ODF 1.2cd01 --- AWOL interoperability conformity requirements

[marbux <marbux@gmail.com>]

Committee Draft 1 includes 3, 449 occurrences of MAY, 20 occurrences
of MAY NOT, 5 occurrences of NEED NOT, and 28 occurrences of OPTIONAL,
used almost exclusively to express options. This total of 3,502
occurrences of terms indicative of options is as opposed to only 272
occurrences of requirement keyword terms indicative of mandatory
requirements.

This condition does not measure up to governing requirements.

Under ISO/IEC JTC 1 Directives, (5th Ed., v. 3.0, 5 April 2007), Annex
I, pg. 145, <http://www.jtc1sc34.org/repository/0856rev.pdf>:

"Standards designed to facilitate interoperability need to specify
clearly and unambiguously the conformity requirements that are
essential to achieve the interoperability. Complexity and the number
of options should be kept to a minimum[.]"

At ibid, pg. 11, those requirements are made mandatory for all JTC 1
standards absent the consent of the Secretaries-General of ISO and
IEC. See also Agreement on Technical Barriers to Trade article 2
section 2.2, <http://www.wto.org/english/docs_e/legal_e/17-tbt_e.htm#articleII>:

"Members shall ensure that technical regulations are not prepared,
adopted or applied with a view to or with the effect of creating
unnecessary obstacles to international trade. ..."

See also ibid. section 2.6 (preparation of international standards is
one method of preparing "technical regulations"). The Agreement thus
raises the issue of whether interoperability barriers created by
international standards are "unnecessary obstacles to international
trade." In my studied opinion, the JTC 1 Directives interoperability
requirements should be understood as the ISO/IEC translation of the
Agreement provision quoted above into the IT international standard
context.

See e.g., World Trade Organization Committee on Technical Barriers to
Trade, Decisions and Recommendations Adopted by the Committee Since 1
January 1995 (May 23, 2002), pg. 28,
<http://docsonline.wto.org/imrd/directdoc.asp?DDFDocuments/t/G/TBT/1R8.doc>
(international standards "should not distort the global market, have
adverse effects on fair competition, or stifle innovation and
technological development").

In the IT specifications sector, full specification of
interoperability conformity requirements is fundamental to fair
competition among implementers. See e.g., Commission v. Microsoft,
European Community Court of First Instance (Grand Chamber Judgment of
17 September, 2007), para. 226, 230, 421,
http://curia.europa.eu/jurisp/cgi-bin/gettext.pl?where=&lang=en&num=79929082T19040201&doc=T&ouvert=T&seance=ARRET
(rejecting Microsoft's argument that "interoperability" has a 1-way
rather than 2-way meaning; information technology specifications must
be disclosed with sufficient specificity to place competitors on an
"equal footing" in regard to interoperability; "the 12th recital to
Directive 91/250 defines interoperability as 'the ability to exchange
information and mutually to use the information which has been
exchanged'").

OASIS ODF v. 1.0 section 1.2 incorporated by reference the  RFC 2119
requirement keyword definitions that bounded each occurrence of the
terms MAY and OPTIONAL with two mandatory interoperability
requirements:

"5. MAY -- This word, or the adjective "OPTIONAL", mean that an item
is truly optional. One vendor may choose to include the item because a
particular marketplace requires it or because the vendor feels that it
enhances the product while another vendor may omit the same item.

"An implementation which does not include a particular option MUST be
prepared to interoperate with another implementation which does
include the option, though perhaps with reduced functionality.

"In the same vein an implementation which does include a particular
option MUST be prepared to interoperate with another implementation
which does not include the option (except, of course, for the feature
the option provides.)"

RFC 2119, <http://www.ietf.org/rfc/rfc2119.txt> (broken into separate
paragraphs for legibility).

The result was that in OASIS ODF 1.0 --- counting each option as
including two mandatory requirements --- the specification had 4,505
occurrences of terms indicative of options whilst having 7,415
occurrences of terms indicative of mandatory requirements.

However, at JTC 1, British Standards Institute included the following
comment on DIS ISO/IEC:26300 along with its ballot:

"The text does not conform to Annex H of the ISO Directives defining
acceptable verbal forms for the expression of provisions (i.e. the use
of "shall", "should", etc.). It uses the conventions of IETF RFC 2119
instead. Our understanding is that the PAS procedure does not allow
exemption from this requirement."

BSI recommended:

"Edit all provisions to conform to Annex H of the ISO Directives."

