Lookup for reserved LDH labels

[John C Klensin]

--On Wednesday, 07 November, 2012 11:04 +0100 Marcos Sanz
<sanz at denic.de> wrote:

>...
>> The term "putative U-label" was used precisely to cover the
>> cast of things that could not possibly be "U-labels".  That
>> includes not only [valid] U-labels (if it isn't valid, it
>> isn't a U-label), but XN-- labels that aren't U-labels _and_
>> anything else with "--" in the third and forth positions.
>> The intent --which I think actually goes back to IDNA2003
>> although I don't have time to check right now-- is to reserve
>> the entire family
> 
> IDNA2003's ToASCII("ad--acta.de") did not fail. It was an
> aware decission  built-in in the design of the algorithm; cf
> RFC 3490 Section 4.1, 1st and  4th algorithm steps:

Sorry for not being clear enough.  My memory is that IDNA2003
reserved all "CC--" ("C" as arbitrary ASCII character) strings.
However, a major difference between IDNA2003 and IDNA2008 was
that the latter basically permitted looking up anything.  It
also did not require a dual relationship between
native-character labels and their Punycode-encoded forms.
Because knowing whether something was valid or not requires
knowing the properties of characters to maintain the inclusion
principle (and because "look up everything" proved inadequate to
'buy' forward compatibility as some people with hoped IDNA2003's
strategy IDNA2008 would do, IDNA2008 changed that decision to
permit looking up only valid labels.

For the record, that summary of the IDNA2003 situation and the
reasons for the changes are my personal opinion.  I don't know
if a survey of the WG participants (or the IETF more broadly)
would show consensus for that interpretation.   But the
conclusion that IDNA2008 does not permit a conformant
implementation to look up something is knows is not a [valid]
A-label is very clear.

best,
   john