Mapping Stability/Storage (was Re: M-Label or MVALID, and dangers with mappings?)

Patrik Fältström patrik at
Sun Apr 12 03:47:11 CEST 2009

On 12 apr 2009, at 03.38, Mark Davis wrote:

> So you are saying that Gmail (my emailer) would non-conformant
> under IDNA2008 to have sent   "http://Ö <>"  
> to you,
> and your emailer was also non-conformant by sending
> "http://Ö<>
> "  back?

First of all, you confuse things by talking not about domain names,  
but URIs, and not only that, "free text" in an email text body. I.e.  
non-typed text.

You can send in a text email whatever you want, but how  
_the_domain_name_ is interpreted depends on when you actually start to  
interpret that free text as a domain name.

I also see it as confusing that you have two text parts there. One  
that is a string that contains something that is not a U-label, and  
something that is an A-label.

If you talk about domain names, then you should talk about domain  
names. Not "text".

> What value would it be to do that? Since we have agreed that the  
> mapping is
> permanent, what gain would that have?

I do not see such an agreement. Specifically as we have not agreed on  
what "mapping" involves.

The gain is to get people to stop using things that is not 1:1 between  
A-label and U-label. So that we do not get anything else than what is  
a domain name stored and saved for the future.

Nothing stops you from sending whatever text you want to me, but  
nothing stops me / my software to interpret what you sent me in  
whatever way I want, according to the context where I happen to be at  
the moment (at the time of interpretation).

I _definitely_ see text that is not clearly a A-Label or U-Label be  
target for all different interpretation that "helps" the user. Just  
like URIs are interpreted in all different kind of ways. Specifically  
when the URI is "hidden" in free text.


> I just don't understand.
> Mark
> 2009/4/11 Patrik Fältström <patrik at>
>> On 12 apr 2009, at 03.13, Mark Davis wrote:
>> (2) I include http://Ö <> in this email  
>> message.
>>> My emailer is IDNA aware,
>>> and recognizes this as a URL. I send the email to yours, which is, I
>>> presume, also IDNA aware. I don't want either one to lowercase it.
>> I think you show a case where you can not know before sending the  
>> email (as
>> the domain name used in the URI is not a U-label) where the body of  
>> the text
>> will end up, so I would say you MUST convert it to a U-label before  
>> sending
>> it.
>> This is exactly the confusion that exist in IDNA2003 that I see we  
>> must get
>> away from. We must get people to use the U-label or A-label formats  
>> when
>> passing things around (when not knowing DEFINITELY how the  
>> receiving side
>> will treat the data).
>>  Patrik

-------------- next part --------------
A non-text attachment was scrubbed...
Name: PGP.sig
Type: application/pgp-signature
Size: 186 bytes
Desc: This is a digitally signed message part
Url : 

More information about the Idna-update mailing list