Re: New Internet Draft on registering IDNs

[Paul Hoffman / IMC <phoffman@xxxxxxx>]

At 3:34 PM +0100 3/27/03, Stephane Bortzmeyer wrote:
> Well, to summary, I find it quite good, much simpler and more
> understandable than draft-jseng-idn-admin-02, specially for non-CJK
> countries.

Thanks!

>  > A "string" is an ordered set of one or more characters.
> >  This document discusses characters that have equivalent or
> >  near-equivalent characters or strings. The "base character" is the
>
> Shouldn't we use "code point" instead of "character"?

Character is more understandable, and I don't see any place in the 
document where a character wouldn't be considered its code point, but 
I'm open to changing it.

>  > If the base character has more than one variant, the variants
> >  are separated by a colon (":", ASCII 0x3A). Strings are given without
> >  any intervening spaces
> ...
> >  U+2202|U+0064;U+03B4
>                ^
>                Isn't it a typo?

Yes.

>  > A registry has three options for how to handle the case where
> >  the registration bundle has more than one label. The policy options are:
> >
> >  1) Allocate all labels to the same registrant, making
> >  the zone information identical to that of the input label.
> >
> >  2) Block all labels so they cannot be registered in the
> >  future.
> >
> >  3) Allocate some labels and block some other labels.
>
> This entire scheme does not discuss financial issues.

Correct. It would be unwise to predict how registries will act on 
this, especially over time.

>  For instance, in
> Option 1, it will mean that a registrant will get more labels than he
> paid for. The registry will not be happy :-)

That's not necessarily true. Registries can decide whether or not to 
charge more for names with bundles. Note that in option 2 and 3, the 
registry is prevented from making money, so they might charge for 
bundles regardless of which option they choose.

> In Option 2, OTOH, it means that there is no option for the registrant
> to activate some variants.

Right.

>  Do you think this case (all variants are
> blocked and some are allocated to the same registrant, if he chooses
> so and if he pays, may be a smaller price than a "real" domain) is
> covered by Option 3?

That wasn't my intention, but it is an interesting thought. My 
intention for option 3, which is what is discussed in the JET 
document, is that the registry decides using the table which names 
are allocated and which are blocked.

> If the variants actually allocated are choosen by the registrant, it
> is up to her to minimize confusion.

You are proposing something different: the registry allows the 
registrant to pick which names from the bundle go in either category. 
It will still have the same problem of option 3, namely typical users 
of the DNS won't be able to predict what they should type, but it 
sounds "nicer" than option 3.

What do other folks on this list think of that idea?

--Paul Hoffman, Director
--Internet Mail Consortium