Dear holymantsala,
You changed a large forest-farmalnd mosaic into a residential area. This contradict the definition of residential area: "An area of land having predominantly residential buildings such as houses or apartment buildings. " (see osm.wiki/Tag:landuse%3Dresidential). This tag is supposed to be used in residential parts of villages and cities.
All the best,
Alkalin

Bonjour Abetsi,
Il semble que ce multipolygone (relation 10731823) a une geometrie incorecte suite a votre modification.
Merci de jetter un oeil.
Bien a vous

Bonjour,
Vous avez fait une erreur ici il semble. Cette zone est une foret et non une zone résidentielle (habitations). Je corrige.

Cher Marie Ngonge,
Je constate que vous avez converti des relations correspondant a de larges zones de foret en zones résidentielle. Comme celle-ci. Il s'agit ici d'une foret et non de zone résidentielle. Svp faites attention a bien garder un tag de landuse ou natural qui soit correct. Je corrige certaines de ces erreurs.
Merci,
Alkalin

Hi JFL,
Thanks for your answer. No problem. To resolve this I resurrected relation 8779. The Congo river banks are complete again. There is a lot of mapping yet to do in order to correctly classify landcover on islands and surroundings of the river.
All the best,
AkalIn

Hi JFL-75,
In this changeset you deleted a large section of the Congo river (relation 8779 with tag natural=water). Are you still working on it, or do you plan to put it back?
Best,
AlkalIn

I realized that now by doing more research. Thanks for the correction. Happy mapping!

Hi syntex
1. I added a field "old name" as Muhoza for Ruhengeri. Source of the name Muhoza was found here:
https://en.wikipedia.org/wiki/Ruhengeri
Maybe alt_name is more appropriate?
2. Previous changeset by another user (osm.org/changeset/130143482) removed the tag "name" of Gisenyi, leaving the city with no name. I added Rubavu in the blank name field to fix it.
3. Changed Byumba to Gikumbi. That was a mistake it seems. Apologies!

And I agree with you for a large section of RN9, given its function, so I have reverted its northern part to primary.

Hello ian727,
A discussion was initiated recently on this topic here: osm.org/changeset/139440761
In that thread I explained the reasoning behind this choice. But I agree it is not optimal.
We should probably open a public discussion somewhere but I am not sure how to proceed. I think the best way forward is to discuss with the community how these roads should be classified in Central Africa.

Thanks. I have already reverted trunk roads to primary in CAR. I will continue with Rwanda progressively.
Happy mapping as well!

Thank you cats_pajamas for the detailed feedback and the time you spend on this.

There are several ways to judge road importance: function, infrastructure and national classification. The regional and country specific pages we discussed above do clarify which one to use, as you rightly pointed out.

To follow decision made by the East Africa community and I can revert changes in ref tag I made (Rwanda, Burundi). I am not an expert in reversal procedure. But I can tag all roads currently tagged trunk in both countries, that are either not paved or not divided, into primary. Same apply for CAR. Do you agree on this, or have a better procedure to suggest?

For Congo, DRC, and other countries where no specific decision has been made, I want to point out that in osm.wiki/Key:highway it is written that "Note that highway=* distinguishes roads by function and importance rather by their physical characteristic and legal classification. Usually these things are highly correlated, but OSM is not obligated to copy official road classifications. " Which align with osm.wiki/Highway_Tag_Africa.
So, I was wrong to give priority on the national classification rather than the function. And will fix this progressively (roads like RN9 in DRC you mentioned). But I am not sure I can consider the state of the road which is often variable along the way of single roads and can be labeled with surface=* and other tags. The RN6 you mention above in Congo is a good example. RN6 plus RN3 are indeed famously unpaved and 10m wide but do function as international link between Libreville and Pointe Noire. I have not taken this road so certainly someone who visited the place can tell more about that. It means to me that we should assess status based on the function of the road (connection between major cities) to remain consistent.

All the best

Thanks ElliotPlack! There is still a lot of farmland mapping to do in the forest zone of Cameroon.

Thank you SLMapper!

Hi cats_pajama,
I based this tag change on the guidline found in the wiki page of highway tags in Africa (osm.wiki/Highway_Tag_Africa). The wiki state that highway=trunk should be used for “The most important roads in a country’s system that aren’t motorways. (Need not necessarily be a divided highway.) Note that such roads can be unpaved.”

Given that in these countries National Roads (RN, NR or N, depending on country) are the most important roads, then trunk should be the most apropriated tag for these.

Do you think another classficiation is justified for these national roads ?

All the best,
AlkalIn

Looking more carefully at the page you reffer to, it appear that osm.wiki/East_Africa_Tagging_Guidelines require trunk road to be paved and divided while osm.wiki/Highway_Tag_Africa (that I followed for changes in Central Africa and here in Rwanda) does not require that. It is unfortunate that Africa and East_Africa guideleines differ,

Hi cats_pajamas,
Sorry for the late reply.
I based this tag change on the guidline found in the wiki page of highway tags in Africa (osm.wiki/Highway_Tag_Africa). The wiki state that highway=trunk should be used for “The most important roads in a country’s system that aren’t motorways. (Need not necessarily be a divided highway.) Note that such roads can be unpaved.”

Given that in these countries National Roads (RN, NR or N, depending on country) are the most important roads, then trunk should be the most apropriated tag for these. As specified in the wiki page the fact that it is undivided and unpaved does not matter. Though I undertand the concern that some (or most) of these nationals are very bad unpaved roads and it look weird that they have the same tag as large paved road in other countries. However I agree with the wiki that what matter the most is the country classification and not the state of the road.

And I might indeed have missed some discussions around these topics. Please let me know if they are specific ressources I should check.

Hi SLMapper,

Thanks for the comment!
This area needs to be more accurately digitized indeed. But I can tell it is mostly farmlands for the following reasons: (1) I have been there to survey and this is a mixture of pure farmlands, oilplam plantations (which should be tagged as orchad and tree=oil palm), and forests converted to agroforest with little trees (not trees with canopy touching each other as in most definitions of forest). (3) According to the tree height calculated in Potapov et al, 2020, Global Forest Canopy Height 2019, this area consists mainly of very short vegetation (crops and shrubs).

This is shifting agriculture so it is sometimes hard to distinguish on high res imagery between forest and the regrowth of small shrubs that happens during the fallow stage between two successive plantation of crops.

But for sure more details can be added as they are still some very small patch of forests inside the farmlands, visible in popatov dataset. These should be digitized.

Hi Fizzie41,
Indeed the multipolygon was not complete. I corrected that.

Hi. It is indeed not a recognized tag (ANTIC). I am not sure what was meant by the person who created it. So I removed it.