There is this issue on GitHub: https://github.com/Plume-org/Plume/issues/228 to make clean slugs/identifiers/URL for users, blogs and articles. The problems are that allowing any UTF-8 character (= any character that exists in the world) is not nicely displayed by many websites when making links (you end up with %-encoded characters everywhere), and that it makes it easier to do phishing/to impersonate someone else, as you can find visually similar characters quite easily.
And as I explained there, I don't know how we should deal with that.
We could only allow ASCII (= english alphabet) characters, but what about people using languages with accents, or even worse, people who don't use the latin alphabet at all?
We could replace too visually similar characters with their ASCII equivalent, but that would require a huge amount of work (there are thousands of characters in UTF-8, how can we handle them all?)
We could also decide to just take the risk of phishing or impersonation. It will always be more or less possible anyway, since you can always create an account with the same name on another instance if you want.
Or maybe there is a better solution?
What do you think?