[Fwd: Re: [Fwd: RE: Consensus Call on Latin Sharp S and Greek Final Sigma]]

Eric Brunner-Williams ebw at abenaki.wabanaki.net
Wed Dec 2 01:31:19 CET 2009

A co-worker, Swiss-German writes:

-------- Original Message --------
Subject: 	Re: [Fwd: RE: Consensus Call on Latin Sharp S and Greek Final
Date: 	Wed, 2 Dec 2009 01:28:22 +0100
From: 	Werner Staub (CORE) <werner.staub at corenic.org>
To: 	Eric Brunner-Williams <ebw at abenaki.wabanaki.net>
CC: 	Werner Staub <werner.staub at corenic.org>
References: 	<4B146353.5090600 at abenaki.wabanaki.net>

Hi Eric,

My view would be to treat the final Sigma and the Esszet differently.
Here are the reasons:

1) The final sigma is extremely frequent, wheras the Esszet is rare.

2) Putting a normal lower-case Sigma instead of a final Sigma is
visually unacceptable.  By contrast, putting "ss" instead of "ß" is
acceptable and avoids confusions. Especially for the Swiss, of course.

3) Greek has an even bigger variant problem because of the Tonos. A
majority of Greek IDN names would need a variant without tonos, both for
uppercase typing (where the Tonos is never used) or comfort of typing
when writing lowercase. As a result, contrary to German, Greek requires
strong registry rules that establish the equivalence of variants. The
German registries in contrast have decided against registry-based
variant equivalence rules. Also note that the Greek variant equivalence
rules in question are one-to-one mappings between IDN characters, where
as the German equivalent rule for Esszet would be between one IDN
character and and two pure Ascii characters.This is another reason for
reluctance in German to do this as a registry policy and the preference
to have it handled in the protocol.

4) If URL rewriting is to work correctly in all cases, the final sigma
must be transmitted to the server. In German, on the other hand, if
confusion is to be avoided, the Esszet has to be converted to "ss".

These are good reasons for me to advocate giving the German-speaking
registries what they prefer, i.e. no change for the Esszet compared to
IDNA2003, and to give the Greeks and Cypriots what they prefer, i.e. the
ability to transmit the final sigma.



More information about the Idna-update mailing list