Stability of valid IDN labels

[Eric Brunner-Williams]

John C Klensin wrote:
> Mark,
>
> For better or worse, Eric is probably right.   
I think I'll frame this.

However, the transformation from the universe of all two or fewer octet 
sequences to the subset of ISO 3166-1-alpha-2 code values contained at 
any point in time in the IANA root zone, whether an ugly table driven 
kludge executed by people using ed(1), or a marvelous shiny syntax 
validator, was an obviously poor attempt to ask about the temporal 
properties of "valid" and "invalid" (and anything neither valid nor 
invalid). Unfortunately I had to destroy the Soviet Union to no useful end.

This is my real interest, which is trying to get an understanding of 
where "forever" and its cognates comes from, as it appears at first 
glance to be significantly different from contracts of a commercially 
reasonable duration, and may therefore constrain mechanisms that are 
capable of being realized over "commercially reasonable durations".

Eric