that could tell us if there are any that don’t exist on there website and probably need deleting.

Guys, I’ve got the solution, 3 little words.

“All rights reserved”

@SunColbat Loved reading this, gave me a chuckle, thank you.

You are doing great work! You’re doing a lot more details than I am, and that’s a good thing. Keep it up.

The website by Rob Whittaker updates about once a week, or just over, so the schools you edited will have moved into section 4 of the website, schools matched by ref. I don’t think the overall number of schools changes that often, it seems most of the time a school changes into an academy with a similar name.

That’s why linking to external datasets is so amazing, we basically get notification when a school changes, or closes. So we can know straight away to make changes to OSM. In terms of the amount of schools in OSM in the UK is not that far from official numbers but in terms of accuracy, we are way off, old names, no names etc. It’s not as bad as NaPTAN, where bus stops in OSM often haven’t been updated in a decade, but it could be much better.

The [Food Hygiene report] (https://gregrs.dev.openstreetmap.org/fhrs/) by Greg is another great tool. It’s primarily for getting Food Hygiene Ratings in OSM clients, but it also an amazing tool for OSM maintainers. Once linked it can basically give notification of POI closures so we can survey and remove from OSM.

I highly doubt Tesla exclusively uses a single source of data. It probably gets it from multiple sources, based on region. I expect OSM is one of those sources, Google Maps and user reports are probably used as well.

1. You might own the car, but Tesla owns any data collected my it. It’s possible that the cars report speed limits to Tesla that then has a policy to require multiple sources or cars to then change the speed limit they push to the UI. As such, you driving the road multiple times wouldn’t modify the data. However I don’t even know if the cars have cameras and computing power to read and parse signs on the fly.

2. I expect there will be an explanation somewhere, on one of the boring pages nobody reads.

3. There could be many reasons why Tesla updated after OSM. Based on that video, it took more than a year to update. So if Tesla is using OSM data they could be pulling it infrequently, sidenote, I remember that Tesla uses OSM data for parking aisles and it is pulled incredibly frequently. Alternatively Tesla could be using another data source, like Google Maps and it was corrected in that source during the year. Or somebody used the OSM data you provided to update Google Maps, not allowed but can happen.

I expect the data is in part from OSM, and your edit is what changed it. But I think unlike parking aisles it’s not directly from OSM. Tesla will periodically pull data from several sources, including data from customer reports, manually or automatically merge the data and only then is it updated in cars. Only Tesla know the exact details.

Which Address section does the County go into?

I’ve seen the county in at least addr:county, addr:state and probably others. addr:county is not mentioned on the wiki, so is probably wrong. addr:state is pretty vague, so I guess you could put it there.

However I’ve not been adding it, as the wiki says it can be interpreted.

Using addr:country=, addr:state=, and addr:province=* are optional, since they can almost always be determined from boundary relations (if present and valid).

Because the OSM map is deliberately de-listed due to concerns over server-load.

If we abstract that to “OSM is limiting it’s self because of technical reasons”, we’ve just discussed that in terms of not being able to click on POIs.

Our opinions seems to line up perfectly, and it’s actually annoying me/upsetting me that OSM isn’t moving forward in this direction. Unless there is something actionable we can do, I’m going to stop talking about it for now. I’ve already ruined my mood and it’s still the morning.

@Robert Whittaker that does make sense in terms of addresses, but I don’t think a client of OSM data can reliably use it. At least not without special casing for Post Offices, Carphone Warehouse, cafes in supermarkets, etc and that is not scalable.

A client needs to be able to say “Carphone Warehouse is located inside Currys PC World” but “name is located inside addr:housename” is far to generic.

My suggestion of https://wiki.openstreetmap.org/wiki/Tag:operator=Currys PC World and https://wiki.openstreetmap.org/wiki/Tag:indoor=yes is wrong. It’s an incorrect use of both tags, operator is a legal company that runs the node, it would line up sometimes but it’s not reliable.

