SIUG
Swiss Internet User Group
Postfach 1908
8021 Zürich

To:
SNV Schweizerische Normen-Vereinigung
Fax 052 224 54 74

Zürich, August 14, 2007

Dear Sir,

SIUG would like to hereby appeal against a set of decisions made at
and in the context of the Comment Resolution Meeting for the
proposed OOXML standard (DIS 29500) [1].


Summary: NK 149 chairman Mr Sebestyen has a conflict of interests
         between the business interests which he represents as
         Secretary-General of Ecma and his duties as NK 149 chairman.  
         He has abused his authority as NK 149 chairman to influence
         the work of UK 14 in a way which has resulted in improper
         dismissal of SIUG's comment expressing the serious concern
         that accepting OOXML as an international standard would go
         against ISO's fundamental objectives and it would also be a
         violation of Swiss and international law, unless OOXML is
         first revised to make it map ODF (in the sense defined below).

         SIUG asks that if it is found that SIUG's concerns are
         justified at least in part, SNV should include the concerned
         comment in the official Swiss comments document, and
         consequently the Swiss vote on DIS 29500 should be "DISAPPROVE
         with comments" until OOXML has been revised to make it map
         ODF.  SIUG suggests that the voting process which the UK 14
         chairman has started within UK 14 may be allowed to continue
         to completion, and that it should be used to determine what
         the Swiss vote should be changed to if/when OOXML has been
         revised to make it map ODF.


1. What happened

1.1 SIUG has submitted, in the proper manner, before the deadline, the
following comment and proposed resolution as a "technical comment",
in reference to "Part 1, introduction (page xii) and entire document":

Comment:

  It is not acceptable for an international standard to be designed
  primarily around the goal of compatibility with a particular
  company's products.  This is particularly inappropriate where, as in
  the present case, compatibility with an existing international
  standard is neglected in favor of the one-sided goal of maximal
  compatibility with document file formats introduced by one company,
  and where the proposed standard does not provide equal opportunities
  for compatibility to that company's present and future competitors.
  Unless this shortcoming of DIS 29500 is fixed, accepting this
  specification as a national or international standard would be a
  violation of Swiss and international law.

Proposed resolution:

  Change the goal from being "fully compatible" with "existing
  investments in Microsoft Office documents" to seeking to attain the
  same high level of compatibility not only with Microsoft's formats,
  but also with the international norm ISO/IEC 26300 (OpenDocument).
  Review the entire draft standard and modify it corresponding to this
  revised goal.

1.2 This comment statement and proposed resolution are together
referred to in this document as "the concerned comment".

1.3 The UK 14 chairman received SIUG's comments and sent an email
response in which he praised SIUG's comments as having been prepared
in an exemplary manner.  (Exact wording in German: "Besten Dank für
Ihren vorbildlich ausgearbeiteten Kommentar.")

1.4 The UK 14 chairman included SIUG's comments in the invitation for
the OOXML Comment Resolution Meeting.

1.5 SIUG informed the UK 14 chairman informally that SIUG's main point
of concern is the demand for fairness.  The UK 14 chairman replied
indicating that he is willing to support such a demand provided that
it can be based on the JTC1 directives and that it is realistic.
(Exact wording in German: "Ich bin für alles zu haben, was sich auf
die JTC1-Direktiven abstützen lässt und realistisch ist.")

1.6 In preparation for the OOXML Comment Resolution Meeting, SIUG
therefore prepared the necessary arguments and documentation for
proving that the concerned comment and its demand for fairness can be
justified based on the JTC1 directives and that it is realistic.  Of
course, the justification for applying the fundamental principle of
economic fairness in regard to standardization (that standardization
must not be carried out in a manner which is designed to move a
near-monopoly market state even closer to an absolute monopoly) to a
concrete situation necessarily involves the economic argument of
establishing that a near-monopoly market state in fact exists.  In
the same way, applying the relevant principles of international law
(specifically GPA, the WTO Agreement on Government Procurement, of
which Switzerland is a member state) and Swiss law (specifically,
the implementation of the GTA in national law as well as the general
principle of competition law that forbids taking advantage of a
market-dominating position) to a concrete situation must necessarily
involve the same economic argument.  SIUG therefore included in the
prepared argumentation a justification based on JTC1 directives that
this kind of economic consideration is appropriate during the
evaluation of a proposed international standard.

