Re: #1273 How do we usefully define "excerpt"?

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

--On Tuesday, 02 May, 2006 21:17 +0200 Simon Josefsson
<jas at extundo.com> wrote:

> John C Klensin <john-ietf at jck.com> writes:
> 
>>> I think the discussion so far has convinced me that the
>>> appropriate way to handle this is to make the correction in
>>> commentary, e.g. "This is specified in the RFC as  BAR =
>>> numchar, but I believe that BAR = 1*numchar was meant, and I
>>> have implemented accordingly" .  I would obviously suggest
>>> putting in an erratum for the RFC as well.
>> 
>> Of course, such commentary is of the nature of a critical
>> review and does not raise any of the same copyright issues as
>> incorporating the original text in modified form.
> 
> Are you referring to US copyright "fair use"?  That doesn't
> hold universally...  granting the same, explicit, rights to
> everyone appear preferable.

No, I wasn't.  I was referring to the ability of anyone, at any
time, to say "This is an implementation of RFC nnnn, except that
section 666 of that RFC says 'BAR=numchar'.  That appears to be
incorrect and this implementation handles BAR as 1*numchar
instead."

As I understand it, that would not violate copyright laws
anywhere.   To do so would require giving the author of the RFC
ownership of particular words, such as "BAR" and "numchar" and
that just doesn't happen absent trademarks. 

One of the places where we apparently disagree is that I don't
believe it is ever _necessary_ to quote extensively from the
text of an RFC to implement the specification in one.  Inclusion
by reference (citation) works except when the material to be
excerpted  is material that cannot be expressed in any other way
without changing its meaning -- code snippets, mapping or
similar tables, header material, or the like -- and that
material is not supposed to be copyrightable.    Now we agree
that it would be undesirable to leave that material (at least)
in a situation in which one would need to take a chance,
implement, and then ask a judge.  But that is why I favor
clearly identifying it.

I also agree with Tom (and I think Ted): the right way to
"correct" SSL-> SSH is to say "He said SSL but I am sure he
meant SSH", not by making the character substitution.   The
difference between the two is substantive: the former alerts the
reader that there has been a change and that there might be a
difference of opinion about its correctness.  The second doesn't
give the reader a clue that she is expected to go back, find the
original document, and compare it with the excerpted or derived
one  to see what changes have been made under the guise of
correcting apparent typographical errors.

What I don't know is how to resolve this: I think we understand
each other's positions fairly well and I don't think our
approaching the issue from different perspectives and then
essentially repeating ourselves is going to convince either each
other or, at this point, anyone else.

      john


_______________________________________________
Ipr-wg mailing list
Ipr-wg at ietf.org
https://www1.ietf.org/mailman/listinfo/ipr-wg