The slide I presented today

[Patrik Fältström]

On 18 nov 2008, at 11.14, Mark Davis wrote:

> Didn't see any mention of this on the email list before now.
> If we really wanted a constraint, it should be generalized: don't  
> allow two
> different numbering systems. This is slightly tricky to state but  
> could be
> done with:
>
> If two characters in a label are of type Nd (decimal numbers), then  
> the
> numeric difference between the code points cannot exceed 10.

>
> This works because the Nd's are allocated in groups of 10, and are  
> more than
> 10 values from one another. (Script differences are not enough.)
>
> However, this seems to me like a well-meant, but overly specific,  
> condition.

This was the conclusion when this was discussed.

> If we were to go this route we could spend endless time coming up with
> similar levels of constraints. Each would be perhaps reasonable, but  
> the
> specifications get more and more complicated. My suggestion would be  
> to
> leave this for the registries.

That is not enough for the indic numbers. Pete Resnick and John  
Klensin is working on text that explain the rationale for it.

You can see more of the discussion when the minutes is released.

    Patrik

> Mark
>
>
> On Tue, Nov 18, 2008 at 07:35, Patrik Fältström <patrik at frobbit.se>  
> wrote:
>
>> Here is the slide you saw at the meeting.
>>
>>  Patrik
>>
>>
>>
>>
>> _______________________________________________
>> Idna-update mailing list
>> Idna-update at alvestrand.no
>> http://www.alvestrand.no/mailman/listinfo/idna-update
>>
>>