node/153754270 Peru, IL would have been more clear though.

Yes, its not in the wiki.
Why? Without any such a tag it is not possible to easily search for this group of polygons by taginfo or overpass. So it was just wrapping the tag to the preexisting relations and hoping it sticks.

Uff, das muss ja schon zwei Jahre her sein, dass ich da vorbei gekommen bin. An der Stelle war vorher ein Spielplatz, der Baum dürfte ein Überrest davon sein. Das Gebäude ist relativ neu, sollte aber eigentlich mittlerweile fertig sein. Wenn da mal jemand vorbei kommt kann er die die anderen Bäume auf dem Grundstück auch mit abklappern.

My modus operandi is switching between the "b" tool for simple building shapes and the usual line tool for more complex buildings.
As the former adds building tags automatically it does happen that I forget adding the tags for the latter, usually the JOSM validator warns about those cases and I fix them before uploading, but it does happen that they slip through.

If there is a reason for me to keep outlines without tags I usually write that down with the "note" tag. So go forward.

Yeah, =residential should be correct, though #070 and #083, at least the southern part, might be debatable. #083 also seems to continue to the north - but I am just an armchair mapper from overseas :D

As far as Iunderstood it till now:
landuse=forest --> logging activity et cetera This is not what I changed and not what I wanted to specify.

natural=wood --> originally any area littered with trees that is specifically not landuse=forest.

If the campus is "wooded except where tagged otherwise", wouldn't it be better to add natural=wood or anything to /way/670613089 ?

Damn - lost a lengthy answer...

I disagree, various paths that were used in these relations are pretty far from being usable.

Sometimes the geometry is rather off:
The connection to the parking lot between Science and Engineering Library and ISB would have gone across a wall.

Sometimes the data disregards real obstacles:
Engineering 2 (when terminating at Baskin Circle next to the bus station) would have crossed stairs [Mapillary photos from 2014] - though I think there is a ramp next to it that would terminate at the south end of Engineering Loop, so that would require an additional [currently missing] footway.

Sometimes the usability of the data is impaired when the paths are not connected to streets where they should:
Parking lot between Biomedical Sciences [The disappearing of that building... should be fixed now..] and Science Lib to Red Hill Road
Parking Lot between Science Lib and ISB to Red Hill Road
Parking Lot east of Earth&Marine Sciences to Red Hill Road

Also something I already changed two days ago, the path across the parking lot 139 north of Engineering 2 overlapped the parking aisles. In that place I made sure to add a wheelchair=yes to the respective lanes - something I admit, I should have been more vigilant about when editing in the Science Hill area.

As useful as the data should be, it's usability is impaired if the geometry it is based on is inherently flawed, see points above and also some other strange cases of Engineering 2 near the Communications building

I probably should have kept the entrance of the ISB next to ISB Bridge.

"assume it is wooded except where it is tagged otherwise" is the problem. The Natural Reserve areas were/are also tagged with "natural=wood" - and I think I can see why it was supposed to be like that, they are in large parts covered by trees and is probably considered a forest as a whole in regards to laws and other relevant topics.

However the actual outline of area covered by trees is different - "except where it is tagged otherwise" - okay, we are kind of missing fitting area=residential, area=highway, landuse=meadow, natural=scrub to fill the gaps but I wanted the natural=wood to represent the outline more accurately than a shapefile originating from the designation of the protected area.

So what I did was copying the geometry of the protected area to have a starting point, included "natural=wood" and then edited the outline of the copy to more closely resemble reality. Finally I removed only "natural=wood" from the original protected area so there are no overlapping areas of "natural=wood" in the data.

Roughly spoken, a forest without trees should not be tagged as a forest in OSM.

Hey, you created a relation for administrative boundaries with the admin level 3 in this changeset. Is there a reason for this? Especially since it's area is completely inside the AL4 Pomeranian Voivodeship?

What I found when searching for "Kaszuby" was a region which should rather be
("type = multipolygon") and "place=region"
e.g.
way/360359268
relation/3734731

Outlining the area that is used for housing.

