4 ms·

oh yes, let's break things to gain nothing of value

gspr 2y ago
Perhaps nothing of value _to you_. I'll hazard a guess that your preferred username can be expressed in a small subset of ASCII? And to hell with everyone else?
- knorker 2y ago
  I'll hazard a guess that your preferred username can't be written by 99.99999% of the world, and would always have to be copy-pasted?
  - Ylpertnodi 2y ago
    Yeah, us foreigners, up to our usual tricks again.
    - knorker 2y ago
      By any definition of the word, I'm a foreigner. So if you meant to imply that I'm an American, you've guessed wrong.
  - gspr 2y ago
    My preferred login name, and in fact also my legal name, are ASCII safe. I've never personally had to fight computer systems built with ridiculous assumptions (in the domain of names, at least). What I'm exhibiting in this thread is called empathy for people who don't happen to land in that category.
    - knorker 2y ago
      Did you appreciate my comment's real meaning, that one can have empathy for everyone else too, who need to work with this data? 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.
- chikere232 2y ago
  If your personal identity is threatened by having to use an ascii alphanumeric login name, you're kind of creating problems for yourself for no reason... 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
  - gspr 2y ago
    You're moving the goalposts here. 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?
    - chikere232 2y ago
      Sure, if you read "nothing of value" as "no one anywhere will get even any perceived value from this ever", but that is disingenious 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
- anon-3988 2y ago
  Its a give and take. If you allow for anything beyond latin, then you have to accept that there will be a class of software that will be difficult to interact with. 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.
  - chikere232 2y ago
    It gets real fun when it's something you need to look up and have match, like a username. 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
    - khaled 2y ago
      > And unicode is an evolving standard where this normalisation sometimes changes between standards Unicode normalization is subject to its stability policy, and Unicode no longer allow adding new canonically equivalent code points. https://www.unicode.org/policies/stability_policy.html https://www.unicode.org/policies/stability_policy.html
  - adrian_b 2y ago
    There are many variants of the Latin alphabet and the English alphabet contains only a subset of the letters contained in the other variants. 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.
layer8 2y ago
The issue is that it has already been broken (read: has allowed arbitrary byte sequences) for a long time, and the debate is about what to restrict it to.

gspr 2y ago

Perhaps nothing of value _to you_. I'll hazard a guess that your preferred username can be expressed in a small subset of ASCII? And to hell with everyone else?

knorker 2y ago
I'll hazard a guess that your preferred username can't be written by 99.99999% of the world, and would always have to be copy-pasted?
- Ylpertnodi 2y ago
  Yeah, us foreigners, up to our usual tricks again.
  - knorker 2y ago
    By any definition of the word, I'm a foreigner. So if you meant to imply that I'm an American, you've guessed wrong.
- gspr 2y ago
  My preferred login name, and in fact also my legal name, are ASCII safe. I've never personally had to fight computer systems built with ridiculous assumptions (in the domain of names, at least). What I'm exhibiting in this thread is called empathy for people who don't happen to land in that category.
  - knorker 2y ago
    Did you appreciate my comment's real meaning, that one can have empathy for everyone else too, who need to work with this data? 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.
chikere232 2y ago
If your personal identity is threatened by having to use an ascii alphanumeric login name, you're kind of creating problems for yourself for no reason... 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
- gspr 2y ago
  You're moving the goalposts here. 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?
  - chikere232 2y ago
    Sure, if you read "nothing of value" as "no one anywhere will get even any perceived value from this ever", but that is disingenious 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
anon-3988 2y ago
Its a give and take. If you allow for anything beyond latin, then you have to accept that there will be a class of software that will be difficult to interact with. 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.
- chikere232 2y ago
  It gets real fun when it's something you need to look up and have match, like a username. 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
  - khaled 2y ago
    > And unicode is an evolving standard where this normalisation sometimes changes between standards Unicode normalization is subject to its stability policy, and Unicode no longer allow adding new canonically equivalent code points. https://www.unicode.org/policies/stability_policy.html https://www.unicode.org/policies/stability_policy.html
- adrian_b 2y ago
  There are many variants of the Latin alphabet and the English alphabet contains only a subset of the letters contained in the other variants. 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.

layer8 2y ago

The issue is that it has already been broken (read: has allowed arbitrary byte sequences) for a long time, and the debate is about what to restrict it to.