From: Bill Manning <bmanning(_at_)ISI(_dot_)EDU>
...
I think things are headed in that general direction and I think it is a
sad state of affairs. Historically, RFCs were used to document ideas,
both good and bad. The series covered the range of idea generation
and expression and this was encouraged by the RFC editor. Now, many
"non-conformant" ideas are being released into the community as defacto
documents since it is too hard to get material documented in the RFC series
as either informational or experimental. This is in part due to the IESG/IAB
review cycle that is now part of the process.
The alternative is many more and far worse circuses.
What happens when someone writes something like
draft-terrell-logic-analy-bin-ip-spec-ipv7-ipv8-08.txt without what seems
to be that author's non-confrontational nature but with the attitudes of
D.J.Burnstein and Mohsen BANAN?
In the old days, the IETF was in fact closed because no one knew about
it, and the RFC editors did real editing. The barriers to outsiders in
1986 were informal but much higher than they are now. Even had he heard
of RFC's and wanted to publish some ideas as one 1985, Mohsen BANAN could
not have, at least not in the forms the modern IETF allowed. As the public
discovered TCP/IP, the RFC printing press was in effect forced open. The
relatively recent attempt to make the IETF a free vanity publisher is a
growing, inevitable disaster that pleases no one.
Things have changed in the last 15 or 20 years. The publishing services
of the IETF and (and ISI) are no longer needed except for official products
of working groups and perhaps except for open mailing lists. Each of us
can trivially operate our own vanity publishing houses. Consider how very
much better things would have been for everyone including Mohsen BANAN
had the IETF said "those ideas are interesting, but until they have been
implementated and tested, and thereby attracted an IETF Working Group to
sponsoring them, all we can do is suggest that you mention a mailing list
and URL for your documents in the IETF mailing lists." And similarly for
TAP vs. Ident and whichever other excitements involving DJB that Mohsen
BANAN is talking about.
This practice leads to
closed environments, documentation, code and operations. Not what I would
have hoped for in an evolved Internet.
That is clearly wrong at least about code. Close vs. open code has nothing
one way or another to do with the IETF's publishing polices, as
demonstrated by the closed and open implementations of offical IETF
protocols, not to mention the "individual submissions." Besides, there
has been practically no code ever published by the IETF. I suspect one
RFC contains most of the code in all existing RFC's.
It's also not clear that open publishing leads to open protocols or
anything else besides open (esp. vanity) publishing. Contrast the
protocols that Microsoft has described with RFC's with the availability
of source implementing them and the chance of anyone outside Redmod fixing
or convincing anyone to fix even their most egregious and obvious bugs.
What made sense as a semi-formal way to slightly polish items discussed
in mailing lists has nothing to do with the polices, procedures and
even goals of a competitor or co-equal of the ANSI and IEEE and member
of the ITU.
The IETF should keep the best parts of the RFC process, including free
access to RFC text. However, for better or worse, the IETF doesn't
have the organization, policies, procedures, people, or billing systems
to operate a major vanity press catering to the under appreciated
computer geniuses of the world.
Vernon Schryver vjs(_at_)rhyolite(_dot_)com