mappings-01

Patrik Fältström patrik at frobbit.se
Sun Jul 5 22:20:02 CEST 2009


On 4 jul 2009, at 17.10, Paul Hoffman wrote:

> It is clearly a good idea because it will prevent user surprise

I am prepared on giving a counter argument.

With mapping, people that use the unicode version of the string do not  
know whether they talk about a string that is mapped to something, or  
not. Because of this, they are surprised if they talk about the string  
that is mapped, and then when the xn-- version is converted back to  
Unicode, they say "this was not what I wanted".

It must be extremely clear that the string that is mapped to a U-label  
is an exceptional case that some applications in some cases, after  
clearly instructing the user the string is mapped, might be accepted  
as input as an IDN.

IDNA2008 must in a much much much better way than IDNA2003 explain why  
it is so extremely important that the strings that can be converted to  
an A-label and back (without changes) is what is an IDN.

Because of this need for an 1:1 mapping, mapping is an exception that  
can help some applications in some cases to make them more friendly  
for users.

    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 : http://www.alvestrand.no/pipermail/idna-update/attachments/20090705/ad6654fe/attachment.pgp 


More information about the Idna-update mailing list