1.7 Unfortunately, on the day before the Comment Resolution Meeting,
the NK 149 chairman sent an email message to the UK 14 members and
UK 14 chairman forbidding UK 14 to discuss any economic arguments,
forbidding explicitly the type of topic that is necessary to justify
or even discuss the validity of the concerned comment.  In this email,
the NK 149 chairman explicitly referred to his authority as chairman
of NK 149, the parent committee of UK, so that it was clear that
this message represented not just his personal opinion as a member of
UK 14, but a decision which would be binding for UK 14 (except, of
course, if SNV makes the decision to overrule him in this regard).
This email is archived here:
http://www.snvlivelink.ch/ll12/livelink?func=ll&objId=1284104


1.8 During the Comment Resolution Meeting, the UK 14 chairman applied
the most unusually demandingly strict standard of proof to already
the first sentence of the concerned comment, which was discussed in
isolation.  Nevertheless, SIUG was given the opportunity to present a
few points of the argument, including the argument which demonstrated
from the JTC1 directive that economic consideration is appropriate
during the evaluation of a proposed international standard.  However
when the point in the argument came where it is necessary to refer to
the reality of a near-monopoly market situation, SIUG's representative
was interrupted in the middle of the first sentence on this topic.
That category of arguments was disallowed, without any justification
besides "you're making too many assumptions" and "I don't allow it".
In view of the UK 14 chairman's earlier promise 1.5 this can only be
explained by the NK 149 chairman's intervention (see 1.7).  SIUG
indicated the intention to appeal against this decision to silence an
essential argument.  Following this episode, Mr Sebestyen (without
clarifying whether he was speaking as NK 149 chairman or as Ecma
representative) made a point about ISO/IEC practice, and from there
onwards the bar for establishing any fundamental concerns about the
proposed standard was raised even higher, to having to show not only a
violation of fundamental principles based on the directives without
any reference to the near-monopoly market situation, but to having to
show that in addition that the situation is significantly different
from previous ISO/IEC practice.  The UK 14 chairman then rejected the
concerned comment for having been "not proved" based on JTC1
directives (such proof was of course impossible because the main
arguments were disallowed), and tried to get SIUG's agreement to
replacing the comment with an editorial comment which is irrelevant to
SIUG's concern.  SIUG did not agree to that.  The UK 14 chairman
completely avoided bringing to discussion the part of the comment text
which refers to national and international law.

1.9 At the end of the Comment Resolution Meeting, the UK 14 chairman
concluded that in his opinion the comments document that resulted
from the meeting does not justify a "DISAPPROVE with comments" vote,
and he emphatically presented his voting recommendation of "APPROVE
with comments" and he initiated a voting process aimed at deciding the
Swiss vote between the options "APPROVE with comments" and "ABSTAIN
with comments".  Event though the majority of the UK 14 members in the
room wanted "DISAPPROVE with comments", the voting process presented
by the UK 14 chairman does not provide for any way in which the
"DISAPPROVE with comments" outcome desired by the majority could be
reached.


2. Definition of when a document format maps another document format

2.1 Definition: Document format X is said to map document format Y if
and only if every valid format Y document can be represented in format
X so that there is no loss of information and so that it can be
guaranteed by a software which implements just format X (and no
compatibility extensions not defined in the format X specification)
that printing out the document from its format X representation gives
a rendering of the original document which is semantically correct
from the perspective of human readers.

2.2 The statement in the OOXML specification itself about its
purpose, which is there phrased as "protecting" an "investment in
Microsoft Office documents" (see DIS 29500 [1], Part 1, page xii)
can be interpreted as implying that it is a key goal of OOXML that
OOXML should map the legacy formats of "Microsoft Office documents"
in the sense of the above definition.


3. The concerned comment is properly labelled "technical comment"

3.1 There are three categories of comments:  Editorial, technical and
general:

3.1.1 Editorial comments are those that can be resolved by means of
one or more changes to the wording of the proposed standard without
changing the technical content of the standard.

3.1.2 Technical comments are those that can be resolved by means of
one or more changes to the technical content of the standard.

3.1.3 General comments are those which can not be resolved by means
of any change to the proposed standard.  (Consequently, general
comments are generally ignored in the ISO/IEC process which has
authority to make changes to a proposed standard but no authority
to apply changes to the rest of the world.)

3.2 The reason why it is proper to label the concerned comment as a
"technical comment" even though it appeals to general and legal
principles is that the way to resolve this comment is by means of
technical changes to the OOXML specification.


4. Justification on the basis of the ISO/IEC JTC1 Directives

4.1 A primary goal of standardization is to facilitate
interoperability between the products of different, competing
vendors.

