Obsoleting Google Maps

rhhs · 2024-03-20T10:12:56.123Z

As said above, I don’t think we should compete with Google Maps. I guess the reason most of us are contributing to OSM because we enjoy the work and making the world a better place. We should focus on the strengths of OSM instead. OSM is filling some gaps that Google Maps doesn’t do very well on. One area is maps for hiking, where Google Maps is often useless because it’s too much focussed on car traffic. Almost all hiking apps use OSM data; I don’t know a single one based on Google Maps. A few other areas are mentioned in this article Finding the ‘invisible’ millions who are not on maps which has been a great inspiration for me to contribute to OSM. We should focus on providing frequently updated map data where there is no commercial interest for Google Maps to provide them. Google Maps is very strong in routing in urban areas because they use the location and speed of android users to propose the best route. OSM and its data users will never be able to outcompete that. Instead, we should focus on providing routing information in rural areas, where it is not the amount of traffic, but the road quality that is crucial for optimising routing. OSM provides this data (surface=* and especially smoothness=*), Google doesn’t. These data can also be used by authorities in planning their road maintenance work.

wicking · 2024-03-20T23:22:24.602Z

Yes, non-paper based maps (= apps/websites) for hiking and cycling are almost all based on Openstreetmap.

Similar to the fact, that most servers, smartphones and all supercomputers use Linux. But gaming desktops not yet.

Google maps is more than just the infrastructure data, it is also reviews of shops, restaurants etc. There is not a single app/website using the OSM nodes/data as an anchor for reviews (as far is I know). http://HappyCow.net for example has its own database. It could have been made like http://wheelmap.org, which uses not only OSM but adds a layer on top of that for photos, which are saved in another database.

Peter_Elderson · 2024-03-21T07:54:24.632Z

rhhs:

These data can also be used by authorities in planning their road maintenance work.

Authorities do not rely on uncontrolled, public data for maintenance planning. If they did, I’m sure my neighbours would not hesitate to alter a tag or two to speed up the long overdue maintenance of the street I live in. I myself would never do that, of course…

Seriously, a few “Safety Regions” (fire brigade+other emergency services) in Nederland use OSM to show routes, passages, entrances and terrains which e.g. a fire brigade vehicle could use to reach emergency locations. Because they depend on OSM tags and objects, which means lives may depend on it, they themselves check and maintain OSM for this in limited regions. This according to specs agreed between the service and the NL OSM community. And that is the only example of official use of OSM data I know.

I have been trying to get hiking organisations to use OSM data for maintenance and promotion of hiking routes and facilities. They simply won’t; they all insist on having their own expensive GIS systems, then export the data and display it on… OSM background. When I tell them OSM already contains all the routes, networks and POIs, and they can control and maintain them in the OSM database, they nod politely and simply don’t believe me.

Even if I actually show them examples of what can be done. They shy away because they see it as nerdy hobby work, not to be taken serious. (And because it’s free, and free stuff can’t be good, right? Can’t put a figure on it.)

dieterdreist · 2024-03-21T08:20:18.000Z

I have been trying to get hiking organisations to use OSM data for maintenance and promotion of hiking routes and facilities. They simply won’t; they all insist on having their own expensive GIS systems, then export the data and display it on… OSM background. When I tell them OSM already contains all the routes, networks and POIs, and they can control and maintain them in the OSM database, they nod politely and simply don’t believe me.

maybe I’ve told it before, the CAI (biggest Italian hiking organization with 300k members) has decided to embrace OpenStreetMap many years ago. They still have their own GIS but it syncs the routes from OpenStreetMap and differences have to be manually approved in order to get applied to their validated copy.

rhhs · 2024-03-21T09:10:24.475Z

Peter_Elderson:

Authorities do not rely on uncontrolled, public data for maintenance planning.

…if they have better data maintained by themselves, of course that’s true! But not many countries are mapped as well, and have databases as well maintained, as the Netherlands. Here in Bulgaria (also an EU member state), building cadastral data of most villages are not yet digitised, the organisation that maintains the European long distance paths still relies on paper maps from the late '80s, and a local party that recently got elected in Sofia city council showed great interest in my smoothness mapping because although they know maintenance is urgently needed, they don’t know where to start.

Peter_Elderson · 2024-03-21T10:07:33.661Z

dieterdreist:

maybe I’ve told it before, the CAI (biggest Italian hiking organization with 300k members) has decided to embrace OpenStreetMap many years ago. They still have their own GIS but it syncs the routes from OpenStreetMap and differences have to be manually approved in order to get applied to their validated copy.

Our Wandelnet is comparable. First they saw OSM as a threat, a competitor. Then they just tolerated it. In the meantime they used OSM-based maps as optional background layer in publications/sites - and as standard background map when Google started to require payment, for maps that are utterly unusable for hiking.
Now OSM is accepted, but only as a free background map, not for the routes and POIs.

