Re: [opendocument-users] New question (2): Reference implementation?

[marbux <marbux@gmail.com>]

On Mon, May 18, 2009 at 9:12 AM, Cody, John (OFT)
<John.Cody@oft.state.ny.us> wrote:
> Paul:
>
> Let's try this:
>
> 1.  You said you are not aware of a legal definition of "reference implementation."  Then how can "governing law" forbid them, as you say, if it doesn't even define them?

That's because without a legal definition of the term, the canons of
statutory construction teach us to use the common and ordinary
meanings of terms unless to do so would create a conflict with other
provisions of the relevant law or produce an absurd result. In this
instance, what constitutes a "reference implementation" is fairly well
understood. I refer you to Rick Jelliffe's excellent post and to
section 3.2.2 on this Gessmer & Updegrove page.
<http://www.consortiuminfo.org/essentialguide/certification.php>.

Perhaps I unintentionally erected a straw man here, but considering
your position and employment, I presumed that your reason for asking
was that you were considering the incorporation by reference of
OpenOffice.org as part of a formal or informal rulemaking procedure,
i.e., in a state SOA technical regulation or a procurement technical
specification.. Did I get that wrong?

> 2.  You base your international trade legal argument on portions of the ATBT discussing "technical regulations."  But "technical regulation" is a defined term, and it is defined as something that is *mandatory.*  So it would seem to have no applicability to for example an appreciation of OO.org as a de facto reference implementation for all practical purposes.

I suspect that you missed the significance of the fact that the same
specification must concurrently serve under the ATBT as both an
international standard and a technical regulation, and under the AGP
as a government procurement technical specification. The legal effect
of the document, e.g., its mandatory or voluntary nature, may vary in
those three contexts, but the specification itself must perforce be
identical in all three contexts.

Therefore, the relevant requirements for technical regulations and
procurement technical specifications provide definition for the term,
"international standard" *as to such a document's content.* So the
Asbestos Panel's ruling reads directly on what content must appear in
an international standard and how it must be phrased.

On "an appreciation of OO.org as a de facto reference implementation
for all practical purposes," I am unsure of what you mean. E.g., if
you had in mind specifying in a procurement tender that the
implementations of ODF you seek bids on must be OpenOffice.org or
implementations with equivalent support for ODF, then you collide with
the AGP requirements I discussed in my last post. If you had in mind
specifying OOo or implementations with equivalent ODF support as part
of New York State's SOA, then you collide with the requirements for
technical regulations discussed in my last post.

Kind of the bottom line here, I think, is that de facto standards and
de jure standards don't mix legally. If one wants a de facto standard
to become a de jure standard, then it must go through the de jure
standardization process and must fulfill the requirements for a de
jure standard.

I realize that you used the term "de facto reference implementation"
rather than "de facto standards."  But you also coupled that term with
"for all practical purposes," which leaves me wondering whether you
intended thereby to encompass New York State's procurement tenders and
technical regulations with the latter term.

> 3.  Finally, using a missing definition to a term which is applicable, and an existent definition to a term which isn't, you draw conclusions as to the illegality of ODF as an international standard.
>
> - Are there any analyses which review these issues and come to the same conclusion you do, Paul, that you can point me to?
> - Are there any analyses which review these issues and come to a different conclusion you do, Paul, that you can point me to?

None that I can point you to. When I've raised such issues in the past
in regard to ODF, all I've encountered are arguments that evade rather
than addressing the merits of the legal analyses. I could point you to
any number of articles that challenge OOXML on grounds of being
vendor- and application-specific, which is very similar to the OOo
reference implementation issue.  See e.g.,
<http://www.sutor.com/newsite/blog-open/?p=1145> and
<http://www.sutor.com/newsite/blog-open/?p=1260>, the latter of which
is a compressed "grandmother's" view of the legal arguments IBM is
making through ECIS in the pending antitrust investigation of
Microsoft Office in the E.U. The linked articles are compressed
dissertations on the relevant antitrust law sans citations.

