Changing the xn-- prefix

Andrew Sullivan ajs at
Wed Mar 19 16:36:59 CET 2008

Hi Simon,

On Tue, Mar 18, 2008 at 10:11:56PM +0100, Simon Josefsson wrote:
> Given ß, I'm not sure German is a simple case.

It is under IDNA2003: it's mapped.  And this problem is why I picked
it as the "simple" case, because it turned out not to be simple at
all, even though everyone claimed it was a totally simple one.  (When
Afilias implemented German registrations, I had a German colleague
rant at me for four hours one day on why this was wrong.  I understood
everything he was saying, but there was nothing I could do about the
mapping already prescribed.  That very sort of discussion is, I
supposel, how we got here.)

> Hm, you have a point.  If this approach is to be seriously considered
> one would probably have to describe exactly how to map IDNABIS names to
> IDNA2003 names, for all the situations where IDNABIS isn't backwards
> compatible.  There are certainly more things to consider as well.

Well, that's part of it.  As soon as there is an official backward
compatibility, then registrants will (quite plausibly) demand that
they get the IDNA2003 mapping whenever they register the IDNABIS
version.  It would effectively mean that there will be new IDNA2003
registrations _after IDNABIS is implemented_, which is surely the
wrong direction.  (I know that someone is going to reply to me that
registries just have to say, "No."  Registries wouldn't have a good
basis for saying that, however, if there were a standard mapping from
the old standard to the new one.)


Andrew Sullivan
ajs at 
+1 503 667 4564 x104

More information about the Idna-update mailing list