ietf-nntp Alternative HDR parameter proposal

[Russ Allbery]

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

>> perhaps we ditch the ALL capability keyword.

> I still think it's a useful thing to have - if you don't want metadata,
> it saves the need for a LIST HEADERS.

Agreed.

>> We can either use a different LIST HEADERS response code in the ALL
>> case (as you suggest above), or a special keyword in the 215 response,
>> eg:
>> 
>> C> LIST HEADERS
>> S> 215 headers follow:
>> S> *                   <= means all headers
>> S> :lines
>> S> :bytes
>> S> .

> I'm not particularly bothered - the two are functionally equivalent.
> Someone needs to make a decision, though.

I like using * better than a different response code.  It's more obvious
to a human what's going on even if they're not staring at a standard.

> It does make me think, though - in the non-ALL case, do we want to
> require all headers before all metadata items? If we take your approach
> to the ALL case I think it would be useful to require the * to be either
> first or last.

Hm, I'm not sure what circumstances this would really be useful.  If I
were writing a client, I'd write it the same either way, since I still
have to handle somehow the case where the server screws up, and if I'm
going to diagnose that it's actually easier to just support * anywhere.

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