ietf-nntp NNTP AUTH draft

[Stan O. Barber]

Andrew,

ANY authentication in current NNTP implementation is a LOCAL extension,
not a standard one. We are specifying a STANDARD one. As such, it needs new
response codes. 

Good server design should be able to accomodate both standard extensions and
local extensions. Good clients should be able to request both standard and local
extensions based on information it gets from the
server (using LIST EXTENSIONS, which is part of the new standard 
extension process). If the client can't get this from the server, then
the server should be thought of as a RFC 977 server that MAY support
some well known, but non-the-less local extensions.

It's your choice as a server designer how you choose to support the legacy RFC
977 + well-known local extensions in the context of the new RFC977bis specs.


Andrew Gierth wrote:
> 
> >>>>> "Chris" == Chris Newman <chris.newman at INNOSOFT.COM> writes:
> 
>  Chris> I'll do a sanity check of the examples and error codes against
>  Chris> the most recent base spec, update the examples to use
>  Chris> DIGEST-MD5, and update the boilerplate text.  Are there any
>  Chris> other open issues?
> 
> Why are you trying to change the response codes?
> 
> This is a major obstacle to any attempt at implementation - as a
> server author I will be unable for the forseeable future to use any
> response codes other than the existing 480 etc., because that would
> break compatibility with clients.
> 
> --
> Andrew.