LouisGac wrote: I don't think the problem will be in DB, we didn't made any change in DB collation since a long time, and it's UTF8mb4:
I still won't ruleout that the issue is not related to LS directly, but lays in the setup of database and access libraries (OS+PHP binaries). The issue occurs, when first using the different language, which demands a broader charset. For bugreports it is important, to know, if the user is already using the language or not. Here it is first time usage.
The japanese bug report seems to contradict my idea with the database.
bugs.limesurvey.org/view.php?id=13904
Some fields are correctly shown. Some not. Perhaps different tables?