I don’t know how the tags are handled internally, so I am creating this issue as it feels that on the off chance it could lead to something serious.
Not sure what I did exactly but I was playing with various tag casing in AnkiDroid and while it seemed to be consistent there, after syncing with Anki I now have a card with the tag ẞ that doesn’t appear in the Browser tag list, which instead has the tag ss that no cards have.
Some sort of compatibility conversion must have happened. ß is the german letter “scharfes S” (or “Eszett” in North Germany) and ss is usually used when the letter is not available.
I suspect a part of the problem is that ẞ can be lowercased to both ß and ss, and wasn’t even a part of Unicode until 2008 so some libraries might not be able to properly deal with it
If you add a card with tag A, A is added to the tag list. If you then add a card with tag a, it is renamed to A to match the existing tag list entry, and no separate entry is added to the tag list - whichever case is first used is treated as canonical. The ss sounds like an AnkiDroid issue that will presumably be resolved when it starts delegating this handling to the Rust code.