ietf-nntp HDR

Ade Lovett ade at lovett.com
Tue Jan 8 13:56:43 PST 2002


On 01/08/02 14:09, "Clive D.W. Feather" <clive at demon.net> wrote:

> Ade Lovett said:
>>> Making one command require another one to be used first is a bad idea
>>> unless it can be justified in contextual terms. Requiring GROUP before NEXT
>>> makes sense. Requring LIST OVERVIEW.FMT before HDR doesn't. What are you
>>> going to do if the client doesn't bother ? Reject it for being awkward ?
>> Then change it to a SHOULD (and be prepared to deal with the consequences if
>> it doesn't,
> 
> *what* consequences ?

The blindingly obvious one which you yourself state one paragraph down,
namely that the client may start seeing 5xx's.

It's also an optimization thing.  A client may not want the entire overview
entry, but a couple of specific headers (perhaps, shock horror, those not
mentioned in the NOV spec).  It does a LIST OVERVIEW.FMT, sees the list of
headers and goes "wow, I can grab this header instead of the entire overview
or, worse, using HEAD).

I had expected you of all people to understand the concepts of providing not
only a complete protocol, but a scalable one.  Seems I was wrong.

-aDe




More information about the ietf-nntp mailing list