But IBM et al have been decidedly uninterested in applying the same
legal principles to ODF and OpenOffice.org, even uninterested in
discussing the subject.

The ATBT and AGP have been far more ignored than paid heed,
particularly in the IT sector. At the same time, one can't do a really
thorough job of researching international law without traveling to New
York City and Washington, D.C. My traveling days are done. I've been
very thorough as to the materials publicly available on the internet
and the University of Oregon Law Library. But I could point you to any
number of related data gaps.

> - Are there any legal decisions squarely on all fours, e.g. determining use of a "reference implementation" to have rendered a standard illegal, that you can point me to?

Nope. Thus far we have precisely two pronouncements on the ATBT by the
WTO Appellate Panel. Neither is on all fours with the issue you raised
unless you contemplate incorporating OOo by reference into a NYS
technical regulation. In that context, the Asbestos Panel's ruling
couldn't be reconciled with the action contemplated. And because of
the requirements that international standards be concurrently used
both as as technical regulations and government procurement
specifications, I'd argue that the same Panel decision is very nearly
on all fours in regard to procurement specifications.

> I'm not saying you're wrong (I don't have enough experience with international trade law or standards law to say definitively).  All I am saying is the legal conclusion you thus far have provided while presented by you as rock-solid reads to me, with my skeptical lawyer's eye, to be based on an incredibly attenuated string of support much of which at first blush doesn't even seem relevant to the question I asked.

It might help to realize that the relevant treaties are largely
competition law, with implementation and enforcement delegated to the
Member nations except when disputes arise between member nations.  And
if you look at ATBT Article 15 section 15.2, you'll see that Member
nations can implement via existing laws, new laws, or combinations
thereof. (You'll probably find an equivalent provision in the AGP but
I haven't looked.)

In that regard, the prohibitions against standards that create
unnecessary obstacles to international trade are largely a restatement
of antitrust law in the U.S. and E.U. that regulate industry standards
development consortia like OASIS and its members. In the U.S., such
bodies are primarily regulated under section 1 of the Sherman Act,
which prohibits unreasonable conspiracies in restraint of trade, in
combination with provisions of the Clayton Act that provide DoJ with
civil enforcement authority and private parties with a right of
action. See e.g., Allied Tube & Conduit v. Indian Head, Inc. 486 U.S.
492 (1988), <http://laws.findlaw.com/us/486/492.html> ("trade and
standard-setting associations routinely treated as continuing
conspiracies of their members"). This leads into a rather enormous
body of antitrust case law dealing with collaborations among
competitors in horizontal and vertical markets.

So while the connection between the ATBT and antitrust law might
appear at first blush to be attenuated, it isn't. The U.S. fulfills
many of its ATBT obligations under antitrust law. And I suspect your
grandmother might miss that even if explained. :-)

I'm sensitive to how daunting and confusing it can be to wade into new
areas of the law. Been there and done that many times over. :-) But it
would help me to be less diffuse if you're in a position to disclose
what specific hypothesized action you are researching in regard to
OOo.

> I'm curious what the IBM standards lawyers, Sun standards lawyers, Adobe standards lawyers etc. (surely they have them) Gesmer standards law firm, etc. think of this?

Haven't seen anything by any of the above on the particular subject at
hand that cites any legal authority but I read far less than
everything they write. You might try Updegrove's site search engine.
<http://consortiuminfo.org/essentialguide/>. His clients include IBM,
Sun, and OASIS.

Can't say I endorse any of them as a source of reliable information on
related subjects. You'll undoubtedly learn that the opinion is mutual
if you talk with them. My relationship with them has been largely
adversarial and I've had issues with their choice of particular
tactics.

Best regards,

Paul E. Merrell, J.D. (Marbux)

-- 
Universal Interoperability Council
<http:www.universal-interop-council.org>