I watch with interest.

One way to reduce the number of ways in the relation is to join some of them. I know the user who did a lot of the intricate mapping of the delta likes to use many small segments. He also uses tracks and other highways as parts of the relation. This means the tracks and highways end up being cut into small segments, if anybody comes along and decides to join up these segmented highways the relation will be broken. The tracks and highways are not part of the relation anyway they just run along the edges. Also any tracks running along the edge of the wetland may not be permanent and in updated imagery they may not appear and so a mapper may decide to delete them and again the relation is broken. This is particularly a problem when novice mappers are part of a HOT project

I understand the concept of a 'super relation' but not seen an example. Would this relation just act as a container without any tags beside the name? Or would it have the tags relating to the object, in this case wetland and wetland type while the 'child' relations have no descriptive tags and what roles do the 'child' relations have within the super relation?

Hi mapwitch,

This is a task I have looked at and shied away from. Just a question, when you breakup large multipolygons such as this how do you ensure they remain associated with one name that does not now appear in each new section?

Another technical point regarding the Okavango Delta. I does not neatly fall into any of the wetland categories, in fact it is part of the Zambezian flooded grasslands ecoregion (https://en.wikipedia.org/wiki/Zambezian_flooded_grasslands) which would imply it is more accurately a wet meadow rather than a marsh, both render in a similar fashion though.

Hi Jomokela Kennedy,

Many of these Police stations already exist. Perhaps you need to check before adding large amounts of data in one go like this.

Hi TsangpoGorge,

These streams are really ephemeral streams i.e. they will generally only flow after a heavy rainfall there is no tag in OSM for ephemeral rivers or streams, this does not mean that the riverbed is dry when the stream is not flowing, the bed is generally sandy and will be wet a few feet below the surface and getting stuck in sand is just as bad as in mud. I am almost 100% sure there will be no ford infrastructure, I have crossed many of these types of streams in Zambia. This particular stream appears to run within what we term a 'dambo' or seasonal wetland just after the crossing point where there may be no distinct stream bed. Being ephemeral they are prone to flash floods during really big storms that can change the river banks completely, although upstream this stream runs in a pretty well defined deep channel but follow it down stream to where it meets the Lupande River and then switch between Bing and Maxar imagery to see how the river bank has been eaten away by some particularly heavy floods. I do not think the tag flood_prone=yes applies to this situation in the way the tag is intended, it is more for rivers that may burst their banks after heavy rains on a regular basis. These streams and rivers will run violently for a short time, they may not burst their banks because their banks have been eaten away over time to accommodate the floods.

I know technically that is correct but these rivers will change within the river bed from season to season and there is no actual nice constructed ford as shown in the accompanying image on the wiki just a point on the riverbank where you can get down to the river to cross so I felt tagging the way more accurately represents what is on the ground. Tagging the riverbanks is also not practical as that will make the river or more accurately the stream appear more substantial than it is. These waterways will be completely dry for more than 6 months of the year and may only have water in after it rains so even the idea of a ford is a bit ambiguous

see the ford wiki
"On a way

The highway should share nodes with the riverbanks where it crosses them and also a node with the waterway=stream where they cross. Split the highway between the riverbanks and apply ford=yes to this segment (in addition to all other information pertaining to the highway).

If the approach (up and down to the riverbank) to the ford is much different from the ford itself, it might make sense to split the way and tag them appropriately. (incline=* might apply for example).

You should not tag ford=yes on the waterway node, when used on the highway way. "

Hi TsangpoGorge,

I usually tag the whole width of the dry riverbed as a ford rather than a single point where the mapped stream/river crosses the way. These rivers will be dry and sandy for most the year and be prone to flash floods when it rains

Hi Charles96,

Don't put anything in the name field unless you know what the name really is

Hi Charles96,

Please don't use the word 'road' as the name or ref of a highway and please try and use the correct classifications for highways.

Thanks

Hi Wire Guy,

Suggest you use MAXAR Premium imagery, it is more current, you will notice most these buildings no longer exist. They tend to be temporary anyway as it is a flood plain and the 'buildings' are built by fishermen when the water levels drop then they move away when it rises again in the rains. It is pointless mapping any buildings in these areas.

