ietf-nntp draft-ietf-nntpext-base-17

[Russ Allbery]

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

> However, an individual server might decide that certain groups (e.g.
> those in Usenet-2) should not be listed to clients with insufficient
> authorization. I think such a server should just be left to get on with
> that, with no help or hindrance from our document.

Right, this is common practice.

> My newsgroups file is actually bigger than my active file (CNews updates
> it automatically from a checgroups message, before asking whether you
> want to add the missing groups - I actually find that convenient for
> knowing what groups are out there).

> But I would certainly expect the newsgroups file to be a superset of the
> active file, even if there were no descriptions for some groups in it.

Then you would expect incorrectly, if you were using normal, currently
deployed news servers.

I think the question is whether we should outlaw existing (sloppy)
practice or not.  My inclination is to say no.  I don't think there's any
question about what the existing practice is like, namely that LIST ACTIVE
is the canonical list of groups, LIST ACTIVE.TIMES is a subset of that
list, and LIST NEWSGROUPS is some other collection of groups that may be a
superset, a subset, or just a disjoint set from LIST ACTIVE (but gets
closer and closer to a matching set on servers that pay a lot of attention
to it).

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