3

RFC6530 defines the necessary steps for "international e-mail" (i.e., especially for UTF-8 e-mail addresses). Apparently Google adopted the RFC back in 2014 (source). Still, most validators I find on the web are having trouble with international e-mail addresses to say the least. One very popular contender is HTML5's integrated e-mail validator of type="email" fields, which does NOT support international e-mail addresses at all.

So my question is: are international e-mail addresses a thing or is this "dead technology"?

We're writing test data generators and are thinking about extending our e-mail address generator but are unsure whether to include international e-mail addresses as well. We could make it a configurable option of course, still, if this technology is actually dead, we are not going to include it.

Community
  • 1
D.R.
  • 231
  • 1
  • 5
  • 1
    State your intended market. Views and conclusions may be quite different from "ASCII only" U,S, "almost ASCII" Germany and "no ASCII" China. – AnFi Nov 30 '18 at 09:55

2 Answers2

3

I share MrSMith42's experience that anytihn outside of the most simple email has a hard time getting used in many systems that currently exist, however I disagree with their assesment.

As it stands the RFC is not widely used, but as the internet keeps booming and non English markets keep gaining in force in the world rather than simply call the RFC "dead technology" I would refer to it as a "future technology".

By refering to it this way I am saying that it should not be outright dismissed but rather than your systems should be set up to deal with such emails, but that for the time being you should not generate emails outside of the typical English character only emails. By doing so, you are helping those that do possess such an email and prepare yourself for a strong future when it becomes more commonplace as well.

TurtleKwitty
  • 301
  • 1
  • 8
  • 1
    What you describe in the last paragraph is the late Jon Postel's _robustness principle_ described in section 3.2 of [RFC 760](https://tools.ietf.org/html/rfc760): "In general, an implementation should be conservative in its sending behavior, and liberal in its receiving behavior." – Blrfl Dec 01 '18 at 12:55
1

My experience is that you would have problems with non ASCII e-mail addresses all the time, so you kind of never see them in real applications.

So my advice is: It's save to treat international UTF-8 e-mail addresses as dead technology.

You might also be interested in this question:
- can-an-email-address-contain-international-non-english-characters
- are-email-addresses-allowed-to-contain-non-alphanumeric-characters
- wikipedia: International_email

MrSmith42
  • 1,041
  • 7
  • 12