oh yes, let's break things to gain nothing of value
I'll hazard a guess that your preferred username can be expressed in a small subset of ASCII? And to hell with everyone else?
Latin-like language system is simply superior for machine purposes. I am sorry, but I don't even want to think of supporting the entire unicode in my software. I am not going to even attempt to reverse that emoji.
There is no reason to consider the English alphabet as superior for machine purposes, in comparison to other Latin alphabets.
Its dominance in IT is caused by the fact that most of the development of commercial computers after WWII has been done at IBM and other US companies, not by any properties of the English alphabet.
Because then it to be normalised in the right way for comparisons to work, or it will only match if your input method happens to produce the exact same variant.
... And unicode is an evolving standard where this normalisation sometimes changes between standards, so the names as normalised in the old version of your standard library might disagree with the new version. So you need to care for that transition.
... And often this is implemented separately for different languages, so you can get names that won't match if you normalise them in python, java or C.
... And as all implementations, these unicode implementations sometimes have bugs, so you need to think not only about matching supported unicode versions, but matching bugs.
... And any change in these normalisations can in theory lead to two usernames that used to be distinct becoming identical.
It's a deep well
Unicode normalization is subject to its stability policy, and Unicode no longer allow adding new canonically equivalent code points.
For whom? Presumably people won't choose a username themselves that is difficult _for them_ to input. It's not like allowing (a subset of) Unicode would force people with near-ASCII keyboards to deviate from what's easy for them to input.
> Latin-like language system is simply superior for machine purposes.
Citation needed. Besides, on the systems we're talking about at least, the machine itself works with integral UIDs, not user names.
> I am sorry, but I don't even want to think of supporting the entire unicode in my software.
What does that even mean? Why is your software doing operations on people's names, beyond storing and retrieving them?
> I am not going to even attempt to reverse that emoji.
Is there any legitimate use case for reversing someone's name??
Yes, when you want to reverse sort a list? In fact, how do you even sort such a name that goes in both direction?
> What does that even mean? Why is your software doing operations on people's names, beyond storing and retrieving them?
A lot of software deals with text. I don't think every software should deal with the entirety of Unicode. There shouldn't be an expectation that all text editor should handle all Unicode. Just like I don't expect a Chinese keyboard and software should work for Latin keyboard and software. In my personal code editor, I handle text byte per byte, so text highlighting, reversing, lookup, etc are simple and efficient. Would it be a problem if I try to open a Chinese text? Yes, but I have accepted that it is a class of problem that I don't want to deal with.
I have not seen an open source software that have gracefully handled Unicode, outside of Chrome or something really big. Pretty much every one of them is "errr how does this work again?". How many times have people simply split a text using whitespace? So lets just say I don't have confidence in people doing this correctly. I know I don't.
There is a field for the full name of the person if you want to, and at least on my linux it warns for non-ascii characters but allows them
You started off with saying that there's no value to login names beyond ASCII, and now you're (a) insinuating that anyone who would wish to have login names beyond ASCII are too heavily invested in their login name, and (b) there's technical downsides to this.
Point b is certainly valid, and the whole reason there's a discussion in the first place. Point a might be true, but very different from what you originally wrote. And again, I suppose, this doesn't apply to you, because your login name – which you isn't attached to at all – is naturally ASCII safe?
What I mean is that it's costly and the benefits are at best emotional.
And if you personally assign a large value to having unicode characters in your login name specifically, I posit that you have the wrong priorities in life. You're free to disagree with that opinion
Another solution is, of course, to give everyone usernames they don't want. E.g. your user name is your employee number. Or your department plus three letters from your name that happen to make your username unique. I've seen both.
I also extend my empathy to those who, after choosing a username nobody can type, don't get invited to meetings, don't get messages, are never referred to as the domain expert, because how can I add "习近平" to a meeting, if I can't even search for their username?
Say even "mårten" is not showing up when I search my email, because those are not the same letters as "marten". Meh, emacron knows almost as much on this topic, so I'll just invite him instead.
If you're already powerful enough, then people will be forced to copy-paste "mårten". If you're not already powerful enough, then you don't get invited at all, because that doesn't scale.
So if you meant to imply that I'm an American, you've guessed wrong.