ietf-nntp HDR parameter proposal

[Russ Allbery]

Charles Lindsey <chl at clw.cs.man.ac.uk> writes:

> I can accept your opinion that you do not like my proposal, but I object
> most strongly when you reject it on the grounds that it contains feature
> X, when in fact it does not contain feature X at all.

Quite understandably, too.  :)

> Read what I proposed. The only things that can be returned by LIST
> EXTENSIONS are
>     HDR
>     HDR ALL
>     HDR OVERVIEW
> where is the "potentially very long list of headers in LIST EXTENSIONS" in
> that?

I'm sorry, I must have had it confused with a previous version of your
proposal or some other discussion that we had before.  Mea culpa.

I have considerably fewer problems with the above approach from a protocol
perspective, but I'll warn that it does not capture one of the most common
cases.  Supernews, for example, offers HDR on various headers not in
overview but doesn't offer it on all headers (last I heard).  I expect
this to be increasingly common.

And LIST OVERVIEW.FMT still has the other problems discussed previously,
and they apply to this use of it as well as the OVER use of it.

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