I created a MapRoulette challenge that lets you find buildings that do not exist anymore:
https://maproulette.org/browse/challenges/41748
This challenge lists buildings where maaamet:ETAK have been deleted from Maa-amet dataset and not replaced by a new building.
It’s important to point out that some buildings have been deleted from Maa-amet map based on latest imagery, but are still visible in the official orthophoto. In such cases use kiirortofoto.
Shouldn’t this building be included in the dataset?
You are right, don’t see any apparent reason why it was not included…
And @qqqqqqqqqqqqqqqqqqqq, I’m not sure on maproulette task comments visibility, so here is my feedback:
Looks like a bunker. Probably is a correct building, just remove etak id? link
Yep, also, it’s better to flag it as Not an Issue.
An error in etak? It’s the neighboring house which is demolished, but it’s present on Maa-amet link
Latest imagery is indeed a bit unclear. But here is historic imagery from 2017:
https://xgis.maaamet.ee/xgis2/page/link/nHgeLFBi
vs 2018:
https://xgis.maaamet.ee/xgis2/page/link/1p6om53
Maybe now it’s replaced by a smaller shed, but to me it looks like the original larger building is no longer there.
I think part of the dataset might have been cut off somewhere and might happen that there are a lot more buildings not removed
https://www.openstreetmap.org/way/213129966
https://www.openstreetmap.org/way/746657991
https://www.openstreetmap.org/way/215131051
https://www.openstreetmap.org/way/228623699
This is very likely indeed. Even the task distribution on Maproulette map view seems quite uneven.
But here 2 additional Maproulette challenges:
Both should be quite simple ones. Basically I double-checked that all buildings that have assigned addresses in ADS are not classified as Ehitatav hoone, Vundament or Vare in ETAK dataset. This means all the buildings where ADS script has updated address, should be in functional state. If someone prefers not to use Maproulette, same data can be queried from overpass: constructions & ruins
Since this data is somewhat indirect, there might be some false positives.