On 22:45 10/05/2005, Dean Anderson said:
The IETF doesn't produce documents that are meant to be accessible to
users. Nor does it produce reference implementations. IETF documents are
meant to be accessible to engineers and operators, creating and running
interoperable services of various types. One and possibly two
implementations are usually required for a standard to be acceptable. The
point of this is to require that specifications be both implementable and
the IETF users are the ones who use the IETF deliverables. And the Internet
is made of the adherence to these deliverable (which do not make the best
documentation around). So, you cannot know or decide who will be the users.
(may be you refer in your mind to "end users"? There are none on the
Internet because it is not a centralised network.
Just remember that Internet users have designed P2P, VoIP, NATs, etc. for
some and that others have more processing and communication powers than the
whole AT&T 20 years ago) and you never know who will need what (even if
99.99% will never read an RFC, they will all read RFC quotes and they must
be clear and consistent with what their consultant, their ISP, their
operator will tell them). What I mean is there is no Internet "gnosis",
there should only be an Internet "gospel".
There is no shame in it, but the IETF mechanic seems to be more:
- a user comes with a working project
- IETF starts maintaining it and documenting it what helps its integration
- if the whole common thing is accepted and works it can become a standard.
Ietf mailing list