[ietf-nntp] Draft 21

[Russ Allbery]

Clive D W Feather <clive at demon.net> writes:

> The idea was that HDR ALL means "you don't need to bother with LIST
> HEADERS if you only want headers, not metadata"; both it and the latter
> command are optimisations when dealing with headers.

> I would say that:
> * EITHER we leave it as is (simple),
> * OR we be completely symmetrical, and modify LIST HEADERS to be:
>     LIST HEADERS MSGID        list headers available with message-id
>     LIST HEADERS ARTNO        list headers available with number/no argument
>     LIST HEADERS              list headers available in both ways
> and make the LIST EXTENSIONS results be:
>     HDR ALL                   LIST HEADERS returns ":"
>     HDR ALLMSGID              LIST HEADERS MSGID returns ":", ARTNO doesn't
>     HDR ALLARTNO              LIST HEADERS ARTNO returns ":", MSGID doesn't
>     HDR                       No form of LIST HEADERS returns ":"

I think the way that you have it in your proposed draft is fine.  It's a
little bit annoying, but I think the above is too complicated.  What you
have now seems like an excellent compromise approach to me.

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