Noting T288897 ... if WMF maps are to be sorted out, it is reasonable to consider whether it is now time to learn from map specialists such as OSM/ACME and implement the easily available and obviously useful map detail which they supply.

Edinburgh: Look - no castle! Mostly, no buildings. Unrecognisable. Entirely useless for my listed buildings use case.

https://maps.wikimedia.org/#16/55.9510/-3.1967

http://mapper.acme.com/?ll=55.95051,-3.19734&z=16

Downtown Lusaka, Zambia

http://mapper.acme.com/?ll=-15.42034,28.28233&z=16&t=M

https://maps.wikimedia.org/#16/-15.4178/28.2820

Victoria Falls. The railway is one of the defining features, as it goes over the gorge from Zambia to Zimbabwe. Users can almost see it on the WMF map, if they know what they're looking for.

http://mapper.acme.com/?ll=-17.92786,25.84826&z=14&t=M

https://maps.wikimedia.org/#14/-17.9287/25.8363

Bergen, zoom 11. Wikidata has 1,000s of hills & summits. use case: show them so that we can check whether WD coordinates point to the peaks.

https://mapper.acme.com/?ll=60.31845,5.31807&z=11&t=M

https://maps.wikimedia.org/#11/60.3531/5.3078

Are my Scottish items with coordinates in the correct P131 area?

https://w.wiki/3unv

Well to be honest, I really don't know b/c the WMF map choice is not to show admin boundaries, so I can't validate them

https://mapper.acme.com/?ll=56.19620,-3.83217&z=12&t=M

https://maps.wikimedia.org/#12/56.1815/-3.8456

I'll end, for now, with: as I've spend many hundreds of hours over several years curating the geocoded items for a country - Scotland - WMF might want to reciprocate by providing maps fit for purpose. The current map set is demonstrably not fit. Boundaries, buildings, landuse, peaks, please.

Eventually I do hope Wikipedia maps will use OpenMapTiles project (with some customization like no admin borders if so desired). There are a lot of efforts by numerous individuals spent on optimizing OSM data extraction, and eventually it hopefully will cover every OSM feature, possibly as extensions (there is a proposal for such arrangement as well).

If I'm going to validate scheduled monuments in Scotland (8,086 items), I need to be able to see archaeological sites.

http://mapper.acme.com/?ll=60.68397,-0.98057&z=16&t=M

https://maps.wikimedia.org/#16/60.6852/-0.9790

And I need to be able to see field boundaries b/c many features of interest, sufficiently notable to be items, are "that lump in a field over these":

https://canmore.org.uk/site/search/result?SITECOUNTRY=0&view=map
Canmore: OS maps showing blue circle features against field boundaries. (Papa Stour, Shetland)

Canmore - next level of detail

OSM/ACME - there be field boundaries
https://mapper.acme.com/?ll=60.32680,-1.68310&z=16&t=M

Wut?
https://maps.wikimedia.org/#16/60.3268/-1.6855

Styles that Wikimedia maps use, osm-bright, is supposed to display some additional elements, including buildings, but lately these elements got lost on Wikimedia maps as apparently styles aren't yet configured to work properly with new data import tool (T288731).

Some more types of elements might be considered, but I doubt that we really want to display everything that openstreetmap-carto style has, which I assume "OSM default feature sets" stands for. The latter styles are used to showcase nearly all data that OSM includes. While Wikimedia maps are most commonly used as an outline map on which markers and particular line or areas are highlighted. For the latter purpose different styles are better suiting, i.e. syles that aren't as overcrowded with information and where colours aren't as contrasting.

Alternatively you might ask for better support of different map layers with different styles, including styles with more information. Some gadgets and user scripts, such as the one you get if you click "Kaart" at the top of this Wikipedia article, already allow selection of both Wikimedia maps and tiles from openstreetmap.org.

T285668 may be related

In T289101#7290502, @Pikne wrote:

Alternatively you might ask for better support of different map layers with different styles, including styles with more information.

Done in T215073 - two years old next month.

In T289101#7290502, @Pikne wrote:

While Wikimedia maps are most commonly used as an outline map on which markers and particular line or areas are highlighted. For the latter purpose different styles are better suiting, i.e. syles that aren't as overcrowded with information and where colours aren't as contrasting.

That's a view, but it's a nonsense view. Maps which show what is on the ground are useful. Maps which do not are not so useful. I've set out very clearly why detail supports my use cases.

What is your use case for "syles that aren't as overcrowded with information"? Is the blank Papa Stour map sufficiently uncrowded for your taste? smh.

I'd like to remind everyone we have no Maps developers. The product is still frozen, the team is only making some changes to make the original status quo more maintainable (which next to fixing a few issues, apparently caused some additional breakages too), but it is unlikely they'll pick this specific ticket up. Please vote in the board elections for someone who cares to fund development of features like these.

Of note is that we ALSO no longer have active volunteers for any map related work on tool forge.. so .....

In T289101#7291155, @Tagishsimon wrote:

In T289101#7290502, @Pikne wrote:

While Wikimedia maps are most commonly used as an outline map on which markers and particular line or areas are highlighted. For the latter purpose different styles are better suiting, i.e. syles that aren't as overcrowded with information and where colours aren't as contrasting.

That's a view, but it's a nonsense view. Maps which show what is on the ground are useful. Maps which do not are not so useful. I've set out very clearly why detail supports my use cases.

