ietf-nntp POST and IHAVE responses (was: Commetns on draft-15.pdf)

[Andrew Gierth]

>>>>> "Russ" == Russ Allbery <rra at stanford.edu> writes:

 Russ> With IHAVE, generally the other server doesn't particularly
 Russ> care whether or not the article was posted, but with POST, it's
 Russ> actually important to let the client know that the article is
 Russ> invalid.  The exception is generally only for dealing with
 Russ> spam, where the intention is to make the server be deceptive to
 Russ> try to fool a spammer.

However, the broken behaviour of some clients (certain versions of OE,
apparently) when given a 441 response does sometimes require extending
this to non-spam articles, e.g. by returning an error the first time a
given article is rejected for a policy reason, but if the same article
is offered again, accepting and discarding it. This is not really a
very nice solution, but a large OE-using user base can otherwise
generate a quite phenomenal number of posting attempts.

 Russ> the following new text:

 Russ>     A response of 240 SHOULD indicate that barring unforseen
 Russ>     server errors the posted article will be made available on
 Russ>     the server and/or transferred to other servers as
 Russ>     appropriate.  In other words, articles not wanted by the
 Russ>     server SHOULD be rejected with a 411 response and not
 Russ>     accepted and silently discarded.

s/411/441/

 Russ> Intentionally deceptive spam filters can then be left as a
 Russ> conscious decision to not abide by SHOULD, which is explicitly
 Russ> allowed in the definition of SHOULD.

yup

-- 
Andrew.