Bundling of Domain Names and DNAME

[Vint Cerf]

Vaggelis,

extensions of DNS in the XNAME direction would require creation of a  
new WG for this purpose.
I menion this only to make sure no one is expecting the IDNABIS  
working group to undertake the idea.

Vint


On Dec 4, 2009, at 8:28 AM, Vaggelis Segredakis wrote:

> Dear Erik,
>
> Actually it does not matter whether the final sigma is PVALID or not,
> although I believe that it will be since both the .gr Registry and .cy
> Registry have asked for it.
>
> We have the same issue with the accent mark "tonos". Almost each  
> word in
> Greek has this accent mark in small letters but if you put it in  
> capital
> letters the mark is omitted, creating two different xn-- Punycode
> translations for each word used as a domain name.
>
> We Bundle domain names already because of this "tonos" and we face  
> this
> bundling issue since 2005 but we had put our hopes on the IDNA  
> revision for
> a better solution. Since this is not possible, we need the XNAME  
> bundling
> regardless of the final sigma in IDNA2008 situation.
>
> Kind Regards,
>
> Vaggelis Segredakis
>
> -----Original Message-----
> From: Erik van der Poel [mailto:erikv at google.com]
> Sent: Thursday, December 03, 2009 7:59 PM
> To: Vaggelis Segredakis
> Cc: namedroppers at ops.ietf.org; Andrew Sullivan; Olafur Gudmundsson;  
> Vint
> Cerf; Sotiris Panaretou; idna-update at alvestrand.no
> Subject: Re: Bundling of Domain Names and DNAME
>
> Vaggelis, thank you for raising this issue on the namedroppers list.
>
> Namedroppers, I would like to clarify one item below.
>
> On Thu, Dec 3, 2009 at 7:56 AM, Vaggelis Segredakis <segred at ics.forth.gr 
> >
> wrote:
>> Recently, as a member of the discussion of the IDNABIS WG I tried  
>> to help
>> resolving these issues by explaining them to that group. However that
> group
>> is not free to completely re-design the IDNA protocol to something  
>> else
> but
>> rather with small steps to reform it to something with fewer issues.
>> Unfortunately, on this process, one of the changes that are  
>> implemented
> lead
>> to even more names that have to be bundled together for each  
>> registrant.
>> This makes it even more significant to break this chain of cost for  
>> the
> end
>> user.
>
> The IDNAbis WG drafts have not been published as RFCs yet. Currently,
> the drafts make Final Sigma (Unicode U+03C2) PVALID, but this issue is
> currently being discussed in the WG.
>
> Erik
>
>