There are many other reasons why using relations is superior when mapping neighbourhoods.

The simplistic way of using landuse=residential as a polygon to map a neighbourhood has got many disadvantages:

1) producing overlapping landuse errors on JOSM checking especially if it overlaps with an apartment lot also tagged with landuse=residential

2) overmapping actual residential land area because much of the land in the neighbourhood is used for purposes other than residential (eg retail, amenities, recreational, forest, lakes, roads, govt offices, etc)

3) GPS routers will route to the physical centre of the neighbourhood rather than the "social centre" of the neighbourhood.

Using place=neighbourhood as a relation and adding a node label into that relation to map a neighbourhood solves the above 3 problems.

From the changeset records below, you can quickly see that under Relations, I have created 9 relations, including the Galleria, Bkt Bangsar, Bangsar Park, Federal Hill, SMK Bangsar. So I hope this put your mind to rest that no data was removed but rather evolved from polygonal to relational.

Relations mapping is inevitable. As the dataset becomes more and more populated with more data, especially with overlapping data - it is necessary to evolve from simple polygonal mapping to the more complex relations mapping.

Hi Rene. Sorry, seeing this message only now. The residential polygons have also been converted into relations. Eg: relation/19349201

While relations editing is more complex than way editing, it is the preferred method of mapping complex overlapping boundaries especially if the length of overlap is lengthy and if there are 3 places sharing the same way. In this instance, Bangsar Park, Federal Hill and Urban Community Forest all share the boundary way - hence they have all been converted to relations.

You will notice that in my relations tag, I have included previously=way:1136336131;way:705649073

So although the relation element is new with no previous history, one can see that the tag values were derived from the previous way polygons and can visit those still existing polygons to check its earlier history before the relation was created.

You have deleted the weather station trail citing "no trail", but many hikers can confirm the trail exists eg https://www.youtube.com/watch?v=PxKaCR2Sep4&t=267s

This is the third time you have deleted this trail without any discussion and despite me putting a note that this trail is not on private property but in Ampang Forest Reserve which can be legally accessed with a permit from the Selangor State Forestry Department.

You have also deleted the 389m hill peak and the origins of Sungai Sering / Seriong despite these land features actually existing on the ground.

Kindly explain your reasons for deleting valid data.

When I say township, I mean something like this: https://theedgemalaysia.com/article/cover-story-demand-penang-bigger-township-development

Wikipedia is mistaken. ( https://en.wikipedia.org/wiki/Mukim ) The closest English translation for mukim is NOT township. The closest English translation for mukim is PARISH. [ https://en.wikipedia.org/wiki/Parish_(administrative_division) ]

residential=rural is not applied to administrative boundaries.

Here is the correct usage of this tag:
osm.wiki/Tag%3Aresidential%3Drural

place=suburb only applies to large urban neighbourhoods / township development projects with plenty of amenities (schools, shopping malls, transportation hubs, places of worship, public park, etc... and perhaps some offices or even industrial area)

Normally townships have their own boundaries and don't follow mukim boundaries.

Great job mapping boundaries.

This is a mukim boundary, so as per guidelines in osm.wiki/Malaysia/Other_tagging_aspects#Administration , it should be given the tag admin_level=8

Hi Rayth, thks for updating Sg Rasau. I've been updating rivers in Selangor for the past 1 month, and I hope you won't mind me giving some pointers:

1) Way: Sungai Rasau way/1282593221 begins correctly until node node/11902084880 where the waterway should continue south towards Blue Lagoon Waterfall and further to Taman Pertanian University where it joins with the already existing Sg Rasau mapped there way/878669711 . After node node/11902084880 it should not be flowing east towards Kinrara Golf Club as rivers don't flow uphill (check against topo lines).

2) Way: Sungai Rasau way/1282593220 is not Sg Rasau but actually Sg Bohol tributary and very rarely culverts run under buildings. If you look carefully on ESRI, you can actually see the monsoon drain running below the trees in between Petronas and BK5 Hening Park node/1531129755 . This stream is NOT connected to Sg Rasau but begins from node/11902084899

3) As mentioned above, there is no connection between node/11902084889 and node/11902084899 . The Sg Rasau tributary begins at node/11902084889 to flow WEST to join the main trunk at node/11902084880

4) Tributary streams way/1112684894 , way/1112684893 and way/1112684888 should not be disconnected from the Sg Rasau main trunk.

5) Where the stream crosses hiking trails ( node/11902084936 , way/1112684882 , node/5156522577 , node/5104609835 and node/11902080155 ) you should map whether it is a ford, bridge or culvert. If unsure, place a fixme node where they intersect.

6) Waterways begin as small streams and usually flow for about 2 km before becoming significantly larger. So usually, the first 2 km of the river should be mapped as waterway=stream (can easily be crossed) instead of waterway=river (shoes fully soaked with water when attempting to cross).

I haven't hiked Wawasan for some time now. What happened to this trail?

I nearly remapped this trail as Strava still shows very strong heat signature. Perhaps, instead of deleting the trail completely, just delete the tags but leave the way on the map and add a note eg: "Dec2023: This trail is no longer usable because ___"

If you are in agreement with the wiki, I'd appreciate it very much if you reverted the changes. Tq.

Hi Thomas, please refer to OSM wiki examples when to use highway=path and when to use highway=footway.

osm.wiki/Path_examples

https://community.openstreetmap.org/c/communities/my/132

Oh, btw, dunno whether you're aware - our OSM Msian forum is up and running again and you've been reelected as moderator. You want to bring this discussion there?

I agree that for simple adjacent landuse overlaps - probably best to keep it simple and not resort to multipolygons.

Can you perhaps share what was it that you intended to edit, and how relations being present had complicated your intended edit - so that I can be more mindful how multipolygons affect other mappers and make a better pro/con consideration when I'm thinking of using multipolygons again in the future?

So I disagree that it is a peculiar use of multipolygon, though I agree it is more difficult to change. :-)

This recommendation demonstrates the use of relations/multipolygons in mapping adjacent landuse plots and administrative boundaries all overlapping each other in complex ways:

osm.wiki/Multipolygon_Examples#Farmland_adjacent_to_forest_and_farmyard_adjacent_to_lake_(Adjacent_multipolygons_and_adjacent_inner_areas)

When dealing with 2 or more boundaries which overlap each other, the recommended mapping strategy is to use relations.

When done using relations, clicking on a way will reveal it as a member of all the overlapping relations that way belongs to.

If you overlap the ways on top of each other without using relations, you won't be able to tell how many ways are on top of each other.

Beautiful work there, hundenvon!