[LINK] Noes from OOXML standards forum

[Glen Turner]

On Mon, 2007-08-20 at 17:55 +1000, Tom Worthington wrote:

> I do not believe that Standards Australia are biased in favour of 
> Microsoft. The standards people make standards in a vast range of 
> areas and can't be experts in the details. It is up to the experts on 
> the committees to advise on what is a good standard and what is not. 
> It does not good to shout abuse at the standards administrators about 
> the technical details of a topic they do not understand.

I have to disagree. 

The membership of IEEE or IETF or W3C would not tolerate an area chair
or a working group chair that could not come to grips with the technical
detail. In fact they are usually expected to be technical leaders in
their field in addition to having the management skills needed of a
committee chair.

So why is the JTC1 committee of ISO's Australia National Body
different?  Standards Australia are a Participating Member, and
this carries the implication that they should be able to perform
the development and evaluation of technical computing standards.

> Also keep in mind that international standards are not made by the 
> more open, free process used for Internet standards. ISO standards 
> are made by small closed committees of vested interests, copies of 
> the standards are sold for money and standards can use patented 
> technology where a licence fee is charged for use. In that context 
> the process for OOXML is relatively free and open.

The MoU with the Commonwealth also requires Standards Australia
to consider the national interest.

ISO wasn't always a cozy little club, as you suggest. The development
and introduction of metric measurements wasn't beneficial to vested
interests, but served the wider community.

Some of us hold ISO up to the principles of its past, not to its
behaviour of its present.  That is because we actually want to *use*
the standards.

> You can't blame SA staff for following their set down procedures. If 
> you don't like the way such standards are made, then what is needed 
> is for those procedures to be changed, or for other bodies with 
> different procedures to be used for making standards. The easiest way 
> I can see to do this is the same "fast track" process being used for 
> OOXML. With this some other body prepares a draft using their own 
> process and then puts it up, completed, to be a standard. That body 
> can use a freer process and make its draft freely available. SA, ISO 
> and other standards bodies can then endorse it officially and sell 
> their official version of the standard. Most people will use the free 
> unofficial version of the standard, a few governments and large 
> companies will buy enough copies of the official standards to keep 
> the official standards bodies going and so everyone will be happy. At 
> least that is what I suggested to Standards Australia, last week 
> <http://www.tomw.net.au/blog/2007/08/australian-innovation-through-standards.html>.

Have you *read* the OOXML specification. I've worked on three
IEEE and IETF standards committees and not even our first
drafts were this poor.  But the Fast Track process has no way
to prevent such a poor specification from becoming an
international standard.

The Fast Track process should be abolished. It only benefits
one organisation -- ECMA International -- and that organisation's
only purpose is to sell access to the ISO Fast Track.  Go have
a read of the ECMA website -- they are quite blatant about it.

The Fast Track is bad for consumers.  Both DVD-RW and DVD+RW
became standards via the Fast Track from ECMA.  Having two
incompatible writable DVD standards was always going to be
bad for consumers, but ECMA doesn't give a stuff about that.

There is an alternative process -- the Publicly Available
Specification (PAS) process.  But even that let through
a deficient specification for OpenDocument spreadsheet
formulas and didn't evaluate OpenDocument for support for
the disabled.

So it has been plain for some time that ISO's procedures are
deficient. I would have expected Standards Australia to be
addressing those deficiencies themselves.

This current farce sets ISO up for irrelevancy.  If any
vendor can make a specification an "international standard"
but still prevent other vendors from making use of it via
patents then what is the benefit of using a ISO standard?

Some other specifications body -- approved by the
international standards community or not -- will take on the
role of developing specifications that people actually use
when they want interoperability.  The ISO has been here before ---
with the IETF and OSI. It is walking down the same road again
with IT specifications and the free software movement.

Microsoft will do want it pleases -- standards or not. Just
as IBM once did with data communications (anyone remember the
increasingly frantic modifications to SNA to keep pace with
the functions provided by the Internet specifications?)
It is not even plain that if a ISO Ballot Resolution Meeting
alters OOXML that Microsoft will implement the altered
specification in Office.  Again, this occured with IBM
data communications -- IBM continued to run SDLC rather
than the ISO's HDLC which was ISO's modication to the
SDLC specification.

In the free software movement ISO has a group of people that
use standards as the key to interoperability between groups
of disparate developers. If ISO have a good standard in the
field then that standard will be used by free software
developers.  ISO are currently doing their level best to
ensure that the free software community invents its own
standards body.

Cheers, Glen

PS: My completed submission to Standards Australia on OOXML
    is available from
<http://www.gdt.id.au/~gdt/presentations/2007-08-07-standardsaustralia-dis29500/>