The label itself doesn't show the extent of the city at all, one might use the mapped streets to get a grasp, but that might also be misleading (weak example, south west of Ashgabat ;))

I've seen that there was another outline of the city roughly adhering to the residential area, similar to what is still left for many settlements in the Ukraine or the Russian Federation. I think of those polygons as an interim stage to completely mapped administrative divisions. Mainly such polygons that should match the administrative boundaries of the city should also contain the industrial areas south of the railway track, in the north east of the city and probably the airport. All of which are probably not used for housing.

However, to begin with I hope be both agree that the residential polygon for the large (and apparently pretty new) area north west of Balkanabat is useful right now.

Pokemon GO-Vandalismus in Änderungen in changeset/58432534 revertiert.

Pokemon GO-Vandalismus in Änderungen in changeset/58432534 revertiert.

Hallo, und willkommen beim OpenStreetMap-Projekt.
Unser Ziel ist es eine detailgetreue und wahrheitsgemäße Karte dieser Welt zu erstellen. Deshalb bitte ich dich keine erfundenen Orte und Gebäude einzutragen.

Nach einer kurzen Überprüfung mit den selben Luftbildern bin Ich zu dem Schluss gekommen, dass sich an der Stelle dieser Höfe kein bisher unbekannter Wallfahrtsort befindet und habe deine Änderungen in changeset/58432534 gelöscht.

I am looking at the relations listed in wambachers missing boundaries analysis: https://wambachers-osm.website/index.php/10-osm-reports/1119-countries-compare-2018-01-31

Current thread (german): https://forum.openstreetmap.org/viewtopic.php?id=61207

I fear the names were copied from the 38north digital atlas (http://38northdigitalatlas.org/ - copyrighted) as claimed by one of the authors Curtis Melvin. (https://twitter.com/CurtisMelvin/status/952947641058516992)

See also history of: node/3723463569/history#map=13/41.0552/126.6610

Hello,
thanks for your contributions and welcome ot the OpenStreetMap project.

However there are two problems with this changeset.
First, you deleted several roads (including their history) and created new lines instead. Please do not do that, we won't be able to connect those new lines to the old history, rather move and edit the road if you change its path. Some other roads you didn't recreate (southern Strada Rulmentului), please keep in mind what you are trying to replace and finish that before uploading.

Second, when creating two ways that cross each other, please connect them at the point they are crossing (except for bridges and tunnels), routers are generally not able to change the road on crossing points if they are not connected.

Please apply those points in the future, if there are any questions do not hesitate to ask and happy mapping :).
- Helmchen42

Hello,
welcome and thanks for your contribution to the OpenStreetMap project.

Given the delicate nature of country borders. I'd like to ask you for your source of the "exact" location of the boundary line.

Also for your information, the changes broke the polygons of the boundary relations of Al Anbar and Najaf. I know that the iD editor only rarely indicates such errors, but please be aware of the possible mistakes especially when editing connecting nodes of multiple border segments.

Greetings, Helmchen

Hello,
your changeset introduced a large amount of leisure=playground north of Vladivostok. Those playgrounds seem to be duplicated with one of the duplicates slightly displaced to the northeast/southwest. A few of the playgrounds I've checked with bing imagery also seems to be plain wrong as they would be in the middle of the forest, in the middle of a field or occasionally intersecting allotments.

Can you explain some of those irregularities or did you maybe made a mistake somewhere?

Greetings, Helmchen

Wie Ich dazu komme: Eindeutigere Abgrenzung von örtlichen Bebauungen. Letztenendes würde ein landuse=residential hauptsächlich Gebiete ausfüllen die nicht landuse=farmland/meadow/forest natural=scrub/wood geschweigedenn landuse=industrial/commercial sind.

Als solches habe ich versucht von Wohnbebauung bedeckte Flächen anhand erkennbarer Grundstücksgrenzen (siehe die Ausbuchtungen in Eisfeld auf Höhe der Raststätte), zugegebenermaßen grob, zu erfassen. Als Quelle müssste ich entsprechend angeben: bing Luftbilder kombiniert mit einer hoffentlich nachvollziehbaren Abschätzung "ja das dürfte noch zum Wohngrundstück gehören.

This changeset was just a fix for broken boundaries listed in Wambacher's "International Boundaries" thread.

I have neither an opinion nor knowledge about the AL2 or Morocco in those disputed regions.