Eszett and IDNAv2 vs IDNA2008

[John C Klensin]

--On Sunday, March 15, 2009 8:45 PM +0100 Cary Karp 
<ck at nic.museum> wrote:

> Quoting Erik:
>
>> If HTML implementations start performing multiple lookups, we
>> will probably start to see Web sites that take advantage of
>> the new xn-- names, and they may not bother with any
>> bundling, since the browser is performing multiple lookups.
>
> I don't understand how the HTML implementation determines what
> happens when one name in a registered bundle is called.
> Doesn't that depend entirely on how the holder of the bundle
> configures the HTTP server to which the names are directed?
>
> In fact, one of the registry headaches that tends to dampen
> enthusiasm for bundling, is the need for name holders to
> understand that they have to direct, by serverside
> configuration, all of the non-canonical forms of a name to the
> single preferred form -- assuming that there is one (c.f.
> John's remarks).

And, while you know this, just to remind anyone who has 
temporarily forgotten...  For HTML/HTTP, that is the relevant 
server for those protocols, not the DNS server.  Since there are 
no service location or MX records for those protocols, there 
isn't a lot the DNS can do unless the variant names are inserted 
by DNS aliases (CNAME or DNAME) and, while it may be hidden from 
the client, even those tend to involve addition lookups and/or 
DNS server-side work.

     john