4.2 What the ISO/IEC directives say: See ISO/IEC JTC1 Directives [2],
section 1.2 and annex I. In annex I, the directives state: "JTC 1
recognizes that interoperability is a major user requirement which can
be facilitated by standardization. Accordingly JTC 1 accepts the
responsibility to identify the key interfaces and produce the key IT
standards at those interfaces (including the relevant content
standards, e.g. ODA, SGML, CGM) to facilitate practical, time and
cost-effective interoperability, consistent with market requirements
and current technologies." The common understanding of the term
"market" is that there should be several vendors in competition
with each other, and in this context "market requirements" refers to
what these vendors need to provide in order to be considered by their
potential customers as a vendor capable of meeting their needs.

4.3 Application to the evaluation of OOXML: A specification which
aims only at making it possible to represent one company's old data
formats in the company's new data format could be described as
providing a kind of "intraoperability" within among that vendor's
products, but not interoperability between the products of different,
competing vendors. In addition, the directive says that the
interoperability provided by the JTC 1 process must be "consistent"
with "current technologies". This is plural and therefore cannot
reasonably be interpreted as referring only to one company's set of
legacy technology at the exclusion of interoperability with ODF which
is the ISO-standardized non-proprietary technology embraced by
essentially all competitors of Microsoft.

4.4 The key economic argument for demanding that OOXML should map
ODF:  In mixed environments with word processing software from several
different vendors which use different native document formats, due to
Microsoft's dominant market position, the common format used for
document interchange is likely to be OOXML regardless of the
technical merits of the standard.  If OOXML is not designed to be
capable of mapping at least OpenDocument Format (ODF), the ISO
standard format that is used as native document format by the leading
competitors to the "Microsoft Word" product, that creates a strong
disincentive against using the word processor software of Microsoft's
competitors in any technologically mixed environment.  This would
have the effect of moving the market situation for office document
software, which is already in a near-monopoly state, even closer to
an absolute monopoly, clearly a highly undesirable outcome both from
the perspective of competitors and from the perspective of customers.
The ISO/IEC JTC1 Directives [2], section 11.1.2 say that when any
action or inaction of JTC 1 or a subcommittee is "not in the best
interests of international trade and commerce", that is grounds for
appeal against that decision.  Certainly, that implies that a
"DISAPPROVE WITH COMMENTS" vote and a corresponding comment are fully
appropriate when a draft standard is not in the best interests of
international trade and commerce.

4.5 OOXML survived the contradiction phase of the ISO/IEC JTC1
process by means of the argument that the purpose of OOXML is
different from that of ODF in that the purpose of OOXML is not only
to provide a document format but also to map existing formats. (See
the "Response Document for Fast Track Ballot of ISO/IEC DIS 29500
(Ecma-376)" [3], section 2.1.)  By contrast, the purpose of ODF is to
simply provide a document format. In addition to the economic
argument of 4.4 this gives an additional justification for demanding
that OOXML should map ODF, although no requirement is imposed on
ODF that it must map anything besides itself.

4.6 The situation with ODF and OOXML is fundamentally different
from the situation with other families of ISO standards with
overlapping purposes which don't map each other, because those
examples don't create the same kind of interoperability problems.  For
example, it is not a big problem when there is a variety of image
formats or a variety of wireless LAN protocols because it is possible
for commercial products to simply support a variety of such standards
without thereby creating problems for end users.  By contrast, when
there are several international standards for document formats, it is
necessary for achieving the interoperability goal that at least one of
these standardized document formats must map all other document format
standards, otherwise it becomes impossible to reliably compose a
document from components which are each in a standard document format.

5. Realistic feasibility of the proposed resolution

The German standardization organization DIN is working on a detailed
study about which elements of ODF can be represented in OOXML, and
how (and vice versa).  Based on the results of DIN's work it will
certainly be realistically feasible to determine precisely which
semantics that can be expressed by ODF cannot be precisely expressed
by the current version of OOXML.  Since OOXML is an inherently
extensible, XML-based format, there cannot be any serious problem with
adding the required features to the OOXML specification.


6. Dismissal of the concerned comment was improper

6.1 The dismissal of the concerned comment was improper because it
violated a promise given previously, see 1.5

6.2 The dismissal of the concerned comment was improper because the
decision to disallow economic arguments was in violation of the SNV
directives which state explicitly that economic aspects should be
taken in consideration.  (Exact wording in German: "Bei der
Normungsarbeit sollen wirtschaftliche Gesichtspunkte berücksichtigt
werden." SNV Richtlinien zur Normungsarbeit [4], V4.2, Punkt 4.1)

6.3 The dismissal of the concerned comment was improper because if
this manner of dismissing monopoly-related concerns were acceptable,
there would be nothing to prevent the standardization process from
being abused for monopolistic purposes.  Such exploitation of a
dominant market position is illegal under competition law, and it
would be highly inappropriate (and probably also illegal under
anti-trust law) if SNV and/or ISO would allow themselves to blindly
go along with illegal "standards war" anti-competitive practices.

