Two-step mapping

[John C Klensin]

Hi.

As I've noted several times, I am not going to start rewriting
text that involves mapping plans until either we converge on
what we want to do or Vint tells me to, whichever comes first.
As has been discussed on the list, it may be sensible to more
all or part of the mapping discussions in Rationale and/or
Protocol into the Mapping document, which is another reason I
haven't been rewriting.

You are looking at text that was constructed well before the San
Francisco meeting and its conclusion that we needed to address
mapping in more specificity than the long-term "do what you
think you need to do on lookup" provisions of Protocol.   The
two-step mapping of that text involves the notion of "optionally
do some mapping, look things up to see if you find something,
then map (some more) and look up again".  I posted an analysis
some months ago as to why it wouldn't work (could easily yield
ambiguous and conflicting results, even long-term).  That
analysis was independent of what seemed to be general distaste
for anything that might involve multiple lookup (if there isn't
multiple lookup, any two-stage process can be reduced in
implementation to a one-stage process.   I had, until now,
assumed the idea was dead.   

Are we reopening the issue or just to clarify text that I
consider dead?

    john

p.s.: _Please_ identify the document you are talking about,
rather than saying "IDNA2008".  Because Rationale provides
explanations, there are inherently overlaps between it and the
other documents and I often cannot guess from reading one of
these notes which document someone is referring to.  Having to
play detective wastes a lot of time and may produce irrelevant
answers or discussions.

--On Sunday, July 26, 2009 04:53 -0400 Andrew Sullivan
<ajs at shinkuro.com> wrote:

> On Sun, Jul 26, 2009 at 03:21:40PM +1000, Wil Tan wrote:
> 
>> user given specific user requirements, and *avoids the
>> two-step mapping* of the original protocol.
>...
> John would be the authoritative source of what the intent
> behind that text is, but my reading of the "two steps" in
> IDNA2003 is more like this:
>...