--On Thursday, 28 September, 2000 02:28 -0400 vint cerf
<vcerf(_at_)mci(_dot_)net> wrote:
perhaps I-Ds are more like elaborated lab notebooks?
very useful for patent references, reviewing dead ends, partly
explored ideas, etc. One doesn't typically throw away lab
notebooks just because you didn't write a published paper
based on them.
That is obviously closer to my model. If I correctly understand
Melinda's counterargument, it is that you don't make a few
hundred copies of your lab notebook and pass it out to your
closest friends, much less make it available to anyone who wants
it, at zero cost, as long as they agree to destroy it (or let it
self-destruct) at the end of six months.
And, if I correctly understand the implications of what Keith is
suggesting, assume you pass the notebook to me with instructions
(and an agreement) that I make it available to others for six
months, and no longer, and with a proviso that whoever receives
it should destroy (or at least not further circulate it) after
that time. Given that agreement, I'm obligated --morally, even
if not legally-- to follow your wishes. And that means that I
can't start a public archive containing your notebook(s)
independent of whether someone else might do so.
I think the notebook analogy, or one involving a
privately-circulated preliminary manuscript where the author
applies some restrictions to the readers/ recipients, are
probably the right ones. In the interest of making life easier
for historians, Melinda doesn't agree and, while I think history
is important, she and I probably aren't going to convince each
other. That said, the conclusions I'm personally drawing from
this discussion --for whatever they might be worth-- are:
(i) There is a fairly strong case for keeping I-Ds as close to a
"lab notebook" or "working preliminary draft" model as possible
consistent with reality and other interests.
(ii) Given the types of nonsense for which patents are awarded
these days, there is a compelling community interest in having
expired I-Ds available for "prior art" searches. It would be
quite rational for IETF to make author agreement to that type of
long-term availability a condition for our making the document
available in our archives or publicizing it on our "announce"
lists. Much as I'd prefer to discourage it, an author who
doesn't like those conditions is able to post a URL pointing
elsewhere, to a repository he or she controls and can destroy.
However, having material available for "prior art" searches does
not require that material be generally accessible without
restrictions.
(iii) While I recognize the value of historical research, I
don't see that requirement as so compelling to the community
that we need to make easy availability of documents for that
purpose a condition for posting using IETF facilities.
(iv) Anyone who uses a networked computer in recent years and
who has even a trace of a clue is aware that anything posted in
a public place, even briefly, is probably archived somewhere.
And, while Murphy's Law may apply, a sufficiently diligent
effort (and, often, willingness to agree to some restrictions),
will usually turn up a copy of any given piece of material.
There is, however, a huge difference between such repositories
and a public, readily-accessible, and unrestricted archive.
(v) The observation above about existing collections
notwithstanding, it seems to me that we are on thin ice (morally
if not legally) if the IETF itself decides to make documents
accepted and posted on the basis of a six-month expiration cycle
and turns them into a permanent, public, archive.
(vi) Putting Melinda's concerns aside for a moment, and looking
at "prior art" searches as the driving issue, it seems to me
that we may be approaching the problem backwards. Some members
of the community, yourself included, have periodically been able
to get the ear of one or more of the world's patent offices.
Would it be at all practical for us to help someone set up a
restricted/ historical repository of old I-Ds (and, for that
matter, old lab notebooks from those who wished to supply them).
(Of course, there need not be only one.) Such a repository
could offer searches of materials that were previously generally
available (such as I-Ds) on a fee basis and might be able to
make other materials available while enforcing author-imposed
restrictions. Would that be sufficient protection that the
community would agree to, at worst, an "opt out" provision for
older materials?
Could we sufficiently draw the attention of patent offices to
the availability of that repository to create the presumption
that anyone who hadn't searched it had not done a competent
search for prior art? Certainly it would be a resource for
someone trying to defeat a patent, whether the examiners were
aware of it or not, but it seems to me that there is a rising
societal interest in preventing stupid patents from issuing,
rather than forcing resources that ought to go into scientific
advancement to be spent on defenses against them.
john