I like the idea of is_in or rather is_in:store however it seems is_in is deprecated. So I guess my argument is that we need a new tag for this case.

@Richard I’ve still not found a source that says all the Currys PC World stores are 3-in-1 stores. The press release sort of implies it, but not completely. But once we see what rebranding happens, you may well be right, only one POI is needed in this case. Then I’d recommend https://wiki.openstreetmap.org/wiki/Tag:alt_name=Currys;PCWorld;Carphone Warehouse https://wiki.openstreetmap.org/wiki/Tag:shop=electronics;computer;mobile_phone with a name that matches the rebrand.

I’m not suggesting that the tags can’t be seen on osm.org, you and I both know that we can right click -> query, or click the weird question mark logo -> click. But someone new to OSM or just looking for details about a POI or a community doesn’t know that. It’s not intuitive, and so it is bad design. (I understand the technical reasons for it in the past)

When I said “it isn’t rendered” I was implying that because it isn’t part of the 2d image of a map, contributors don’t care about the data. You see evidence of this when people commit changes saying “Changed the icon of x” instead of “Changed x from a restaurant to fast food venue”. A lot of people contributing to OSM only care about that image of a map, the old school map, whereas we are talking about a digital map, intractable. Nobody uses a map to look up that in y the school is called x, now I’m going to get the phone number through a different service like 118 118 or the Yellow Pages. However that was a workflow for people in the past, that people actually did, but it makes no sense now.

You are right, a map HAS to be able to serve up that basic information in order to be useful, but obscuring like on osm.org currently does has the same effect of not having the data in the first place.

There are other reasons why this data is unprioritised, like why isn’t OSM indexed by search engines? Why aren’t OSM POIs in search engine results? You defiantly know more about limiting markets, target audience, and reach with your sales training than I do so I’m gonna stop typing. Sorry for the wall of text.

@alexkemp You probably know more about this than me, but I expect there are several reasons for a lack of contact information in OSM. One of the main ones being that it isn’t rendered, and so it’s excessively hard to see the contact details on osm.org. I believe if you could simply click or tap on POIs there would be more contact data in OSM.

In similar situations I would do something along the lines of https://wiki.openstreetmap.org/wiki/Tag:operator=Currys PC World and https://wiki.openstreetmap.org/wiki/Tag:indoor=yes, however I don’t think this is technically the correct use for indoor.

@Richard I don’t know any details about Carphone Warehouse, but if the opening times are different I would argue for a separate node. This is how I do it with cafes inside Morrisons stores, typically named “Market Street Cafe”.

@Wynndale same, a precise page of website is better than a generic homepage, in the case of chain shops. I think there are bots that change http to https sometimes, but to doesn’t take much extra time to do it ourselves.

I have now done the bulk of the S postcode, this is quite boring, but hopefully it helps somebody.

I also don’t copy websites links straight from Edubase, they are almost always the http, not https version and often go through redirects of old domains. I go to the site, then copy the link, the browser should go to https when available and follow the redirects.

That’s a good goal to have, good luck and have fun doing it.

I’m new to the OSM community, and a white guy, so a bit out of my depth in a diversity discussion.

Does anyone know how “wheelchair” became so popular? Most editors request that data, some even give warnings and errors when it’s not present. Did OSM mandate it, or just came together as a community, decide it was the right thing to do. OSM is probably the best map for physically disabled people now.

Can the same be done for other things that we don’t usually thing about? The one that comes to my mind is “lit”, I rarely avoid a path because it’s dark, or feel scared in public. But huge amounts of people do. How proud would you feel if you heard somebody say they use OSM because it makes them feel safe?

In the case of “lit” I imagine satellite images at night would produce huge amounts old data. But I don’t know if there are any such images available for OSM to use.

How does OSM drive things like this? How do we get mappers to use the tag? How do you encourage developers of the editors to feature certain as a priority? How can we convey the importance to renderers?

The above is obviously about users rather than contributors. I don’t know if it goes without saying, but attracting users from different backgrounds and diversities will in turn result in some of them becoming contributors.