Comments on idnabis-rationale-01

[Frank Ellermann]

Tina Dam wrote:

> I would be ok with a different term as well, but I don't
> have any good ideas.

The funny thing is that we all like LDH-label as intuitively
clear, and then don't agree on the same definition.  Here's
some simplified ASCII art:

+--------------------------------+
| DNS labels (octets)            |
|                                |
| +------------------------------+
| | LDH labels (LDH)             |         
| |                              |
| | +----------------------------+     +----------+
| | | A-labels (IDNAbis valid)   | <=> | U-labels |        
| | +----------------------------+     +----------+
| |                              | 
| +------------------------------+
|                                | 
+--------------------------------+

What I see is "A-labels are a proper subset of LDH-labels,
as specified in IDNAbis.  LDH-labels are a proper subset
of DNS labels, as specified in RFC 1123" (or similar).

What John sees is "A-labels are one thing, and LDH-labels
are the DNS labels consisting of LDH which are no A-labels,
as specified in IDNAbis".

IOW John has no name for the union of A-label and LDH-label
in his terminology (the middle box in the ASCII art).

I've no term for those LDH-labels which are no A-labels in
my terminology (the middle box excluding the innermost box).

 Frank