[NNTP] [2501] Reader commands in transit servers and vice-versa

Russ Allbery rra at stanford.edu
Fri Dec 3 10:51:48 PST 2004


Ken Murchison <ken at oceana.com> writes:

> I really don't see a problem with READER and TRANSIT, since the are just
> protocol tokens which only the client software (and developer) ever
> sees.

> I suppose I could live with something like USER" and TRANSIT or USER and
> PEER or USER_MODE and PEER_MODE, but I really don't see the
> point. Everyone already knows that POST is a reader (user-agent) command
> and READER gives us symmetry with MODE READER.

These are also the terms already used by NNTP software implementors.

> This seems like a bridge we only need to cross if we ever come to it. In
> order for a new transit command to be added to the core spec, the new
> document, with presumably a new VERSION 3 token, could handle this.

Agreed; any other transit commands short of a base protocol revision would
be extensions and would therefore come with their own capability labels.
So this is a problem that we can safely defer.

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



More information about the ietf-nntp mailing list