Re: Resolving the "TIFF Issue"

[John C Klensin <klensin@xxxxxxx>]

Scott,

Several comments inline below...

--On Tuesday, 25 September, 2001 16:30 -0700 Scott Foshee
<sfoshee@xxxxxxxxx> wrote:
>...
> 1.  It is important to distinguish between two issues (per the
> IAB chair's presentation in London):  interoperability and IP.
> It is my understanding that the primary reason the WG in
> London chose to scale things back was interoperability. 

The nature of the IETF policy on intellectual property rights
and claims is, in essence, to make IPR claims an
interoperability issue.  In other words, it is the
responsibility of the WG to select solutions that can be
independently implemented and interoperate.  The WG may select
solutions that require licensing (either within or outside
license grants to the IETF); if it does, then the
interoperability demonstrations must show that whatever
licensing is needed can be successfully exercised.  In
principle, that might even involve demonstrations that the IPR
claims were not sustainable, that litigation against
implementors was unsuccessful, etc.

Given that orientation, once IPR claims have been raised in the
context of a decision already made by a WG, and especially if
those IPR claims include the assertion that the technology
chosen is outside an existing license grant or other licensing
assertions, it is not possible to separate "interoperability"
and "IPR" issues as clearly as your note implies.

> I
> refer you to his briefing for his rational, but I understand
> it to be based on general IETF interoperability principles and
> specific IETF WG goals for TIFF FX that resulted in the
> selection of TIFF as the basis for work.

See above.  Certainly I also expressed a preference at the
meeting for a base specification that interoperated as broadly
as possible, with both fax and email implementations (especially
given some other choices made by the WG).  My understanding at
the London meeting was that the preference of the people present
was to go with a two-track approach, moving the portion of the
specification that had greatest interoperability (and no
collisions with the Adobe claims as I, at least, understood them
at the time) into Draft Standard, and recycling the more
problematic data representations at Proposed.  I didn't
understand there to be any plan to drop those additional forms
entirely.

But, as you know, WGs don't make final decisions at IETF
meetings, they make them on mailing lists.  And my understanding
of the mailing list discussions is that the additional forms are
sufficiently implemented and deployed that there is strong
sentiment in the WG for preserving things as a single standard,
recycling at Proposed if needed.  That decision --one way or
another-- is well within the scope of the WG to recommend to the
IESG.

>  Adobe is not the
> cause of this decision or action, although we support it.

You may be misinterpreting the decision.  See above.

> Adobe's IP issue was specifically not addressed. The
> implementation of the London WG decision has the side effect
> of mostly avoiding the IP issue (deferred until the remaining
> functionality of TIFF FX is investigated).  It is my
> understanding the the IAB, IESG, and working group chairs want
> us to evaluate this decision primarily on interoperability
> grounds.  My impression from the London WG meeting is that
> they support the London decision.

I will support _any_ decision the WG makes that is consistent
with the general interoperability principles stated above and in
London.  I believe that the relevant IAB and IESG members will
too.

> 2.  If the Internet fax working group implements the
> recommendations of the London meeting (focusing on the
> immediate progression of mostly of S/F), this
> "interoperability choice" would mostly avoid the IP issue.
> The Adobe IP issue remains significant for all other profiles
> of TIFF FX, and the working group implicitly considers these
> issues when it considers progression of those profiles.

I am not sure what you are intending to say here, but, if I
correct understand it, I agree.

> 3.  It has been emphasized to me that the IETF as an
> organization does not make statements nor take positions
> regarding the IP claims of companies.  However, it is
> certainly true that the members/companies of the working group
> are making decisions regarding what technologies to include in
> TIFF FX and are making decisions regarding implementation of
> TIFF FX by their companies given (1) our original license, (2)
> our communication with IETF editors, and (3) our communication
> with the IESG and IAB.  Adobe continues to indicate that TIFF
> FX is outside the scope of our license grant.  The IETF is
> deferring a position on this to the membership/implementing
> companies and whatever process they use to evaluate IP and
> licenses.

