[NNTP] LISTGROUP

Russ Allbery rra at stanford.edu
Mon Apr 25 10:14:02 PDT 2005


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

> I'm happy with the concept. I'd rather we hammered down the semantics
> and then let me draft text rather than having to edit someone else's
> text so as to preserve consistency of style. To that end, I *think*
> we've settled on:

> * Range is an optional second argument, only available if a group has
>   been specified. It has the same form as with OVER.
> * The *only* effect is to remove some entries from the list following a
>   211 code; those entries that don't occur within the range.
> * Consensus is to leave the current article number as if no range had been
>   specified.

I agree with the above as a working group participant; as chair, I do want
to point out that there have been a couple of people who wanted to see the
current article number set to the beginning of the range.  I don't know if
my argument persuaded them?

> * Still undecided what to do if:
>   - the range is valid but empty (e.g. 1-5 when the LWM is 10);
>   - the range is inverted (e.g. 5-1).

I don't think we really were undecided about the first -- LISTGROUP
returns an empty multiline response and otherwise behaves just like GROUP.
I think we should do the same with the latter, since returning an error
would mean making a decision about whether the group was selected anyway,
and I think it's best to make it behave the same as the first.

Both of these are different than OVER and HDR, though, which is worth a
bit of nervousness.

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



More information about the ietf-nntp mailing list