You are being too cynical. There's a lot of data outside of open streetmap that complements open streetmap. A lot of that data is open but hard to combine with openstreet maps. And some of that data is owned by companies that are willing to license the data in an open way.
Overture basically is a joint effort by several companies to combine all that data.
Speaking as someone who has worked with OSM data. It's great for maps but severely lacking elsewhere with a lot of incomplete data, poorly/inconsistently tagged data, lot's of regional variation in tagging, etc. All this presents challenges for users of this data wanting to build stuff on top of this data. And there are lots of companies that are replicating efforts to fix this between each other. Been there done that. This is hard, non trivial work.
Overture is an attempt to move on from lots of companies reinventing this wheel to get to a state where there is a decent data foundation to build their applications on.
And they are releasing that data under an open license. So, there's a lot to like here. The process of how this data is produced is not as open unfortunately and it is unfortunate that they are doing this outside of the openstreetmap community.
But then perhaps that community wasn't that welcoming to get such a thing done? Nor have they seem capable or willing to do such work themselves. Overture are clearly working around them and it's worth spending some time reflecting on whether that could have worked differently and what would have had to change on both sides for that to happen.
They seem quite happy to use OSM data. We'll see how this stacks up w/r/t license: from here (and obviously IANAL), it looks very much like "sure, give us all the ODbL data you have, but we're doing a magic trick here, and the end result is not ODbL, hee hee". https://docs.overturemaps.org/release-notes/data-attribution...
That page seems to save that almost all of the data except places IS OdBL and requires attributing OpenStreetMap?
OSM POI data is definitely severely lacking a lot of regions -- so it makes sense why they'd source that from somewhere else. It's also pretty cool that meta and microsoft's POI data is now open -- from look at it, it has the opposite problem of OSM: there are a lot of businesses in it that don't exist anymore or whose info isn't totally accurate -- but it's very usable for certain use cases (and, funnily enough, could certainly form the starting point for a clean up effort -- that COULD then be stuck in OSM and licensed under OdBL).
it seems to be making new map data under a permissive licence instead of openstreetmap's copyleft one, which makes it possible to coopt the work of volunteers
Looks like the only permissively licensed part is the "Places" dataset provided by Meta and Microsoft https://docs.overturemaps.org/release-notes/data-attribution... so the volunteers being coopted are e.g. people who made a Facebook page for their business?
Any new data not in OSM is under a separate dataset and not under a copyleft license, so anyone contributing to that in the future means that their data might be used in a proprietary map.
In general, this seems to be more aimed as a corporate-friendly replacement of OSM where instead of volunteers, its companies with satellite data and phone data and AIs and only the end result is shared, not the process. So you're right that its not really being unfair to the OSM community whose data they seem to only reluctantly want (because it's copyleft) and have kept as such.
I am worried though that if if this gets popular, that OSM will be starved for funding and contributors
> anyone contributing to that in the future means that their data might be used in a proprietary map
But overture doesn’t generally allow the type of contributions OSM allows. There is no volunteer to map out a set of houses or streets. Allowed types of contributions are large mapping datasets, typically of the type explicitly disallowed from being imported to OSM.
