Re: ietf-nntp Initial draft FINALLY available

[Ben Polk <[e-mail]>]

At 04:04 PM 10/3/96 +0000, chris (c.) lewis wrote:

>From another perspective, I don't believe AUTHINFO SIMPLE is actually
>implemented.  Isn't it purely something that got lost when the older
>NNTP-II spec got abandoned?  Both NNTP 1.5.x and INN 1.4 support
>AUTHINFO USER/PASS, as do a number of newsreaders.  Which I think meets
>the requirements for "at least two independent implementations".

The Netscape News Server uses AUTHINFO USER/PASS.  Not surprising, 
since it's based on INN 1.4.  But I think it's fair to claim that
it has diverged enough to be another independent implementation.
It uses the 480 response to indicate that authentication is
required.

The Netscape newsreader responds to 480 with AUTHINFO USER/PASS,
and 450 with AUTHINFO SIMPLE.  The 48x number space is too polluted
by common use to do anything but agree that it will be used for
authentication (IMO).  I don't know about 45x, but I suspect it
will cause a lot of problems as well.  These authentication required
responses are more problematic than a general 4xx error, because 
they often trigger a specific authentication response by the client, 
rather than just indicating an error has occurred. 

Should we add the response codes that NNTP extensions will use to
the information that is required to register them?