The latter is the only thing the IETF ever does as an IETF
decision.  Since these situations can result in proto-standards
getting into an impasse, especially if one or more company
actors take positions that seem to shift to block progress (I am
not suggesting that is happening here), we also have a history
of trying to encourage the companies involved to work the issues
out, creating licensing grants that make it feasible to develop
interoperable implementations of whatever the WG considers the
best choice of technology.

> 4.  Finally, a bit of history (in response to an e-mail that
> asked how we got here).  Adobe provided a license to the IETF
> and ITU in 9/97 for the use of TIFF as the basis for the
> interchange of FAX data on the Internet.

I have been assuming that included the "interchange of FAX data"
in whatever transport format, including as ordinary email
messages.  Part of your message seems to imply otherwise, and I
would appreciate it if you, or some responsible party who can
speak for Adobe, would clarify this once and for all.

>  The 3/98 draft of
> TIFF FX presumed the publication of TIFF 7 with certain
> content which was never incorporated into TIFF. 

It is my understanding that it presumed the use of certain
public tags that the WG had been told, by a participant and
Officer of Adobe, that it could use for this purpose.   The
binding of these tags to TIFF 7 (or any other unreleased version
of TIFF), occurred in subsequent communication between Adobe and
the WG (and IETF more generally).

> TIFF FX was
> progressed despite this disconnect... and despite repeated
> Adobe/IETF editor (Xerox) discussions that inclusion of these
> features were not certain....and thus TIFF FX was left outside
> the scope of the Adobe license.  

That issue appears to hinge critically on whether the tags that
an Adobe Officer and participant in the WG gave to the WG and
advised the WG to use were included in the scope of the license.
If they were not, the question has been raised as to whether
Adobe properly, and in a timely way, complied with the IPR
disclosure provisions of RFC 2026 and its predecessors.   As
with the IPR claims themselves, the IETF will take no formal
position on that issue.

> Other than cautioning the
> editors, Adobe has not participated in the working group since
> the beginning of 98.  When the editors progressed the document
> despite objections, we elevated objections to IETF management
> last December and the IAB this Spring. The listed Adobe editor
> has not reviewed the document since prior to the 3/98
> publication and recently asked that his name be removed from
> the latest version (it has yet to be removed by the continuing
> editors).  Early this year Adobe provided the IETF, at the WG
> chairs request, a formal process by which any third party can
> request content in the next version of TIFF.  We have yet to
> receive any requests via this procedure for us to incorporate
> the presumed/missing content into TIFF.

My understanding from conversations with you was that (i) it
would be pointless to make such a request unless all IRP claims
impacting the technology used in that content was released so
that Adobe could utilize it in _any_ TIFF context, not just Fax
and (ii) there was no guarantee that, if such a request were
made, Adobe would, in fact, incorporate the relevant content,
nor when a new version of TIFF would appear, if ever, nor as to
when (or if) any decisions Adobe made on whether to incorporate
the relevant content would be disclosed to the WG.   If I have
gotten any of that wrong, please correct me.

>       Note that:
>       TIFF 6 is available on our web site.
>       Our license is available on the IETF and ITU web sites.
>       Our December 2000 e-mail should be available from the
> IETF.       Our process should be available from the WG chairs.
 
> In close, I think the London proposal is a good way to move
> things forward fast.  While splitting the document has
> disadvantages, it would allow rapid progression of the parts
> that are interoperable with existing TIFF (but non-TIFF FX
> specific) readers.  The alternative leaves the IETF and anyone
> who wishes to implement TIFF FX with a candidate TIFF FX
> specification that Adobe  has identified as being a use of
> TIFF that is outside the scope of the 9/97 license grant to
> the IETF.

We understand this to be your position.   My understanding of
the WG discussions is that the active participants in the WG
have reached a different conclusion, based on the need to
support advanced features and formats in Internet fax and the
presence of those features and formats in deployed equipment.

> Adobe is committed to a timely review of any requests for
> changes to TIFF and is committed to working with the IETF to
> ensure a TIFF standard that is interoperable.

I, at least, appreciate this additional committment on Adobe's
part.

   John Klensin
   Chair, IAB