Re: who has the,problem here, really?

[John C Klensin <john-ietf at jck.com>]

--On Thursday, March 12, 2009 21:49 -0400 Marshall Eubanks
<tme at multicasttech.com> wrote:

>> We largely agree. The idea that authors/editors can reasonably
>> be expected to give such a guarantee is untenable in the
>> general case. The issue seems to be whether they (and as
>> Harald rightly insists, persons helping them) can reasonably
>> be expected to do so in straightforward cases, or whether we
>> just forget about it and stick a red flag on any draft
>> containing old material.

I suggest that, for WG-produced materials, the cases that are
really straightforward may be few enough to not be worth
defining an extra case.  The situation may well be different for
independent and individual submissions.  But, having spent a lot
of time in recent years playing WG document editor/author and
editor/author of collective, multi-author works, text comes from
all over the place, it is very hard to remember how much was
significantly rewritten (retaining the original ideas but not
any of the original text or phrasing) and how much might contain
enough of the original to justify some copyright-based
attention.  

I just did a count on my own more-or-less technical RFCs in the
last five years and got three whose authorship/contributors I
could positively identify (two of those were independent
submissions) and probably find and twelve where I could not do
so with any confidence (for two of those, I'm absolutely certain
that I cannot find and get releases from one or more of those
from which they might be necessary).   I don't know if that 4:1
ratio for all of those documents (and 12:1 when only IETF or
IAB-track documents are considered) is typical but, if the ratio
is no better than that, I am not convinced that cluttering up
the world with additional cases is worth it.

Of course, if you add to the "straightforward" category
documents that people are willing to sign off on the model that
the odds of anyone actually coming after them are non-existent,
the category gets better.  On the other hand, if we actually
believe that well enough to encourage people to make that
assumption, we could probably dispense with the extra 5378
rights entirely, let the Trustees get whatever they get (whether
3978 language or 5378 language with or without disclaimers,
grant whatever rights they want to grant on the theory that no
one will ever challenge them, and we can go merrily on our way.

> My suggestion would be that it is sensible to start off asking
> for a reasonable effort. It
> should rapidly become apparent whether this is more trouble
> than it is worth.

My hypothesis is that the IETF is in for hard times as the
economy deteriorates, times in which people are pressed to make
IETF work as small a fraction of their jobs as possible
consistent with getting the IETF work done.  While they are
anecdotal, I've heard several stories of cutbacks, increased
pressure, and even lost positions in organizations that were
once very supportive of the IETF.  Against that backdrop, the
difference between "not absolutely necessary" and "not
reasonable" may become very small.  Put differently, if we
require some level of effort in hunting down prior authors and
Contributors and that costs us a single person in the critical
path for some piece of work who concludes, or whose organization
concludes, that it just makes document editing too burdensome,
then we will have shot a pretty big hole in our virtual feet.

    john