In DNS, the domain name has to be ASCII, so unicode characters in the domain name are converted to Punycode and prefixed with xn--. So really, blåhaj.com is really xn--blhaj-nra.com (put that in your browser and watch the name change).
I would imagine that most things would just work, but there would probably be some annoying bugs with different clients who aren’t using libraries which support internationalized domain names, or aren’t expecting them. It’d probably be a good thing to have an internationalized domain name for a popular instance, as that would be a good test case for servers and clients to support that standard.
I’m personally convinced limitations like this are why English is becoming such a dominant language, because the internet and most coding was all designed in English for English, without consideration for other languages. Other languages have to get tacked on with semi-complicated workarounds like this.
Adapting keyboards to non-alphabetic languages for example. Forced Chinese to adopt a romanized way to type things out and learn a new alphabet.
Although they actually eventually became some of the fastest typists on earth thanks to the predictive auto-complete (as I understand it) they adopted decades prior to the rest of the world using it on phones.
It was actually more that early computers had such limitations on compute and memory that they had to limit what they could work with. The internet is still built from the ground-up on tech from 40-50 years ago.
It can appear and act as cultural imperialism, but there’s actually a lot of evidence that it was just people making a new thing and not thinking that far forward with what they were doing.
Usually cultural imperialism is considered something a society does on purpose to force other societies into their mold. There is plenty of evidence this one is honestly purely accidental and not purposeful.
I mean, I don’t blame the creators. They didn’t know how these things could affect others nor know their standards would last so long.
But the US and other Western powers became the tech powerhouses they did through imperialism and exploitation, so really these things can be viewed as an extension of that imperialism.
Intentional? Perhaps not. But it’s still reality that it makes things harder for much of the world, and it’s something that should be pushed back against.
Never attribute to malice that which is adequately explained by stupidity.
All of this stuff is 50 years old and there’s no way the people who put these systems in place could have ever predicted the Internet as it is today. Famously in 1973, there was a “map of the ARPA network”, that’s how few machines were “online” at the time.
It’s really only in the past 10 years or so that security via TLS has been widely adopted by websites, and it’s been over 20 years since IPv6 was announced and still is not entirely supported everywhere. There is so much inertia with the underlying infrastructure of the internet, even if there is a serious issue (like DNS cache poisoning ), systems will not be reworked and replaced, but hacks will be added to fix the issue of the day.
Agreed, I just said the same elsewhere. Early computing was filled with memory and compute limitations that meant having every language under the sun supported was just not physically possible.
In DNS, the domain name has to be ASCII, so unicode characters in the domain name are converted to Punycode and prefixed with
xn--. So really, blåhaj.com is reallyxn--blhaj-nra.com(put that in your browser and watch the name change).I would imagine that most things would just work, but there would probably be some annoying bugs with different clients who aren’t using libraries which support internationalized domain names, or aren’t expecting them. It’d probably be a good thing to have an internationalized domain name for a popular instance, as that would be a good test case for servers and clients to support that standard.
I’m personally convinced limitations like this are why English is becoming such a dominant language, because the internet and most coding was all designed in English for English, without consideration for other languages. Other languages have to get tacked on with semi-complicated workarounds like this.
Cultural imperialism.
Adapting keyboards to non-alphabetic languages for example. Forced Chinese to adopt a romanized way to type things out and learn a new alphabet.
Although they actually eventually became some of the fastest typists on earth thanks to the predictive auto-complete (as I understand it) they adopted decades prior to the rest of the world using it on phones.
It was actually more that early computers had such limitations on compute and memory that they had to limit what they could work with. The internet is still built from the ground-up on tech from 40-50 years ago.
It can appear and act as cultural imperialism, but there’s actually a lot of evidence that it was just people making a new thing and not thinking that far forward with what they were doing.
Usually cultural imperialism is considered something a society does on purpose to force other societies into their mold. There is plenty of evidence this one is honestly purely accidental and not purposeful.
I mean, I don’t blame the creators. They didn’t know how these things could affect others nor know their standards would last so long.
But the US and other Western powers became the tech powerhouses they did through imperialism and exploitation, so really these things can be viewed as an extension of that imperialism.
Intentional? Perhaps not. But it’s still reality that it makes things harder for much of the world, and it’s something that should be pushed back against.
Abso-lutely.
All of this stuff is 50 years old and there’s no way the people who put these systems in place could have ever predicted the Internet as it is today. Famously in 1973, there was a “map of the ARPA network”, that’s how few machines were “online” at the time.
It’s really only in the past 10 years or so that security via TLS has been widely adopted by websites, and it’s been over 20 years since IPv6 was announced and still is not entirely supported everywhere. There is so much inertia with the underlying infrastructure of the internet, even if there is a serious issue (like DNS cache poisoning ), systems will not be reworked and replaced, but hacks will be added to fix the issue of the day.
Agreed, I just said the same elsewhere. Early computing was filled with memory and compute limitations that meant having every language under the sun supported was just not physically possible.