[NNTP] LIST EXTENSIONS (again)

Russ Allbery rra at stanford.edu
Sun Nov 7 09:19:59 PST 2004


Howard Swinehart <binaryboy at binaryboy.com> writes:
> From: "Mark Crispin" <MRC at CAC.Washington.EDU>

>> The server must advertise both NNTPv2 and NNTPv3.  A server which just
>> advertises NNTPv3 is not OK to an NNTPv2 client.

> This returns us to the original problem of identifying broken servers.

> NNTPV2 was supposed to distinguish between Netscape Collabra (has LIST
> EXTENSIONS but wrongly advertises OVER) and a truly compliant
> server. Without NNTPV2 and without a version number, a client based on
> this proposal won't see a difference between a future incompatible
> NNTPVx server and Netscape Collabra because neither will return NNTPV2.

I think that's actually the correct behavior.  That future server is
speaking a protocol that the client doesn't understand, which is, if one
thinks about it in the right direction, exactly the same as the case for
Netscape Collabra.  The client can give up, or fall back on RFC 977 and
try using commands at random and hope, which is pretty much the case for
Collabra today.

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



More information about the ietf-nntp mailing list