I didn't see your second post about musem item Q20196262 right away. To me link to that Wikidata ID doesn't seem more suitable either. Statements in that item currently somewhat conflate the building and museum (organization) but it seems that the intended subject in that item is rather the organization. Wikipedia articles attached to it say that the museum is housed in the Berlin Palace.

Regardless whether previous and current building should be considered the same, item Q1636886 currently includes data about current building, most importantly Wikipedia articles attached to it (including dewiki article) use present tense to define its subject. I see that contradictinly a couple of statements in that item also suggest that the building is demolished, which is confusing indeed.

Nothing in item Q1636886, nor in attached Wikipedia articles, on the other hand suggest that that item is for a building. That item is for particular reconstruction process that happened in the past.

This doesn't seem right. You changed Wikidata ID of the building (Q170119) to Wikidata ID of reconstruction (Q1636886).

I don't know which name form exactly you refer to "regarding Ukraine". Every recommendation of course doesn't have to be follwed and is not follwed. News outlets and other publications can make an editorial choice to follow or not follow a recommendation, at least in countries like Germany and Poland this should be the case.

Anyway, we can really rely only on verifiable information, like usage analysis in reliable sources, not on "most people, as far as I know". Sure, some (old) names are/were in common use for long time but language can still change, whether we like it or not.

Whatever are the reasons (campaign or not), name usage has definitely shifted, a while ago already. It is reasonable to analyze usage based on recent media texts just because these texts are usually copy-edited, and are therefore generally more reliable compared to random texts on the internet.

It doesn't matter, Latin-script languages normally use endonyms of other Latin-script languages as are, without simplification. "Klaipeda" is an exception: as said, is different as this name form (without ė) is actually common in German-language texts.

name:et should be corrected, too. I'll do it later if someone else doesn't get to it before. Whether we like it or not, that's the same of this town. "Kingissepp" by the way is a genuine Estonian-language name, Russians have just transliterated it.

"Belarus" is about 40 times more common, is what I wanted to say.

It's a controversial edit, as you very well know in light of recent discussions. It probably needs to be undone.

German Wikipedia prefers "Belarus", as per Wortschatz it is much more common than "Weißrussland" in modern German, in recent news texts about 40 more common: https://corpora.uni-leipzig.de/de/res?corpusId=deu_news_2025&word=Belarus

What "ASCII rules"? German alphabet itself includes several non-ASCII characters that are definitely not avoided in German names in OSM.

If you perhaps have some other names like "Riga" in mind then this is just a traditional name in German language, not due to some ASCII rule. As per Wortschatz a few others like "Klaipeda" are also fairly common, though "Klaipėda" is not significantly less common these in German. But "Jekabpils" is rare, there probably isn't a good reason to prefer it over an endonym.

You refer to a news article about proposal(s) that were never adopted. Also the propsal (or rather an idea of possible proposal) was about recommended name in Estonian language, irrelevant to common name in German language that you edit here.

Some back and forth between bicycle=designated and bicycle=yes has been going on here for way/76559569. Previously we established that rather bicycle=designated should be kept, see changeset/144441885

This ("Estonian government recommends") is quite surely false, and it also unclear why any recommendation by Estonian goverment is relevant to German name here.

Whatever source you have in mind here, you most likely misunderstand it. If you perhaps refer to Estonian language normative dictionary then this provides "Jamburg" only as historical (ajalooline) synonym for "Kingissepp" (https://sonaveeb.ee/search/os/Jamburg/1).

Also, I doubt that it's a good idea to add name translations like "City center" and "Keskusta" for the city district (changeset/170331119). That kind of place names are usually left untranslated. E.g. respective Wikipedia article titles are not translated. These translations are literal and descriptive, they may as well be used for some smaller area not the entire city distirct, or (in some other occasion) possibly also for some larger area.

Here and in some preceding changesets you moved label nodes for "Tallinn" and "Kesklinn" closer and closer to one another. Now they are about exactly in the same spot. Why is this better than before? Current placement doesn't seem more accurate in any way while now it is difficult to render both names concurrently.

Outside of cities in Estonia normally highway=tertiary is used for roads of type "kõrvalmaantee ", as documented here: osm.wiki/International_highway_classification_equivalence
Hence these road segements also were classified like that.

At least for this example "... ēka" seems to be used occasionally, based on Google search. But I'm not saying that every building needs to be named like that of course, nor that it is necessarily needed in this particular case. Not every building where some organization(s) operate needs to have a name at all.

Label for Wikidata item of given building can be improved fairly easily, as said. Other than that it seems clear to me that given item is for the building. Long gone gymnasium is subject of third item in Wikidata, linked from building item as previous occupant.

I agree that consistency is important but at the end different name variants exist in actual usage, and we probably can't force Wikidata and OSM to prefer exact same name variants in every occasion. In regard to consistency, I find it more important how the data is modelled, e.g. whether consumers get building data (not organization data) if they query for buildings.

Yes, occupant's name is often also used for the building. Which is why I mostly kept "name" keys untouched here.

I think it isn't particularly imporant that OSM name and Wikidata label were exactlly the same, as long as based on other attributes it is clear which object/item is for (primarily) for the building and which for the organization. But if distinctive name is needed for the building then I suppose for this example "Jāzepa Vītola Latvijas Mūzikas akadēmijas ēka" is an option.

Yes, the idea is to clear up cases where Wikidata ID is for building's (current) occupant, not for the building itself. In number of cases, like way/39432730 mentioned above, separate item exists for the building.

Also schools and other organization occasionally operate in multiple buildings and there should be 1 to 1 correspondence between OSM and Wikidata objects (see also "one feature, one OSM element" principle). Here I mostly only moved the data to keys with operator: prefix but I think it would be actually better to add all organization data consistently to separate OSM objects, to a node within building, or to wider area (not individual building) if some compact and delimited area like school perimeter around school buildings exists.

As for "name" key values and Wikidata labels, these can be further refined. Label for this example probably should be something else, an actual and natural name used in modern language, rather than a description from cultural heritage register ("Aleksandra ģimnāzija, tagad - augstākā mācību iestāde").

It is unclear what you intended to do here. These three latter inner elements should not have been addded to the building relation as actually there isn't a courtyard containing a building that has another courtyard. Also the amphitheater probably shouldn't be considered a building part.