6.4 The relevant document of international law is the WTO Agreement 
on Government Procurement [5] (to which Switzerland is a member
country) which is applicable since the ISO standardization of OOXML is
pursued precisely in view of government procurement.  This treaty
states: "Technical specifications laying down the characteristics of
products or services to be procured [...] shall not be prepared,
adopted or applied with a view to, or with the effect of, creating
unnecessary obstacles to international trade." (Article VI, paragraph 1.)
Preparing and/or adopting the OOXML specification in view of
government procurement is a violation of this treaty unless the
unnecessary obstacles to international trade are avoided by means of
amending the specification so that it maps not only Microsoft's legacy
formats but also ODF.

6.5 That unless revised to make it map ODF, the OOXML specification 
has the effect of creating unnecessary obstacles to international
trade is seen as follows:  That there is the effect of creating
obstacles to international trade follows from the argument of 4.4,
(since everything that significantly disincentivates procurement of any
products which compete with the market leader is an obstacle to trade
in such competing products) together with the observation that most
countries don't produce their own office software suites, i.e. office
suite software is traded by means of international trade.  That these
obstacles to international trade are unnecessary follows from 5.
above.

6.6 The combination of 6.4 and 6.5 shows that the dismissal of the
concerned comment was not only improper but even illegal.


7. Mr Sebestyen's conflict of interests

7.1 As NK 149 chairman, Mr Sebestyen has very significant influence
on the UK 14 process, both by his power to appoint the UK 14 chairman
and also by direct interventions as evidenced in 1.7 and 1.8.

7.2 It should not be assumed without evidence that Ecma's interests
are directly aligned with those of ISO and SNV.  In fact, there
is strong evidence to the contrary, as follows:

7.2.1 At the UK 14 launch meeting, Mr Sebestyen said, in the context
of introducing himself as Ecma's representative, "Of course we want
that Open XML gets approved".  This unconditional desire for approval
without consideration of third-party concerns is very different from
the interests of ISO and SNV.

7.2.2 In Ecma's marketing materials, Ecma offers to its customers
among other benefits the following: "Bonus: International ISO/IEC
standard via 6-month Fast-track procedure".  See
http://www.ecma-international.org/activities/General/presentingecma.pdf

7.2.3 During informal discussions within UK 14, Mr Sebestyen has
strongly promoted the "RAND standards are the norm now" agenda which
reflects Ecma's patents policy, but which is very different from the
much more cautious attitudes towards patents expressed in the
directives of SNV, ISO and IEC.


8. Conclusion: What should happen now

8.1 If it is found that SIUG's concerns are justified at least in
part, SNV should include the concerned comment in the official Swiss
comments document, and consequently the Swiss vote on DIS 29500 should
be "DISAPPROVE with comments" until OOXML has been revised to make
it map ODF.

8.2 The UK 14 members expect to be able to participate in the vote
which has been started by the UK 14 chairman, and they expect the
result of the vote to be used for something.  SIUG suggests that the
voting process which the UK 14 chairman has started within UK 14 may
be allowed to continue to completion, and that it should be used to
determine what the Swiss vote should be changed to if/when OOXML
has been revised to make it map ODF.  This is in fact the proper
interpretation of that vote since it has been started and is being
carried out under the assumption of the concerned comment being
excluded from consideration.


9. References

[1] http://www.ecma-international.org/publications/standards/Ecma-376.htm

[2] ISO IEC JTC 1 Directives 5th edition Version 3
    http://www.snvlivelink.ch/ll12/livelink/1261374/
    ISO_IEC_JTC_1_Directives_5th_editon_Version_3.pdf?func=doc.Fetch
    &nodeid=1261374&viewType=1

[3] Response Document for Fast Track Ballot of ISO/IEC DIS 29500 (Ecma-376)
    http://www.ecma-international.org/news/TC45_current_work/Ecma%20responses.pdf

[4] SNV Richtlinien zur Normungsarbeit, Ausgabe 2007
    http://www.snvlivelink.ch/ll12/livelink/1244540/
    SNV_2007%2D06%2D11_Richtlinien_zur_Normungsarbeit_V4%2E2_.pdf
    ?func=doc.Fetch&nodeid=1244540&viewType=1

[5] WTO Agreement on Government Procurement
    http://www.wto.org/English/docs_e/legal_e/legal_e.htm#procurement


Yours sincerely





Norbert Bollow
President, SIUG (Swiss Internet User Group)