So... Do you mean that because it's hard to protect one against it, it won't be done? I'll be mean for a second, but aren't you circling around the problem?

I would understand if you told that it's on the bottom of the priority list (for instance due to high-cost/low-profit, for instance). But without this added message, it looks like you're more not fond of in tackling that issue, so you reject it on the form rather than the substance. A bit like a sports fan would be annoyed about having to go to the supermarket to resupply on potato chips just before a soccer match, because they are already on the couch and the trip is too long. It's something I understand, people have one thing they hate doing; And this thing doesn't seem particularly fun, but...

Also, is it really that hard? It's a common and long known web problem as users are everywhere now, so... And hackers, too. They like weird characters, as far as I know. There surely are a lot of libraries doing that, right? Right?

Tortliena‭ The point is that there's nothing to protect against. There's simply no way to harm using just Unicode. After all, it is ultimately just text. At most, they can create funny looking text, but that's the extent of it.

Moshi‭i Do you think there is no way to harm or this is a normal behaviour when -I quote ArtOfCode- it is something "easy to abuse" and that you should "expect your username to be reset if you do (use weird letters)"?

My point is, yes, unicode is working as intended, strictly speaking; However, it's just a structure. It isn't the features it serves as the soil of : User creation, search, @-ting people... That's what actually matters to the eyes of the asker, mine, yours on the other post, and maybe ArtOfCode, based on how they said things. After all, we don't build things that are bug-free but impractical, annoying or useless, right :p?

Tortliena‭ Here are some things people believe about names. All are false. The point is not that I don't want to do it, it's that there's no way to do it (a) in a sane manner, (b) without breaking important functionality for some use cases (non-ASCII names), and (c) in a way that we won't have to keep fixing when someone finds another weird way to break it.

Now, in terms of mitigation, there are things we can look at doing - several people have mentioned the three-character ping trigger, which we could change; we could look at bringing up a list of any non-ASCII names for easy pinging. You're right, every system deals with this problem - and most just do so by building in workarounds instead of trying to "fix" Unicode.

@ArtOfCode Indeed, they're mitigations, not really something which solves the core of the issue (which isn't unicode per say). You know you (or the next one) will have to work on it later, and face the issues of restricting something which wasn't, both technically and socially. And this, even if it means straying out of this ideal of encompassing all possible user names.

But in the end, it's you who choose to cut the leaves and branches or uh... unroot the roots. Not me, so do whatever you wish!