I have great respect for the work of others, so I will never interfere with anyone. But I do not have a personal interest in starting a battle, I just expressed my opinion.
I am writing these lines for those who are also looking for all useful tags.
In my opinion, the service https://taginfo.openstreetmap.org/ is completely unhealthy.
So I myself continued to create such an equivalent. The basis of the stat consists of elements that have 3+ different names (in different languages)
From the stats, the names and wikis were deleted:
^name|alt_name|int_name|old_name| wikipedia
Then tags were removed from the stat, the values of which are numerical, they are not interesting to me.
Removed tags with fewer than 3 occurrences, as well as keys, there are a lot of garbage or erroneous (from my point of view)
Here is a sample with 3+ languages on the label:
place (284617)
= village (142810)
= locality (59632)
= hamlet (47668)
= town (17516)
= city (4175)
= suburb (3796)
= region (2138)
= county (1911)
= island (1214)
= state (1060)
= isolated_dwelling (1014)
= neighbourhood (876)
= country (224)
= farm (172)
= islet (114)
= quarter (93)
= sea (77)
= yes (23)
= province (14)
= allotments (12)
= municipality (9)
= subdistrict (9)
= peninsula (8)
= continent (7)
= archipelago (7)
= district (6)
= ocean (5)
= islands (4)
= bay (3)
Surprisingly, very normal statistics, which I believe. And this is exactly the tags that I was looking for and what I expected to see. More: https://a.safe.moe/VJzmf.txt
If I increase the number of languages, then the stat changes, the spaceports and volcanoes disappear already by 10 (the spaceport in general is one - the Baikonur, so it is removed from the sample), but on the other hand, such “top sampling” is also interesting
place (1880)
= city (812)
= town (335)
= state (302)
= country (223)
= sea (59)
= village (58)
= locality (26)
= suburb (19)
= island (11)
= region (10)
= county (8)
= continent (7)
= ocean (5)
= hamlet (3)
is_in:continent (911)
= Europe (329)
= Asia (269)
= Africa (128)
= South America (71)
= North America (68)
= Oceania (22)
= Central America (14)
= Australia (9)
is_in:country (829)
= Japan (62)
= China (50)
= Germany (44)
= Italy (37)
= Spain (34)
= France (28)
= Russian Federation (26)
= България (25)
= USA (24)
= Brazil (23)
= Україна (22)
= Republic of Korea (17)
= Brasil (16)
= Belgium (15)
= Poland (11)
= United Kingdom (11)
= Lithuania (11)
= Canada (11)
= România (10)
= Danmark (10)
= Chile (9)
= Czech Republic (9)
= Россия (9)
= Austria (9)
= Switzerland (8)
= Iran (8)
= The Netherlands (8)
= South Africa (7)
= India (7)
= Greece (7)
= Australia (7)
= Algeria (6)
= United States (6)
= Georgia (6)
= Ukraine (5)
= Egypt (5)
= Bolivia (5)
= Slovenia (4)
= LV (4)
= Saudi Arabia (4)
= Cyprus (4)
= Russia (4)
= Colombia (4)
= Argentina (3)
= Mauritius (3)
= Croatia (3)
= Abkhazia (3)
= Vietnam (3)
= Uzbekistan (3)
= North Korea (3)
= Syria (3)
= Hungary (3)
= Pakistan (3)
= Syria سوريا (3)
= Norway (3)
= Afghanistan (3)
= Denmark (3)
If someone is interested, then you can see the sample https://a.safe.moe/nfC1s.txt - the features of mapping in different countries are clearly visible
If you are interested in a selection of different languages, I created different selections up to 30: https://a.safe.moe/YItYl.zip
Immediately I say that my data is old, I do not have a space for a new version of the planet. However, it looks like it’s not necessary for anyone. How to use Taginfo, I still do not understand, it totally not useful for me (and looks what data wrong).