Historic Scripts - New version: draft-ietf-idna-tables-01.txt

Shawn Steele Shawn.Steele at microsoft.com
Wed May 7 18:16:22 CEST 2008

> Can we also hear from others on this please?

I think that there is no need to restrict historic scripts (or other groups in general).  If there is a small study group that thinks it would be "fun" to register a domain name in a script they are enthusiastic about, I have no problem with that, and I don't see any need to discriminate against such potential users of a script.

>From a security view I could see how someone may try to take advantage of an esoteric script, but frankly there are just as many, if not more, "legal" combinations that are interesting from an attackers perspective, so the increase risk is very small.  (Heck I get legitimate mail from vendors I do business with asking to go to "mail.surveys-r-us.com" or something equally silly.  The last step of some storefronts takes one to a 3rd party payment system that has nothing in the name in common, etc.)

So my "vote" is to allow as many scripts/characters as possible.  I think the focus needs to be on unambiguous representation and display of bidi text and making sure that characters are represented in a single unique way (eg, the NFKC step).  I would restrict the control and formatting characters that aren't necessary to represent a specific language/script, but otherwise I'd be fairly liberal.

- Shawn

More information about the Idna-update mailing list