Re: [idn] Adding "optional" characters in draft-ietf-idn-nameprep

[Brian W. Spolarich]

On Tue, 15 Aug 2000, James Seng wrote:

| Keith already said this during the meeting. Wires and bits don't care
| about I18N. If we only focus ONLY on the wire format, then the DNS
| packet are 8-bit clean already. So lets shutdown the list, go home and
| pat yourself on your head.

  I think you're taking my point past where I intended it to go.  What I'm
suggesting is that we should identify any gaps in the standards and
technical reports produced by the UC and work with them on producing
the necessary extensions or new TRs to cover the missing cases.  

| The reason I18N of Domain Names is because of Human Interaction. And
| that is more important than the wire format. 

  I agree completely.  What I'm suggesting is that a proposed IDN standard
in which the IETF defines the data format, protocol interactions, etc. and
which follows all of the Unicode Consortium TRs in terms of character
sets, c15n, etc. would be more easily accepted as the 'right' solution by
the humans who are going to have to live with whatever is implemented.

  -bws