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

[John C Klensin]

--On Saturday, April 11, 2009 08:57 -0700 Mark Davis
<mark at macchiato.com> wrote:

>...
> I think what we can say is programs SHOULD convert to the
> canonical U-Label form before transmitting to other IDNA-aware
> programs, and to the A-Label form before transmitting to
> non-IDNA-aware programs. (The above isn't formal language, but
> you can see what I mean.)

Can you identify the use cases for exceptions to this rule
(i.e., what justifies a SHOULD, rather than a MUST) rather than
making the non-conversion behavior non-conforming?

   john