@woodpeck, do you kwow Todo Plugin in JOSM. Find landuse, add in Todo list, and you will be able to revise quickly.
Quality is our concern, and we are looking at various ways to improve. See https://opendatalabrdc.github.io/Blog/index.html#!Bulding_Geometry_Analysis_to_Support_OpenStreetMap_Quality_Analysis.md

@woodpeck What's make you think that way 614459980 do not express reality? This landuse polygon corresponds to a certain density of houses in the area while less dense areas are ignored. I revised all polygons and they all correspond to housing.

clairdelune is doing a great job in RDC coordinating / responding to the #Ebola2018 outbreak and such information is important for a quick response, to identify populated places. And, Has said in the comment
"with a manual review of areas #Ebola2018"

merci de collaborer a opencity kinshasa. Ðans ce changeset, le tracé des immeubles est plus grand que sur l'image, Si nécessaire zoomez davantage pour afficher le batiment en plus gros et tracé plus près des bordures du batiment.

Import limites des territoires protégés vient compléter les contours du lac Albanel.

Important cependant de réconcilier avec ce qui est déja tracé. Lorsque Ile existe deja, exemple chemin=46906913, on ne doit pas creer une nouvelle ile en parallele tel chemin 598692720, mais plutot mettre a jour ce qui existe. Pour nouvelles iles ajouter, elles doivent etre ajoutées a la relation du Lac Albanel avec role=inner.

Well this developper is answering two years later without opening discussion too much.

Do you pretend that power substations are not industrial areas?

Ce changeset a aussi enlevé le nom de la relation waterway=river et créé un doublon chemin waterway=river qui recouvre plusieurs chemins admin_level. Un tel doublon est difficile à repérer et crée souvent des problèmes d'édition par la suite. Je corrige immédiatement cette portion.

Rivière des Outaouais Barrage des Rapides-des-Quinze
2018-06-14 http://overpass-turbo.eu/s/zGJ
Votre commentaie indique que vous voulez corriger des zones non fermées. Voir requête overpass ci-haut. Avant vos modifications, Le contour de la Rivière Bécancour était complet. Beaucoup de travail de ma part pour réviser cette rivière. Quel était votre objectif de réviser les modifications que j'ai fait dans ce changeset et d'effacer les chemins que j'ai ajouté ?
------------------------
Your comment says that you correct areas for ring not closed. See Overpass query above where we see that the ring was closed.
There was a lot of work to revise this river.
What was your objective to revise this changeset and to delete the ways that I added ?

Barrage de Rapide-7
Vous avez modifié la relation de type=multypolygon a type=building. A noter que c'est un barrage avec un immeuble sur une partie du barrage. Dans ce contexte, les roles part et outline sont inappropriés.
-----------------------------
You have modified the relation
from type=multypolygon to type=building. Note that this is a dam with a building standing on one part of the dam. In this contex, the part + outline roles are inappropriate.

Rivière Bécancour 2018-06-17
http://overpass-turbo.eu/s/zGI

Votre message indique que vous voulez corriger des zones non fermées. Avant vos modifications, Le contour de la Rivière Bécancour était complet, Il ne l'est plus. Et vous avez effacer de nombreuses données.

Je suggère que vous fassiez un revert de ce changeset.
---------------------------------
Your comment says that you correct areas for ring not closed. Before your modifications, this was closed. It is not anymore and you haved deleted many ways.

Do you plan to revert your changeset ?

Thanks for your contribution to support the OSM-DRC community in Kinshasa. Some suggestions to learn how to trace more precisely the buildings.
1. be carfull to zoom enough
2. If perspective modify the shape of building, remember that they are rectangular, trace rectangular and with iD, use the Rectangle button to square the builiding once finished tracing it.
3. do not make one square fo many buildings. Trace each individual building

To be more precise, 594283985 is a simple split of way 23359415 to stay under 2,000 nodes. Badly we lose history when doing so. Since I did not trace way 594283985 , I will leave the source as is, but after your observation I did remove source indication on way 23359415 were I used sources as indicated in my changeset metadata.

Thanks to contribute to this OSM response in Congo. You have traced at least two large polygons in this edition with tag landuse=residential using Esri imagery. Note that this imagery is imprecise. You should be carefull to spot regular patterns that correspond to houses. Ways 591053318 and 591053319 are traced over a wetland where no houses can be seen.

Living_street regulation from UK do not apply to Africa villages. A road that cross a residential area is still a minor road highway=unclassified. It should not be cut for each of these places to apply highway=living_street. In villages, if only pedestrians, not large enough for cars to circulate and connect to other roads, we apply highway=path

Thanks to collaborate to this response. Be carefull with your interpretion of blurry images like Esri is in this area. In this changeset, you added two houses over a road.

I also have used DigitalGlobe Premium is some areas. Here it is clearer but not very detailed. I would say, for the three polygons it looks more like residus.
For other areas, Blury Esri image is blury, imprecise. More or less rectangular objects can be interpredted as houses but with no certainty.
I try to anticipate NGO's needs in these various areas. One solution would be to create a task for more experienced contributors to revise rapidly each polygon adding a variable that confirms that houses are clearly observable.

Both you and me are experienced contributors and did contribute a lot to this Response in Congo. And with outdated images (Bing) or blurry images (Esri), it is difficult to interpret. Your insigth can help to understand.

Residential polygons surface can be used to estimate population in various areas. There is a considerable number of residential polygons. Some with clear image are quite easy to interpret and we can confirm that there are houses . For other ones, we more or less make an estimate that there are house. The smaller ones, more recents are difficult to interpret. Esri has more recent imagery but fuzzy.

I validate rapidly the polygons to see if any problems. For the three polygons I listed, Esri shows green area. Would this mean there were woodcuts first and vegetation is regaining? Sometimes, in woodcuts, wood residus can be confused with houses.

Il y a beacoup de coupes de bois a ne pas confondre avec zones résidentielles. There are many wood cuts to distinguish from residential landuse.
polygons > woodcuts by nearby population ?
w.589811868, w.589811867, w.5bb84443

Thanks for your collaboration to this OSM-DRC response. Note that buildings are generally rectangular but might look distorted when the satellite camera is not directly over it. Use the iÐ editor rectangular fonction - a button with a square to do that.

Thanks for your collaboration to this OSM-DRC response. Be carefull to square buildings even if sometimes they dont look square because of the perspective. In iÐ editor, thir is a button with a square to do that.

Thanks for your collaboration to this OSM-DRC response. Please look how to square buildings. In iÐ editor, thir is a button with a square to do that.