[NNTP] requirements for standardization of NNTP
Mark Crispin
mrc at CAC.Washington.EDU
Mon Apr 25 11:56:39 PDT 2005
Before NNTP can be standardized, it must go through a great many rigorous
steps, in which individual opinions are not going to count (and this
includes the opinions of the document editor).
These are IETF requirements; the NNTP working group can not override them.
Among these requirements is an implementation report which shows multiple
independent interoperable client and server implementations of every
aspect of the protocol. Those aspects of the Proposed Standard protocol
which do not have multiple independent client and server implementations
will be removed from the document that describes the Draft Standard.
The above needs to be kept in mind. I disagree sharply with what the
document editor considers to be "clutter"; but ultimately neither his
razor nor mine matter -- the Draft Standard protocol will be cut with a
much more aggressive razor.
Once again, I recommend that we split this document into two separate
documents:
One document, to be published as Informational, will attempt to describe
"what is"; that is, the NNTP protocol as is implemented by various client
and server implementations that are deployed around the world. The fact
that an implementation does (or depends upon) such-and-such is sufficient
to merit inclusion in this document.
The other document, to be published as a Proposed Standard, will attempt
to describe "what should be"; that is, the NNTP protocol as disambiguated,
and cleaned up. It is expected that no current client or server will
conform perfectly to this document; thus, there must be a way for a
conformant client and server to identify each other.
It is not necessary for a non-conformant implementation to interoperate
with a conformant implementation. It is desirable for the short-term; and
some effort should be made in producing an interoperability document
focusing on how a conformant server can interoperate with an old client.
Nevertheless, the long-term will require conformance. Old software,
especially old servers, will eventually die.
-- Mark --
http://staff.washington.edu/mrc
Science does not emerge from voting, party politics, or public debate.
Si vis pacem, para bellum.
More information about the ietf-nntp
mailing list