I'm referring to most maps in Wikipedia that highlight markers or some objects above Wikimedia maps, such as examples given for Maplink template. On such maps highlighted objects generally come forward better if on the background there aren't too many distracting objects or other contrasting objects. That said I don't claim that Wikimedia maps shouldn't display more objects than it currently does, it should display at least buildings and other features that were present on it recently (ticket referenced above).

Different map styles, including those with different selection of information, obviously fit for different purposes. If we'd try to create map styles that fit for all purposes, then likely it wouldn't fit for anything. Hence I'm just saying that we should first figure out what given map styles are mainly for. Your use cases are fair use cases, but some of them might be better served using entirely different map layer with different styles and information.

Can you describe any use cases, @Pikne, which are frustrated by the normal convention of showing progressively increasing amounts of information as the resolution of the map increases?

@TheDJ noted. It's up to the WMF to decide whether they're going to support users and improve the product in obvious ways such as the above; or whether they're just going to sit back and count their astronomically huge pile of money. It's really only all the users and all the editors who are currently being ill-served. it doesn't take much imagnation to see that wikimedia would be better served by appropriately detailed maps, and so the question is, why is this a backwater for WMF? I live in hope that WMF, replete with a plurality of communications managers, gets back to us on this one.

@Pikne this is not accurate -- while it does help to reduce the number of OSM features to some extent, it is still far better to include more. What we do want is not to keep small number of features, but rather style map in such a way that it doesn't get in the way of the overlayed content. Here are some sample styles OpenMapTiles has, and I think Positron style would be a better fit than the bright style.:

One other thing that few mentioned -- current maps in Wikipedia are NOT open source. You cannot take the code and run it yourself -- Mapbox will sue you because they claimed copyright over the DATA STRUCTURE of how data is organized in the vector tiles/styles. Another reason to switch to OMT.

In T289101#7291465, @Tagishsimon wrote:

Can you describe any use cases, @Pikne, which are frustrated by the normal convention of showing progressively increasing amounts of information as the resolution of the map increases?

Why do you ask? As far as I can see this convention, which I don't have a problem with, is already followed on Wikimedia maps: as zoom level increases, more place names, more roads, more waterways etc. appear. Hopefully buildings and some more objects will also reappear at higher zoom levels once bug that I mentioned above gets fixed.

@Tagishsimon: Please feel free to bring up meta-level discussions about WMF at better suited places - thanks a lot! :)

Now buildings and some other objects appear again on Wikimedia maps.

In relation to what to add and what to omit, see also T113912: Outline cartographic goal for style.

following roughly the pattern established by OSM as their default tile set

Note that this goal can be quite dangerous. Trying to display everything ends with quite severe legibility and understandility problems. You can render plenty of things but it makes much harder or outright impossible to clearly differentiate and to keep map readable.

For reference, I am one of authors of OSM Carto (OSM default map style).

Though it is quite irrelevant anyway ( https://phabricator.wikimedia.org/T215073#4922001 "a developer (@Pnorman) was hired to work on better display of the disputed territories, and after over a year of work, after it essentially being 90% complete, WMF decided to stop the project without any explanation to the community." )

Of note is that we ALSO no longer have active volunteers for any map related work on tool forge.. so .....

Given that work funded by WMF is blocked by WMF it is not a surprise that noone is volunteering to make work that will be ignored.

In T289101#7594363, @Mateusz_Konieczny wrote:

Given that work funded by WMF is blocked by WMF it is not a surprise that noone is volunteering to make work that will be ignored.

No, i mean that a substantial set of currently existing tools used by the community, that were developed outside of WMF (but running on WMF hardware) are unmaintained.

Maintainers have left or take over half a year to respond. What this means is that the community no longer has the capacity to fill the gaps that the foundation is leaving either. The expertise has vanished.

Im mostly talking about things like the various poi apis, but also the community tile layer servers. The community osm database isnt even up to date and no one knows how to debug the problems.

@TheDJ this does feel like an egg-and-chicken problem, does it not? If we take maps seriously, and someone who wants to + is able to implement clean solutions reaches out to map enthusiasts in our community, like Mateusz and others, we could consolidate maintenance forces. Good WP integration remains valuable to the maps ecosystem too.

To be clear: egg-and-chicken problem is even worse and WMF taking maps seriously would not be sufficient for me to contribute beyond some feedback like above about not trying to fit as much as OSM Carto (OSM default map style), unless you want the same problems or figured out way to avoid them.

I am definitely not planning any substantial time for any WMF-led projects as long as WMF budget completely lacks transparency. I am not going to work[0] for free for organization with 150 000 000 $ revenue[1] refusing to explain how money is being spend.

Materials at https://wikimediafoundation.org/about/financial-reports/ completely lack explanation where money actually ends.

"Salaries and wages" (67 million) may be covering anything from board paying tens of millions to itself in salaries, through paying programmers to develop projects unwanted by community to maintaining actually needed projects.

[0]I am an active contributor at Wikipedia/Wikimedia Commons/etc with activity including removal spam and improving things that I need to exist, WMF is not so bad to avoid project it is controlling out of spite and communities around this projects are great. But sadly WMF is not a part of this communities.

[1] https://en.wikipedia.org/wiki/Wikimedia_Foundation