I have been working the digital route quality angle, with some successes, and I just learned that volunteers outside the central office of Wandelnet are starting to use OSM-difference analysis, just like I promoted - they just didn’t bother to inform me. They should, because I can fix OSM-side problems in minutes.

This is all far from “embracing OSM”. I won’t give up, but the usabiliy of OSM remains an important factor - without specialised nerds, organisations can’t use it fully, other than as a background map or a comparison tool.

Currently, Thunderforest Outdoors is often used, it’s free and OSM based. But it doesn’t show benches at the right zoom level for hiking, let alone routes or Node Network routes and nodes. If you want to change that, it’s possible of course, you get pointed to libraries and toolboxes, in other words: the nerd world.

That’s just one example of the low practical usability of OSM. It maybe just a perceived problem, but IMO a presentation shell is missing. Currenly, volunteer organisations need to hire an expensive programming company to make proper use of OSM, which sort of defeats the free for all principle.

Minh_Nguyen · 2024-03-21T14:24:37.284Z

Our wiki does a decent job of documenting large-scale, global usage of OSM data by open-source projects and tech companies, but we could more thoroughly document and celebrate local and regional adoption by government agencies, media, nonprofits, and hobbyists. Some local wiki pages for North America have an “OSM in the wild” section that lists this usage. Lists like these can give local mappers ideas for improving map coverage and provide us with existing examples when trying to convince other local organizations to adopt OSM.

RAytoun · 2024-03-21T14:42:09.550Z

“And I’d say a shift in the goal of the project towards developing end-user products is neither realistic nor desirable.”

Even if we did try to consider an end user, which end user would be referring to? Every end user has a different requirement.

The cartographer that wants to create a map for the local park run running event only wants a base map to download and mark the route.

The person coming out of the theatre and wants to find their way to the nearest pub will need the route finder function and detailed pub information.

The researcher that wants to understand the area may be looking at the landuse to identify residential, industrial or commercial areas of a town.

All we can do is create a data base and they will select the data they need for their project.

pb07 · 2024-03-21T17:06:50.634Z

9p8:

See… I wish when people visit openstreetmap for the first time they see a list of projects like FacilMap, and OrganicMaps right off the bat. I’m surprised I’m finding out they exist for the first time in a random forum thread.

Yes I fully agree with that idea.
I have always been surprised that when opening openstreemap.org we immediately come across the choice Edit, History… before even knowing what we can use all of that for.

alan_gr:

I see your point, but I suppose part of the problem would be deciding what to list.

It seems to me that this problem is already there. When you open openstreetmap.org, apart from the standard base map you have the choice between two cycling maps, a map from tracetrack and a humanitarian map. Who decided on this choice? Why not add others in a wider range?

Peter_Elderson · 2024-03-21T17:36:53.330Z

pb07:

It seems to me that this problem is already there. When you open openstreetmap.org, apart from the standard base map you have the choice between two cycling maps, a map from tracetrack and a humanitarian map. Who decided on this choice? Why not add others in a wider range?

I guess because osm.org is, by design, still a nerd catcher rather than a showcase. Which is fine with me - I’m a nerd - but it does not help OSM’s position in the Outer World.

Anton_Khorev · 2024-03-21T18:16:50.361Z

pb07:

apart from the standard base map you have the choice between two cycling maps, a map from tracetrack and a humanitarian map. Who decided on this choice?

https://operations.osmfoundation.org/policies/new-tile-layers/

SimonPoole · 2024-03-21T18:24:10.979Z

pb07:

… Why not add others in a wider range?

@Anton_Khorev has already linked the policy, though that is, as these things go, quite recent. But outside of policy questions the providers of the layers need to be able and willing to spend money on providing a free service with doubtful returns in any form.

JMGonk · 2024-03-22T05:03:14.036Z

I don’t have too much to add except my experience as a Google Maps “contributor”. Maps is not what it used to be. The rate of garbage data making it into Maps exceeds the rate that their unpaid volunteers can remove it and the experience of suggesting edits has become incredibly frustrating. Eg. Maps shows a business - which has been closed for a couple of years - as being located in the middle of a broad avenue, even though the attached address is 25 blocks away. Suggest that it doesn’t exist? Not accepted. Try to move the map marker? Not accepted. It says that adding a photo will help support your suggestion. A photo of what?

Escalating within Maps doesn’t work. Sometimes it fails with a server error, other times the request just never gets acted on. The only way to get a human to look at the problem is to go to the community support forums where a “product expert” (unpaid volunteer with no authority than you) might submit the issue to the Maps Team - ie. escalate, minus the chance of a server error.

