Høringsvar til Dansk Standard vedr. ISO og OOXML

[Martin Schlander <sslug@sslug>]

Dansk Standard er blandt de ca. 30 nationale standardiseringsorganer der
skal afgøre hvorvidt OOXML skal ISO-certificeres. De kører høring frem til
2/7.

Høringssvar skal afleveres på engelsk i en skabelon i et *.doc. 

Jeg har skrevet et udkast som nok er lidt af en tynd kop te, men hvis nogen
vil læse 6000 sider og komme med sidehenvisninger og konkrete
ændringsforslag skal de være meget velkomne.

-----------

We believe OOXML should not be ISO-certified at all. There's already an
ISO-standard for document formats. Having two different standards with huge
overlaps goes against the whole idea of standardization. Doing so could
damage the credibility of ISO as well as Dansk Standard assuming they voted
in favour of it.

Of course the feature set of the formats are not completely identical, but
it would be better to expand existing standards than to create new ones
everytime someone needs some feature.

ISO is a very respected and influential organization. Having two
ISO-standards for office documtents would effectively mean that all
developers of office software would have to implement both standards to
compete in the market place. This would lead to a big waste of resources,
bad for businesses and more expensive and complex for customers than having
just one standard. It would also be imbalanced and favour large vendors.
Having two different ISO-standards would most likely lead to more
interoperability problems than having only one.

The market share of Microsoft Office is not a viable argument either. At
least two gratis plugins that provide ODF-support exist already ? and
nothing is stopping Microsoft from implementing native support for the
existing ISO standard in the field.

Besides the above points of why two duplicate standards is a bad idea, there
are a number of reasons why OOXML is a bad format in itself. If an ISO
certification should happen at least these problems should be adressed.

- The sheer size of the specification is a problem. This complexity means
it'll be very difficult to implement properly ? perhaps impossible. And
again it will favour large vendors over smaller ones.

- The specification has numerous undocumented sections which is well known.
Apparently for reasons of backwards compatibilty. We must insist that these
sections be fully documented.

- OOXML specifies it's own standards instead of re-using existing standards,
for example wrt. handling calendaring, dates and Mathematical Markup
Language (MathML). Leading to unnecessary complexity.

- The best argument for it seems to be the promised backwards compatibility
with existing binary Microsoft document formats, but even this  backwards
compatibility is questionable.

-----------

Hvis man er interesseret i at se formularen som den ser ud pt. kan den
findes her:
http://suse.linuxin.dk/sslug-iso-svar.doc

Her er en howto til at udfylde formularen, jeg synes ikke det 100%
selvindlysende:
http://suse.linuxin.dk/Template_for_comments_and_secretariat_observations_-_User_Guide.pdf

Her er DS' oplæg til høringen:
http://www.ds.dk/3654

Happy feedbackin'


-- 
openSUSE begynderguide: http://suse.linuxin.dk