Message sent to the Standards Board as referenced above:
(1) It was evident from the "Tagung Fuer Informatik und Recht" held in
Bern (24.10.06) that there are quite strong currents towards Open Source
and general openness of information and its retention. The arguments
were pragmatic, driven by interoperability needs. However, the legal
profession has another trait which has not often been addressed: it does
NOT like change. Hence, forcing the legal profession to change how it
works every 2..3 years by releasing a new version of MS Office does not
compare to the relatively stable "learn it once" interface that
OpenOffice.org offers - not to mention that OpenOffice.org is free and
less prone to virus infections.
(2) Archiving will be at risk with OOXML. It does NOT offer full
specification, as it includes referral to external specifications which
are not publicly available (and could thus be rewritten later without
affecting the "Standards" status). OOXML has been independently produced
by a formally convicted monopolist who even now has EU punishments
pending for failing to comply with court demands for openness ("SMB
specifications"). One could call this a deficient track record.
ODF, by its nature, allows the use of open and/or alternative software
that can be inspected for information leaks. The events at Crypto AG
<http://biphome.spray.se/laszlob/cryptoag/crypto_ag.htm> show that such
inspection may indeed be desirable for government and legal use.
Indeed, in more secure environments the use of OpenOffice.org software
has increased just because of that inspection ability, and its
aforementioned lower sensitivity to virus infections.
(3) OOXML is a specification by one company only. ODF is defined by a
consortium and put in practical use by Open Software communities - and
accepted as EU standard. ODF is also the default file format in most
charitable offerings such as the One Laptop Per Child (OLPC) project and
the school project in the Spanish Extremadura region. In other words,
it has had real world exposure, use and correction. OOXML is too new to
be trusted, and certainly too new to be then voted as a formal standard
with wide ranging economic and social impact. OOXML could only ever be
considered viable if it had as many years 'in the field' as ODF and
could demonstrate cross-vendor interoperability.
(4) OOXML has extreme deficiencies in the way it handles mathematics
(see <http://www.robweir.com/blog/2007/07/formula-for-failure.html> for
a sample analysis). For a nation whose technology exports are renowned
for world leading precision this should raise questions. Flaws like
this have the potential to lead to economic damage even more than the
lock-in to one provider, and may only have emerged as a result of the
pressure for openness - we will never know if these are new problems or
are exposures of existing ones.
I would recommend to address the Swiss vote with care. The extreme
lobbying is an indication that the arguments for and against a specific
choice have to be very clear to avoid the impression of supplier induced
bias. The focus has to be Swiss interests, none other. As someone who
has worked with automated, policy and secure environments I know at both
a tactical and strategic level what works, and what causes (usually
costly) problems long term.
No decision should ever be taken by a vendor.