And don’t get me started on the dumpster fire that is photos of places in Maps.

Don’t try to compete with Maps because being like Maps is not a worthy goal.

dieterdreist · 2024-03-22T08:16:36.000Z

Don’t try to compete with Maps because being like Maps is not a worthy goal.

the thing is that on top of all the outdated and otherwise wrong data, they also have all businesses that are actually existing (or at least all businesses for whom it is important that they are found). Everybody (unless they hate google and can afford it) is going to add their shop or restaurant or hotel etc. to google maps because it is the goto address, and this is the reason they see so much seo and spam. If we could manage to replace them in this position we would have won, but currently we neither have the volunteer workforce to handle the amount of seo and spam we would see if we were the number one. Being in a kind of niche makes it cosier (not saying we don’t have spam or seo, but it is orders of magnitude lower than it could be).

rhhs · 2024-04-16T19:06:33.791Z

Peter_Elderson:

Our Wandelnet is comparable. First they saw OSM as a threat, a competitor. Then they just tolerated it. In the meantime they used OSM-based maps as optional background layer in publications/sites - and as standard background map when Google started to require payment, for maps that are utterly unusable for hiking.
Now OSM is accepted, but only as a free background map, not for the routes and POIs.

I guess organisations like Wandelnet are not willing to use OSM as a sole window to display their data because they want to keep full control of their data and are afraid of the “anyone can edit” principle that is the foundation of OSM and Wikipedia. Would it be an idea to get such organisations on board by allowing them to have some degree of exclusive control over their data? I.e. not everybody can edit the route relations that “belong” to Wandelnet. Something similar to the “claim this business” link on Google Maps. If it would be possible to have some degree of “ownership” of data on OSM, it might encourage organisations to use OSM to display their data, or even save us a lot of work by encouraging business owners to adding their businesses to the map and maintaining them? Of course there should be conditions to this “ownership” so we can take back control in case of misuse or if the business goes out of business.

Peter_Elderson · 2024-04-16T19:29:56.308Z

I have tried again and again to arrange a meeting about how we could coöperate and enhance each others work. They simply never want to talk. In the meantime they do appreciate my work on “digital route quality”, not directly but they now have a route quality team, and they now started to run analyses of differences between surveys and their own digital route maps, as I have shown that this will detect 50-100 errors for a route of say 200 Km. That’s 5-10 route errors per hiking day!

alan_gr · 2024-04-16T19:36:53.786Z

rhhs:

Would it be an idea to get such organisations on board by allowing them to have some degree of exclusive control over their data?

How would OSM contributors then update the data if it becomes out of date? Or add new tags as new tagging schemes or use cases emerge?

Mateusz_Konieczny · 2024-04-16T19:39:36.623Z

rhhs:

I.e. not everybody can edit the route relations that “belong” to Wandelnet.

That is not viable without entirely changing OSM data structure, even if would assume that such data ownership makes sense.

Why someone would be unable to say map another route or add detailed surface tags just because it is on way belonging to route claimed by someone?

Peter_Elderson · 2024-04-16T20:26:45.214Z

Mateusz_Konieczny:

Why someone would be unable to say map another route or add detailed surface tags just because it is on way belonging to route claimed by someone?

I think the owner/operator of routes is the primary source of the information. If a mapper finds that OSM differs from the situation on the road, (s)he can change it freely, as seen on the ground. The operator should monitor changes, there is tooling for that. I’m sure any differences can be easily resolved.

Tooling: Knooppuntnet Monitor can show the differences (on a map and with links editors) between the current route relation and a ‘frozen’ reference version of the same relation, for exactly this purpose: notice unscheduled changes and act on it.

rhhs · 2024-04-17T06:29:01.924Z

Mateusz_Konieczny:

That is not viable without entirely changing OSM data structure, even if would assume that such data ownership makes sense.

That I cannot judge, I have no idea how to implement that in code as I have little IT expertise. It might be a showstopper.

Mateusz_Konieczny:

Why someone would be unable to say map another route or add detailed surface tags just because it is on way belonging to route claimed by someone?

The “ownership” could be limited to having exclusive access only to the data that are crucial for that owner, such as the route relation information and which ways belong to it, but not restricting anyone from changing less crucial info such as surface tags, adding new tags, including a way in another route relation, etc. If some independent mapper wants to change crucial data such as adding or removing a way to a route relation, or change the name of the relation, the “owner” of that relation could have the right to approve or reject the proposed change. The “owner” is expected to react to such proposals, and if he doesn’t, the ownership rights could be revoked. The “owner” of a shop node could be required to update a checkdate tag once a month/quarter/year (and update any other tags that have changed since the last check), and if he doesn’t, the ownership will expire and the node will become editable for anyone.