RE: Only include implemented features in a draft standard

Jacob, what seems to be lacking in this discussion is an application of
judgement.  The IETF has never defined interoperability rules as formally as
you seem to desire them.  Most RFCs (including the MIME specs), contain what
formalists may see as a huge mishmash of requirements, implementation
advice, background discussion, non-normative definitions, and even humorous
asides.  (My personal favorite is the discussion of incorrect
implementations in RFC 1982: "Nothing can be done with these
implementations, beyond extermination.")

But, while some might say that this is no way to run a standards
organization, many (including myself) find that it the only way to do so,
and any other approach to be stiflingly bureaucratic.

So, to your specific question, no, my suggestion below is not the official
IETF policy on to how to deal with an obsolete features section.  But, it
seems reasonable to me and it may very well seem reasonable to the IESG and
the AD responsible for overseeing the interoperability report.

If you want to review all the different approaches that have been taken to
IETF interoperability testing, you might start with
<http://www.google.com/search?q=interoperability+testing+ietf&num=30&sa=Goog
le+Search>.  Instead, I would suggest that you use your judgement in
conducting the testing and then produce a report like
<http://www.w3.org/Protocols/HTTP/Forum/Reports/rollup.txt>, although
hopefully far less detailed since MHTML is a much simpler standard.

If you really want help from this mailing list or ietf-822, then I would
suggest writing up your plan for interoperability testing and having it
reviewed by the list (with the understanding that the IESG are the audience
that really counts at the end of the day).  But these abstract conversations
are not getting us anywhere, and therefore I too hereby drop out of the
thread.

		- dan
--
Daniel Kohn <mailto:dan@dankohn.com>
tel:+1-425-602-6222
http://www.dankohn.com


-----Original Message-----
From: Jacob Palme [mailto:jpalme@dsv.su.se]
Sent: Sunday, 2000-06-18 11:21
To: IETF Applications Area Discussion List
Subject: RE: Only include implemented features in a draft standard


At 03.25 -0700 0-06-18, Dan Kohn wrote:
> Two different implementations must be able to correctly parse obsolete
> messages.  Implementations should never generate those messages.  The
> easiest way to do the test (and hopefully every MUA vendor will do this)
is
> simply to telnet to port 25 and feed in the obsolete message from the
> example.  Thus, the obsolete section would be interoperability tested if
> both implementations produced the same parsed message.  There is no
> requirement for interoperability on generation, and in fact, MUAs are
> required NOT to generate such messages.  Thus, your second sentence
doesn't
> follow.

I thought interoperability meant that one implementation
could receive what another implementation produced. You say
that interoperability means that two implementations will
handle the same manually generated input in the same way.
Is that really right? Or is that a special variant of
interoperability, to be applied to "receive-only" specs,
while "send-and-receive" specs will still be tested by
testing that one implementation can receive what another
implementation sends?
-- 
Jacob Palme <jpalme@dsv.su.se> (Stockholm University and KTH)
for more info see URL: http://www.dsv.su.se/jpalme/

Received on Sunday, 18 June 2000 18:28:06 UTC