See BSI comments in zipped attachment to ISO/IEC JTC 1/SC 34N0728rev,
Summary of Voting on DIS ISO/IEC 26300 - Open Document Format for
Office Applications (OpenDocument) v1.0 (13 June 2006), pg. 1,
<http://www.open-std.org/JTC1/sc34old/repository/0728revc.htm>.

Unfortunately, BSI's "all provisions" phrase was ignored and only
section 1.2 was modified, dropping the incorporation by reference of
RFC 2119 and substituting in its stead a reference to ISO/IEC
DIrectives Part 2 Annex H definitions. ISO/IEC:26300-2006 was left
liberally strewn with undefined requirement keywords.

More importantly, no editing occurred to restore the up to 7,192
instances of mandatory interoperability requirements snuffed out of
existence by the flip in the requirement keyword definitions. Under
ISO/IEC Directives Part 2 Annex H definitions, OPTIONAL is undefined
and MAY takes its common and ordinary meaning of "permission,"
unbounded by mandatory interoperability requirements.

While the Editors have made progress in Committee Draft 1 in replacing
the occurrences of OPTIONAL with MAY, there appears to be no progress
at all on restoring mandatory interoperability requirements phrased in
ISO/IEC Directives Part 2 terminology.  See Committee Draft 1
statistics above. The "conformity requirements essential to achieve
the interoperability" made mandatory by JTC 1 Directives remain
unspecified.

Indeed, the situation calls into question whether DIS ISO/IEC:26300
(OASIS ODF 1.0) or ISO/IEC:26300-2006 is the existing international
standard. OASIS ODF 1.0 was adopted as an international standard
without dissenting ballot. Although a ballot resolution meeting was
scheduled, it was canceled because of the lack of dissenting votes and
the final changes including those under discussion were approved by
ITTF rather than by NB ballot. See Alex Brown, The Maintenance of ODF
– an Aide-mémoire, Where Is There an End of It? (6 November 2008),
<http://www.adjb.net/post/The-Maintenance-of-ODF-e28093-an-Aide-memoire.aspx>.

A very similar issue arose in the processing of DIS ISO/IEC:29500
OOXML when India appealed the decision to adopt that specification as
an international standard, arguing that because the specification
resulting from the changes agreed to at the specification's ballot
resolution meeting was not published by the close of the balloting
period. The ISO Technical Management Board recommended rejection of
the appeal on that basis, stating:

"6. 'Final' text of ISO/IEC 29500 or ISO/IEC DIS 29500, or 'revised
FDIS text', not released

"6e. Correct but irrelevant. The decision being appealed is the JTC 1
decision to approve
the draft. The text mentioned in the Directives and by the appellants
is not germane to
that decision, *which must be taken on the basis of the original DIS
text and the actions
taken by the BRM on the comments.* The provision of any revised text is not for
purposes of further decision by NBs."

ISO Technical Management Board Secretariat, Appeals on ISO/IEC DIS
29500 Open Office XML, No. 078/2008 (4 July 2008), pg. 5,
<http://www.groklaw.net/pdf/ISOAppealRecommendationTMB.pdf>; see also
India appeal at ibid., pp. 13-14.

In the case of DIS ISO/IEC:26300 ODF, it was adopted as an
international standard without dissenting ballot and there was no BRM
convened for the NBs to act on any comments received along with
ballots. The NBs and not ITTF have the authority to adopt alterations
to DIS specifications as was recognized by the ISO TMB in the DIS
ISO/IEC:29500 case.

ITTF therefore appears to have acted outside its authority and in
disregard of JTC 1 Directives interoperability requirements in
approving the wholesale abandonment of mandatory interoperability
requirements contained in DIS ISO/IEC:26300, without the approval of
NBs and without the required consent of the Secretaries-General of ISO
and IEC.

Recommendation: Issues such as those discussed above should be mooted
by fully specifying the interoperability conformity requirements ---
including implementation behavior --- essential to achieve the
interoperability, with particular attention paid to existing options.
Per JTC 1 Directives, the number of options should be minimized and
the degree of complexity in the specification should be drastically
reduced through reduction of options.

OASIS ODF 1.0 relied far too heavily on the RFC 2119 definition of MAY
and OPTIONAL, requiring interoperability despite options without
specifying the methodology for achieving that state such as
identification and preservation of relevant markup. Such deficiencies
need to be rectified in ODF 1.2.

Doing so should not be unduly difficult. E.g., the mass of options in
the specification mask hard-coded SHALL and SHALL NOT programming
decisions in the market leading implementations. Disclosure of those
programming decisions and translation into SHALL and SHALL not
provisions in the specification would considerably advance toward
completion of the interoperability editing task.

Best regards,

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

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