The EAI Working Group (http://datatracker.ietf.org/wg/eai/charter/) is making rapid progress toward standardizing Unicode email addresses. Unicode email addresses are a terrific feature for people in many countries that don’t use Latin/ASCII as a native script. Ironically, in the US its easy to miss the importance of non-ASCII email addresses. Many other Latin script users may also think EAI is neat, but not critical since they make do with Latin today. Cyrillic or Chinese users on the other hand may have a difficult time using Latin/ASCII.
I expect that people in other countries would like to have a native script email address. In addition they’ll probably require an ASCII address, especially in the near term. Even long term though, you’ll likely see a big demand for multiple addresses. Today we have an ASCII solution for “aliases”. Shawn.Steele and shawnste both seem to work the same to me, but contacts and mail clients may see them as different users.
For example, today a Japanese business card often is double sided, with a Latin script and a Japanese script side. It’s easy to imagine that user wanting a Japanese email address for one side of their business card, and a friendly ASCII address on the other side. I’d expect that when choosing an ASCII script alternate address, most users will want something similar to today: Either a transliteration, nickname, or other easily communicated address. xn-- worked for DNS because you can always have CNAME, DNAME or A records for an ASCII address as well, but it’s not going to fly for email addresses.
In an EAI world, it’s pretty easy to see how a user would want a native Cyrillic address to use, and also a Latin address to use when talking to their importer/exporter, or to do other business outside of the country. Unfortunately US businesses will have little incentive to update to EAI support, so users may need Latin addresses for simple software support or other issues for some time.
What’s less obvious are the more complicated cases. A Mongolian user may prefer a Mongolian address to share with their customers in their community. To communicate with the government apply to a university they may need a Chinese address. To reach friends or business partners in the US, they may also need an ASCII address. So one user may have aliases in three (or more) scripts pointing to the same mailbox!
Some users may actually be fluent in two non-Latin scripts and have multiple addresses, neither of which are ASCII. It is even possible that those users don’t need to communicate outside of that environment and may never use an ASCII address at all.
In conclusion, when thinking about international email addresses:
- International users will probably have multiple addresses for the same mailbox. Contacts and mail client applications probably need to recognize that.
- There might even be more than two addresses, with ASCII as a last choice.
- In some environments users may not need an ASCII address, so apps can’t rely on one always being present.
- Users can probably only type scripts they’re familiar with. (Writing your chinese address on a business card wouldn’t help me at all)
- Even when other addresses are assigned, they may not be available. (The App could have an address cut & pasted from a document or forwarded or something else.)
Hope someone finds these thoughts helpful,