Fuck AI companies! There is good faith use and then there’s whatever the fuck this is. For some reason apps that are actually useful (OsmAnd, CoMaps, Mapy, …) all figured out how to mirror OSM data on their own servers so that it’s only downloaded once per app once per month or so, and then users never interact with osm.org. Meanwhile a trillion-dollar industry can’t figure this shit out!
It’s not a secret, this page lists like 20 ways how to download data, depending on your use case: https://wiki.openstreetmap.org/wiki/Downloading_data
Unfortunately it requires an actual brain to read and understand that article.
can’t figure this shit out!
Oh they can. They just won’t.
I can’t help but wonder if this isn’t an effort being done surreptitiously by a well-known producer of online maps, with the intent of strengthening their monopoly.
Unlikely, there are Weekly compilations of all Data/Diffs dedicated for downloading provided. This is most likely from vibe coded stuff just using the live api
That is their point. It’s stupid to crawl this way. So it must be some sort of DDoS attack
Every CS kid, and many hobbiest vibe coders are running fleets of agents that do stupid stuff in high volume.
I think it isn’t:
- First of all Hanlon’s razor.
- It doesn’t affect map users and rendering clients directly, as they display extracts, they don’t use the main website directly.
- Power users also work with offline clients, e.g. I download the data, work on it offline, then upload it, so what I notice from this is sometimes the download or upload times out, but I just retry it and it works after a while. It’s annoying, but not a big deal from my point of view.
- I think it mostly affects casual mappers, who add a lot of very important data, but by volume it’s much smaller than the data contributed by the big mappers. There was a very good presentation about this in 2014: https://wiki.openstreetmap.org/wiki/SotM_2014_session:_The_Long_Tail_of_OpenStreetMap
- It affects mostly sysadmins, who will burn out.
Slightly off topic, but do the large Map apps (Google, Apple, etc) donate to OpenStreetMap at all? I’m pretty sure they use their data, or at the very least check for updates to the open map. I’ve had it a couple times where I’ve updated the map, then a couple weeks later, it’s updated in Apple Maps and Google Maps. If that is the case then surely they can pay to support the service.
Google doesn’t use osm officially. I don’t know the status of Apple, while they may use some data from osm, their map is very different
There are big companies who use it (Microsoft, Meta, Amazon), and they donate money and manpower, they have employees actively working on osm.
You can see organised editing activities here: https://wiki.openstreetmap.org/wiki/Organised_Editing/Activities This is where they should document how they contribute data.
Financial sponsors are listed on the webpage of the foundation: https://osmfoundation.org/wiki/Corporate_Members
About your perception of copying from osm: a lot of data on those maps are contributed by users, who may copy it from osm or may not. This problem is actually as old as mapmaking, the only way to prove they copy your map is to include “trap streets” and see if it shows up in other maps.
Apple maps used to use OSM for many parts of the world, and still uses it in some remote areas. They even have a team of employees being paid to edit/update it: https://wiki.openstreetmap.org/wiki/Organised_Editing/Activities/Apple
Bing at least provides their imagery as a resource you can use to update osm too. Wonder how that happened and how long it will last
Bing is owned by Microsoft, and Microsoft is a platinum corporate member of the foundation, it means they pay at least € 30,000 annually.
They won’t disappear soon, they need map data and the other option would be paying to google.
Actually Ms, Meta, Amazon started to support osm because they figured out they can’t really compete against google individually, supporting an open project is cheaper than developing their own maps, or paying millions to google (remember, google maps is free only to individuals, companies have to pay a lot). They don’t really have other option, and the license of osm is very corporate friendly.
@infeeeee @Munkisquisher Bing providing access to their imagery goes back when Steve Coast worked for them.
deleted by creator
They used to use some data from osm, public transport in some parts of Poland, and they displayed the required attribution correctly:
https://xcancel.com/MapAmorePH/status/1453863816110952450
I have a feeling, it’s just some users who copy from osm to gmaps, and they don’t know it’s illegal. They are already not the smartest people on earth, as they work for free to a for profit company…
I think Google’s fault is mostly just not enforcing copyright laws. I’ve never seen something on gmaps copied from osm, but I reverted a lot of changesets because they copied from google. So your experiences are not as general as you think
In addition to the other companies already listed, Lyft does have an OSM team for providing updates to the map based on their own street level imagery:
https://wiki.openstreetmap.org/wiki/Lyft
I see them updating the roads in my area from time to time. So, while they’re not one of the Corporate Members, they are at least donating their time to help update the maps.
They definitely take additions from osm however Google seems incapable of noticing changes or deletions for osm. Which as led to some roads going through new houses near me. So I’d be cautious if they provide mapping data back to the project
@Theoriginalthon @DrCake I imagine google is very careful about ever adding anything from OSM.
If they take OSM map edis, they have to release all their data
How I understand it, they wouldn’t have to release it if they do it correctly. ODbL allows some mixing with proprietary sources, as listed here: https://osmfoundation.org/wiki/Licence/Community_Guidelines/Horizontal_Map_Layers_-_Guideline
The principle that a map maker can make and publish a map made from several distinct horizontal layers without being obliged to share data from the non-OpenStreetMap layers has been established for several years and is much clearer with the switch to the ODbL. Here we formalise and state what we are happy with and give examples that we hope will help potential OpenStreetMap users with incompatible sources.
They have definitely taken my edits on osm and added to Google maps. Some farm buildings got demolished and housing built on the spot. I was the one who added the buildings in the first place then removed then as they got demolished, then added roads new outlines etc. Google has the road going through the farm building. On the licencing front I’ve got no idea
should we add fake roads to osm to see if Boogle will add them (like paper map publishers used to do) ?
thanks for the link
Does OSM have copyright easter eggs?
Although some OSMers have deliberately added errors to the OSM data as well (see posting on legal-talk by User:80n), this is strictly discouraged and contrary to OSM policy. It’s also very unnecessary. Even if we seek to represent reality perfectly, that will never be the reality of map data. With OpenStreetMap there’s a very unique and distinct fingerprint evident in the data coverage and details included, and yes, in the errors made. For example in 2012 the OpenStreetMap foundation could issue a statement with absolute confidence that Apple had used our maps (without crediting us) based on very evident copying, and without the need for introduced errors.
It’s illegal, you should report it if you are sure it was from osm. But I guess they get data like that directly from municipalities, or road construction companies, or from the traces of users they spy on
