The Navajo Nation and Cherokee Nation both have flags to use that could theoretically be incorporated into Unicode if they didn't draw the line at National flags, but you could always draw flag representations in SVG instead.
But I also think this discussion has gone so far into the theoretical it's kind of ignored a very practical point: how many languages do you need to present in your software, because your software is actually localized in that language? Most software isn't going beyond the top two or three recognized languages in a given UN-recognized country (and even some of those are relatively obscure), and countries that have >1 languages usually share an official language with another country that uses it more predominantly. At this point in time, the economies of scale do not favor actually localizing software in 1600 different languages, so we don't. At all.
In fact, the usual pattern for developers from Europe and the Americas is to favor their dominant language and localizing in other dominant languages that originated in Europe and some of their more popular variants, and then maybe also some combination of Simplified Chinese, Traditional Chinese, Japanese, Arabic, Hindu, Urdu, and Korean.
So when you're presented with a number of languages and variants exceeding 30 in a list, flags probably aren't going to cut it anymore, but there's still some design space to use flags as language representations if you want.
I am already aware of that, but stating that here does not contradict what I said, neither that line specifically nor that line within the full context of my comment. Indian software written by Indians for the Indian market might be different though, but I'll defer to any Indians that want to chime in on that subject themselves.
It's irrelevant. The topic is a defense of equating a country flag with a language. Even if you have 2 in one country then you've already proved why its a bad idea
The thing that makes it a bad idea is ambiguity. But most types of ambiguity in this situation only apply to the list of supported languages. If the languages aren't supported, the problem doesn't exist. So yes it's relevant.
And there's only a couple countries in the world where this is a serious issue anyway. It doesn't make flags a terrible idea for the rest.
The Navajo Nation and Cherokee Nation both have flags to use that could theoretically be incorporated into Unicode if they didn't draw the line at National flags, but you could always draw flag representations in SVG instead.
But I also think this discussion has gone so far into the theoretical it's kind of ignored a very practical point: how many languages do you need to present in your software, because your software is actually localized in that language? Most software isn't going beyond the top two or three recognized languages in a given UN-recognized country (and even some of those are relatively obscure), and countries that have >1 languages usually share an official language with another country that uses it more predominantly. At this point in time, the economies of scale do not favor actually localizing software in 1600 different languages, so we don't. At all.
In fact, the usual pattern for developers from Europe and the Americas is to favor their dominant language and localizing in other dominant languages that originated in Europe and some of their more popular variants, and then maybe also some combination of Simplified Chinese, Traditional Chinese, Japanese, Arabic, Hindu, Urdu, and Korean.
So when you're presented with a number of languages and variants exceeding 30 in a list, flags probably aren't going to cut it anymore, but there's still some design space to use flags as language representations if you want.