ietf-nntp HDR replacement text

[Russ Allbery]

Andrew Gierth <andrew at erlenstar.demon.co.uk> writes:
>>>>>> "Russ" == Russ Allbery <rra at stanford.edu> writes:

>  Russ> I'm not sure I understand: You're saying that the current XHDR
>  Russ> behavior is broken in this regard?

>  Russ> The above is essentially what XHDR does now.  XHDR doesn't
>  Russ> distinguish between nonexistent articles and present articles
>  Russ> that don't have that header.

> This does not agree with my observations or experiments. What server
> behaves in that way?

> Those servers I've tested produce no response line for articles that
> are missing, and an empty response (or the deprecated "(none)"
> response) where the article is present but the header is missing.

Ack.  You're right.  Mea culpa; my testing methodology was flawed.

Sounds like we should be returning an empty response for both a missing
header or a header with no content, then, just like OVER will.  I'll make
that change in my writeup.

-- 
Russ Allbery (rra at stanford.edu)             <http://www.eyrie.org/~eagle/>