Hi Ed_Nrs,

Welcome to OpenStreetMap. Are you sure these hospitals are where you have mapped them? In the imagery they are in the middle of nowhere.

Hi PiazzaDellaSignoria,

I have corrected the erroneous one way. As regards links, I have previously had a discussion with remote mappers regarding these in Zambia and feel they do not apply to many of what we call 'slip roads' here in Zambia but if I changed them somebody would come along and re tag them as links so for peace and quite I have left them as links. You may notice from the Mapillary images there are not many road markings at all on the roads, they do not get repainted often, makes driving at night a really tricky business. These 'links' will have been marked separating the traffic with a painted line at some point but the markings will have worn off. If you are observant you will also notice the traffic lights/signals are not operational at this junction, also a common problem in Zambia The vehicle in front has put their hazard lights on and will edge their way into the traffic hoping the vehicles in the main road will allow them in, if you are too timid you will not get across. Some drivers will treat a junction like this, where the lights are not working, as a 4 way stop but you can not be sure this will always be the case. At the next junction with Burma road the driver has actually ignored the traffic lights and turned left when they should wait for a green arrow. The driver has done this because the lights are not correctly synchronized and traffic from the right, in Burma Road, should be stopped by the red light you can see once the driver has completed the left turn and only traffic making a right turn off Burma Road into Nationalist Road should be able to go as indicated by the green arrow, if you think of it this should allow traffic turning left from Nationalist Road into Burma Road to go as well but the lights are not synchronized to allow this so the Mapillary driver has just ignored the lights. It pays to keep your eyes on the minibuses with the orange stripe, they are a law unto themselves. Lessons in road safety in Zambia

Hi PiazzaDellaSignoria,

I have reconfigured the layout to more accurately reflect what is on the ground. The other examples you have mentioned are a sightly different as they involve intersecting dual carriageways and a single carriageway road but I will change these to a similar layout to this junction as I think it is more accurate. I found that the junction at (way/685800626/) and way/671587449 was not correct as the link roads were not shown so have reconfigured it.

Hi PiazzaDellaSignoria,

I think these particular ways are more a link to the primary ways from the residential way. It is difficult in this situation as these ways are primarily creating connectivity for routing purposes and don't really reflect the situation as it is on the ground, I have used this junction a few times. In general I think the residential tag is over used by remote mappers in Africa, they tend to use the logic that if there is some form of building in the region of the road it is residential, particularly in rural areas. I would suggest to remote mappers that they ask the question "Would I call this a residential road in my home town?". The other problem is in a rapidly developing city like Lusaka many of the areas are mixed residential and commercial which makes classification difficult and there is not a lot of town planning involved.

How can this be a residential road?

I have reverted your change. The whole section of the way crossing the dry river bed was already tagged as a ford. There is no need to create a node with the tag ford

Hi rab,

Have fixed the one way error. It was my error as there have been so many new roads and when I was adding them I missed that section.

Hi animalkingdom,

There is no need to add a ford tag at these nodes as the way is split to span the dry river bed and the section within the riverbed has already been tagged as a ford

Hi rab,

I see that you have changed the section leading to the Kafue Hook bridge to construction. This is not entirely accurate, it is just the bridge being repaired and has been in this state for at least 6 years as far as I can ascertain, no funds, and I am not sure when work will resume. I saw your repairs were based on routing problems, I am not sure when these started to occur but suspect it is when the Chunga Gate barrier was created. I think the gate is now before the detour road, where you can see a queue of trucks in the Maxar imagery. I am not sure how the tag 'construction' affects routing as some roads are often usable while still under construction here in Zambia as they just close one lane but in this case it is not routeable as there is no access to the bridge. I will move the barrier to where I think it may be, I did travel through the gate Christmas last year but can't remember exactly where it was, in fact there is a sort of traffic control system here as the detour bridge is one lane only. The problem we have here is information is being added, I presume based on a trip made many years ago. I have removed the ferry route added further down the Kafue as it does not exist. It was the ferry for the original road west before the bridge was built in the late fifties, I think. While it did sort of operate in the '70s as a tourist entry into the park it has not been in use for decades. A way to spot if a ferry route is still in use in Zambia is to look for the pontoons moored at one of the banks in the imagery, when you know what to look for they are obvious.