Consensus Call on Latin Sharp S and Greek Final Sigma

Erik van der Poel erikv at google.com
Wed Dec 2 04:00:18 CET 2009


Should Latin Small Letter Sharp S and Greek Small Letter Final Sigma
be PVALID in IDNA2008 or not:

My position:

(2) Both characters should be DISALLOWED

It is too painful to change the mappings of these characters at this
point. While it may be frustrating for registrants if they cannot
register names that only differ in these characters (and ZWJ/ZWNJ), we
have similar restrictions in other languages, such as the English
"polish" vs "Polish".

A display preference can be developed, possibly in DNS, definitely in
HTTP, starting with the infamous 4 characters mentioned above and
possibly the other problematic characters (characters that only had
upper-case versions in Unicode 3.2 and Han characters that had their
normalizations changed after Unicode 3.2).

Later, the display preference can even be extended to AAA.com and
Ecole.fr, as long as the security implications are understood.
(Upper-case has more confusables than lower-case.)

Another advantage of mapping away ZWJ and ZWNJ is that we don't need
to implement contextual rules for them.

The TRANSITIONAL idea strikes me as a double transition -- first we
wait for users to stop using the current browsers, then we wait for
them to stop using the newer browsers that made the problematic
characters illegal. That is rather a long wait, when the display
preference might be implemented more quickly.

Erik


More information about the Idna-update mailing list