I've got very mixed feelings about this topic: I understand the
reasons for an archive, but I'm extremely concerned that, as
Keith, Ran, and others have pointed out, turning I-Ds into a
permanent record may make it harder to get half-baked ideas
exposed in the community rapidly enough that we can make
progress.
I think that this is probably the heart of the
problem. It's not a matter of turning drafts
into a permanent record - for better or worse
they already are a permanent record, at least
until such time as the technology is developed
to obliterate every copy everywhere stored on
every medium, blah blah blah.
It seems to me that the question at hand is
whether or not this gives drafts any additional
weight in standards-land. The answer to that
question lies in how references are controlled.
Given that they can be controlled quite tightly
through the review process, I'm not sure I see
the problem.
But, when they are retained and available at all, they
are traditionally also severely encumbered as to access.
This situation is quite different, I think. An internet
draft is not limited in its distribution. It's
limited temporally, and I really can't think of anything
analogous elsewhere.
Restrictions such as "no access during the author's lifetime",
"no access while the copyright on the final publication is
still
in effect", or "no access without special arrangements that
require identification of the researcher and purpose of the
research" are all, as you must know, fairly common.
Sure, but this is different. We're not talking about
private documents being passed between two private
parties. Internet drafts are available to anybody
who wants them, and there are effectively no restrictions
on reproducing and archiving them. What is very different
is that they "expire" after six months. I think that
one thing that might come out of this discussion is
a definition of what it means to expire a document in
the context of the IETF.
Instead, I
think it would be helpful if the people and communities with
experience in the area, especially the library one, would help
us find appropriate analogies to the restrictions on access
that
often characterize collections of what may be similar
character.
This just strikes me as wrong. I think that the
most desirable and realistically achievable result
would be what we've already got - you can't cite the
things in standards (and I might note that the ITU-T
and ETSI already specify that you can't reference
internet drafts in their documents), but otherwise
the information is out there. Heaven knows I've said
things on mailing lists that I'd like to retract, but
some things are just beyond my control.
Melinda