Last update Tue, 11 Sep 2007 21:15:00 GMT
This page contains our
collected information and our
comments about Microsoft's proposal for ISO standard, ECMA-376 (OOXML)
with comparisons to ISO/IEC 26300 (ODF).
The information presented here is collected by
us
and it is our views which are expressed here.
That is,
this site is independent from
the <NO>OOXML campaign site.
For ongoing news regarding OOXML we
recommend that you visit the
<NO>OOXML site as
the page you are reading now will only be updated rarely. Only when
we get new significant facts,
such that Microsoft would have improved the specification draft,
which could possibly change our opinion somewhat about OOXML,
or for instance, that Microsoft would decide to drop OOXML.
Background:
The 27th of Aug 2007 there
was a big scandal at the
Swedish Standards Institute, SIS
when
23 new members, mostly Microsoft partners suddenly
appeared
at the voting at SIS for recommending ECMA-376
for fasttracking towards an ISO standard.
Here is a blog from
a member
of one these new companies, with some of ours (Roland Orre's) comments.
When 23 new members appear just before a decision about whether
a standard should be recommended or not, it certainly indicates a problem
with the review and voting procedure, but it also indicates
a serious problem on how Microsoft handles technical issues.
Microsoft
has also acknowledged that there was a problem and
there are
evidence that Microsoft has influenced the voting procedure.
Here is an
article from Wall Street Journal
which indicates that this hijacking of the voting procedure
has happened in many countries.
Anyway, this page is a result from a debate article in
the Swedish newspaper
Computer Sweden,
where Klas Hammar, business area manager of Microsoft Sweden
writes a
debate article in Computer Sweden the 7th of Sep 2007,
where he claims that OOXML is "future safe".
We certainly do not approve this claim and
replied on Klas Hammar's article
the same morning. When the signature "Johan Wallqvist" then replied to
our comments, claiming that we had just taken our arguments
from the <NO>OOXML site
as well as claiming that we had got them from IBM,
we decided to make an official statement about this through this
web page as we had collected information about OOXML on our own,
reaching the conclusion that we can not support OOXML.
There is also an article in DN
"Microsoft vägrar ge upp striden"
11th Sep 2007 (in Swedish)
where Klas Hammar says:
"One could ask the question why it shouldn't become
a standard".
Well, our view is that this question is
redundant when one has understood the issue.
It is a standard proposal which in its current shape locks people
to platform dependence. Our view is that documents should
be freely exchangable between any platforms, now and in the future.
Summary:
Listed below this summary are the technical comments and the
technical documentation which has been used as a reference
for our opinion upon OOXML in its current form, together
with the links from where this documentation can be fetched.
We have also put the standard documents here, both for ODF and
OOXML, as they are zipped otherwise.
-
As long as the flaws in the standard proposal
according the documents below are not corrected we
consider the standard to not have quality enough
to qualify as a proposed ISO standard.
-
As long as there exists no open source, platform independent,
reference implementation of the standard,
as is announced
but not yet available, we consider the OOXML format being an unsafe
lock in and should not be used for long term storage of documents.
-
We as a company only handles and develops platform independent formats
and we will not be able to handle OOXML in its current form.
-
Our view is that it would require a considerable effort and cost from Microsoft
to be able to fix the flaws in the current draft proposal.
-
We don't see the benefits having two standards solving exactly the same problem.
Therefore a recommendation from us would be that Microsoft joins
OASIS
and participates in developing the already existing ODF standard
towards a better ODF-2.0. This will benefit all users on all platforms.
Technical comments on OOXML:
These documents are collected here locally, but also linked to orig. site.
If you want some brief, straight to the point documents start
with Google's, Oracle's and Italy's comments. They are quick and
easy to read.
- Google's Position on OOXML as a Proposed ISO Standard (3 pages)
-
version 1: local,
orig site
version 2:
local,
orig site
- Oracle Comments on DIS 29500 (8 pages)
-
local,
orig site
- Open Document Format Alliance comments about OOXML (5 pages)
-
local,
orig site
- Spain comments on OOXML from FFII (12 pages)
-
html,
local ODF,
ODF orig site,
PDF orig site
- Singapore's comments on OOXML by Anand Vaidya (91 slides)
-
(this is a good overview of the most serious flaws with ECMA-376)
local ODF,
orig site Flash and ODF
- Italian (Associazione PLIO) comments about OOXML (6 pages)
-
local PDF,
orig site PDF
Standard definitions:
- The ISO/IEC 26300:2006 standard (OpenDocument v1.0 and v1.1)
-
-
First standard, ODF 1.0, ISO in May 2006: local ODF,
local PDF,
orig site ODF (722 pages)
-
Latest standard, ODF 1.1 ISO in Feb 2007: local ODF,
local PDF,
orig site ODF (738 pages)
- The ECMA-376 (Office Open XML) standard
-
- Part1 Fundamentals (178 pages PDF)
- Part2 Packaging_Conventions (131 pages PDF)
- Part3 Primer (473 pages PDF)
- Part4 Language_Reference (5220 pages PDF)
- Part5_Compatibility_and_Extensibility.pdf (43 pages PDF)
orig site ZIP archives
Other documents and links:
- A white paper on the differences between ODF and OOXML by Edward Macnagthen
-
local PDF,
orig site PDF,
orig site ODF,
link to background of white paper
- A study in comparing file sizes: "Why is OOXML Slow?"
-
link to html
- OOXML stalls at ISO
-
O'Reilly article
by
Nat Torkington)
- How a standard can kill a standard (OOXML versus ODF)
-
O'Reilly article by
Andy Oram
- The <NO>OOXML campaign site
- The <NO>OOXML campaign site
led by FFII (Foundation for Free Information Infrastructure)
-
A petition you can sign, organized by
FFII, to recommend that OOXML does not
become an ISO standard. With the eight most serious flaws of OOXML listed
in 39 languages.
Roland Orre
CEO, Research Director
NeuroLogic Sweden AB