Re: [Fwd: I-D Action:draft-carpenter-5378-old-text-02.txt]

[Brian E Carpenter <brian.e.carpenter at gmail.com>]

John,

On 2009-05-12 15:22, John C Klensin wrote:
> 
> --On Tuesday, May 12, 2009 09:29 +1200 Brian E Carpenter
> <brian.e.carpenter at gmail.com> wrote:
> 
>> This version of the draft describes a completely modified
>> procedure compared to the one presented at IETF74, which did
>> not obtain consensus. It contains small and large changes
>> throughout.
>>
>> Discussion is invited on the ipr-wg at ietf.org list.
> 
> Brian,
> 
> My reading of the rough consensus during and after IETF 74 was
> not "no consensus for draft-carpenter-5378-old-text-01.txt,
> please try again".  It was "accept the 'interim' Trust
> disclaimer as permanent, without further attempting to tune
> details, and stop" -- with the "decide we are done and stop"
> part being important.

Well, the intention of this draft, I think, is to say exactly
that, and to leave the rest to the Trust. I don't see how, when there's
a known gap in the BCP, we can fail to amend the BCP; a feeling in
a room in the SF Hilton doesn't do that. But basically, the intention
is to achieve exactly what you say.

Of course, there would be another form of amendment which simply says
"the Trust is directed to fix the gap we left in 5378". I won't be writing
that draft personally, but we could take that approach.

> 
> In addition, if Larry Rosen's theory is correct, this entire
> discussion is immaterial.   To the best of my knowledge, we have
> received no advice or instructions from the Trust or its Counsel
> on that subject.  But, were advice to appear that we can call
> rely on Larry's theory, we would also be finished and your draft
> would be irrelevant at best and a serious distraction at worst.

Unless we hear from our counsel and the Trust, I think we have
to assume that Larry's theory doesn't apply. BTW that means
internationally, not just under US law.

> 
> It is possible that, despite that impression from IETF 74, there
> are a significant number of active IETF participants who want to
> pursue this.  If there are, I hope they will identify
> themselves.  In their absence, I trust the rest of us can and
> will return to productive work.

That's entirely the intent and the spirit in which -02 was written.
I've spent a *lot* more time on technical drafts since SFO than
on this.

> 
> If you are going forward with this nonetheless, let me suggest
> three things for you to consider in your next iteration:
> 
> (1) In the first paragraph of Section 2, you appear to lay a
> foundation for claiming that all Original Contributors who need
> to be considered will be listed in Acknowledgments sections.
> However, there has never been a requirement (at least prior to
> 5378) that authors or editors track these Contributors in detail
> and document Contributions down to the Section level so that the
> author of editor of a follow-on document can identify that work,
> as your second bullet on Page 4 (toward the end of Section 2)
> apparently anticipates.  Indeed, many IETF documents contain
> generic acknowledgments to unspecified participants in a WG.
> Neither the IESG nor the IETF Trust have ever objected to such
> acknowledgments, at least not to the extent of blocking
> publication until specific names are provided.   My guess is
> that we would be hard-pressed to identify all participants in a
> WG during the time a particular document and its predecessors
> were being worked on, even though I could imagine somewhat
> painful ways to approximate that list.

That is supposed to be covered by the various escape clauses:
"to the best of their knowledge and ability" "This requirement
does not extend to small fragments of text culled from IETF
discussions." "This requirement does not apply if the text
concerned is scattered throughout the document."
Personally I'd probably include more weasel words of this kind,
but Jorge's advice has rather been against that. The intention
was not to add to the quantity of acknowledgements already required
by RFC2026 and its descendants, and you may have noticed "shoulds"
in front of the new requirements.

> Remember too that one of the topics on which I believe that we
> had fairly clear (although not complete) consensus prior to IETF
> 74 was that a model that put more burden --either of extra work
> or of legal liability-- on authors or editors was not in the
> best interests of the IETF.  Both the suggestion that authors
> should start tracking and documenting contributions down to the
> paragraph or section level and the implication that authors
> should make statements on behalf of "all Contributors" based on
> who is explicitly listed contradicts that principle.

You're stating a consensus that I only saw as an opinion. I did
read the minutes of the BOF quite carefully, and the open question
at the end was:

" Is it incumbent on authors to identify the source of material beyond normal
 attribution? "

So, a consensus call on section 2 of the new draft is really a precise
version of that question from the BOF.

> 
> (2) A second key principle is that we should not leave authors
> or editors in situations in which their own legal Counsel is, in
> good faith, advising them that Contributing work to the IETF is
> ill-advised because of IPR traps.  Unless the IETF Trust (or, in
> principle, you and your University and Harald and Google) are
> willing to give legal advice and protect authors against the
> consequences of following that advice, the disclaimers in 5378
> and elsewhere apply.  When your document imposes a requirement
> based on a November 10 cutover and effectively insists that, as
> an author, I may rely on you for a belief that any material
> contributed after 10 November was contributed consistent with
> the rules and licenses required by 5378.

I really don't see what's new here. People have always had to
take responsibility for what they submit, and the fall-back advice
from the IETF has always been "if in doubt, ask your own lawyer."
The draft is fairly clear that it doesn't offer legal advice,
and it does suggest using the disclaimer in case of doubt. The
problem for a submitter is caused by the change of rules in
5378.

> 
> Now consider a situation that one might hope was purely
> hypothetical (it is not).  You (or the Trust) publish a document
> that says "you can rely on publication after November 10, even
> if the publication occurs on November 11th".   The relevant
> author consults his or her own Counsel, as 5378 and its
> predecessors rather strongly recommend, and is told "you
> absolutely cannot rely on the November 10th date because, if
> someone wants to make a claim against rights not granted before
> 5378, they will point out that there is no reasonable basis on
> which they could have been expected to know that 5378 was
> applicable, especially since the old Note Well was used at IETF
> 73 a week later".   It doesn't make any difference whether the
> author's lawyer is "right" or not(that would be determined only
> by a judge if someone actually did make such a claim).  The
> problem is that, if you insisted that the author could rely on
> November 10 and the author got legal advice that such reliance
> was unwise, the author would have two choices: to obtain formal
> releases from everyone who might be considered a Contributor (a
> requirement I believe the IETF has rejected) or to not post the
> relevant document (which hurts the community).
> 
> I think the conclusion from this is the one others have
> suggested: either the Trust has to offer people the legal advice
> that they can rely on the November 10 date _and_ indemnify them
> against any bad consequences of that date being challenged or
> the date is meaningless.

I think the draft is phrased to allow the use of the disclaimer
in case of any kind of doubt including this kind of doubt.
(If it isn't, that can be fixed, of course.) That seems necessary
and sufficient.

> 
> 
> (3) One of the key mandates to this incarnation of the IPR WG
> was that having the WG, or any other amateur or collection of
> amateurs, write legal text or specific legal provisions was a
> bad idea, characterized by repeated failures in the past.  The
> WG was chartered to not try that again but rather to specify
> goals and objectives with the expectation that the Trust and
> Counsel would then write text.  So now we have a document,
> written under the names of two such amateur lawyers, that
> specifies legal text.  That is inappropriate and unwise at best.

Well, it was reviewed by Jorge before submission, and the intention
was to have as small a normative section as possible with an
explicit direction to the Trust to do the rest. If you trust the
Trust to take over the whole mess, then a much shorter draft
would do the trick.

     Brian