Friday, January 12, 2007

Microsoft EOOXML Hits the ISO Contradiction Wall

EOOXML -- What is a 'contradiction' at ISO and what are its procedures?



Here we are just six days into the ISO Contradiction Review Phase, and the legendary GrokLaw legal expert Marbux comes stampeding across the windswept steppes of standardization, across endless arrays of routers pulsing with the soft wave dance of light streaming through fiber, and hordes of angry netizens shouting for freedom behind him.

So what has so riled the keepers of the Open Internet? Well, they sit at the threshold of a new age, the age of collaborative computing, where truly open and unencumbered standards, open source community participation, and the open Internet meet. And they see dark clouds blowing in hard from Redmond. They see a threat to their beloved universal XML file format, OpenDocument.

The issue here is the ISO Contradiction Review Phase for consideration of Microsoft's EOOXML file format specification now under way - with twenty four days left on the review calendar. The clock started ticking on January 5th, 2007, and ends on February 5th, 2007. During this short review period, NB's (National Standards Body Members) of the ISO JTSC1 must review the 6,000 plus page EOOXML specification and submit their contradiction concerns.

It's actually much more than 6,000 pages. Microsoft was kind enough to pad their submission with volumes of highly persuasive, professionally cut presentation and marketing materials. All of which the legendary Marbux covers in another post; Microsoft/Ecma's submissions to ISO for Ecma Office Open XML

With these two posts, the magnificent Marbux provides everything an NB needs to quickly work their way through the muck and mire of EOOXML. And he gives the rest of the world all we need to provision our National representatives with contradiction comments and insights.

Thanks Marbux. This is serious business. The future of mankind's digital civilization is at stake. And you've perhaps saved the Open Internet with this prescient intervention. And not a moment too soon!

The MNC ODF-EOOXML Translator Project: What's not to like? And what does this have to do with the ISO Contradiction Phase?

In spite of the mighty Marbux's great work, i'm still confused about the relationship between the woefully obvious problems Microsoft anticipates with the ISO Contradiction Review Phase, and the work being done at the MS-Novel-CleverAge Translator Project?

How come our friends in Redmond think the MNC Translator Project is going to pull them through the rough waters of the Contradiction Review Phase?

I don't get it. If the MNC Translator works (which it doesn't) then it proves EOOXML is a duplicate of ODf, and therefore in contradiction with an existing ISO product. If, as is the case, the Translator Project bombs, then that would prove EOOXML to be a rather poor and inconsistent subset implementation of ODF. Worse than a duplicate contradiction.

We know that Novel fully intends on implementing the MNC EOOXML/ODF Translator as a plugin for OpenOffice. The first iteration for OOo Writer is promised for release any day now. What a challenge. Except that the Translator plugin transformation process is crap. Translator tries to do an XSLT conversion of ODF to EOOXML. Since EOOXML is non compliant with XSLT, with presentation styles damn defiant of the structural integrity XPath requires, this is near impossible to do with any halfway acceptable measure of fidelity. It's so bad that the poor folks at the Microsoft sponsored CleverAge are resorting to C# routines to get the MSOffice version of the translator plugin working.

Here's the thing: Even though EOOXML and ODF are designed to do the same thing, the interoperability and quality of transformation between the two is beyond awful. It's unacceptable. Not only does EOOXML contradict ODF, the inconsistencies it brings to the table are embarrassing to ISO.

Microsoft will of course argue that ODF was designed for OpenOffice features and is inadequate to meet the demands of the more advanced and feature loaded MSOffice (pull the string on your Alan Yates interoperability by design doll and this incessant apply directly to your forehead message will replay itself. Beware, once triggered the Yates nonsense can not be turned off: Microsoft’s Alan Yates about ODF).

This is of course not true. The Yates interoperability by design doll does this drive by shooting of ODF and then adroitly shifts to proof of this file format inadequacy claims by comparing OpenOffice features to MSOffice features. Wait a minute, isn't he supposed to be comparing ODF to EOOXML? Or proving that ODF can't handle everything the legacy of MSOffice operations has to throw at it?

Yes he's right about one thing in his slight of hand prestidigitations; the feature sets of OOo and MSOffice are not the same. But what does that have to do with ODF and MSOffice?

The truth is that ODF can handle everything OpenOffice throws at it. And, it can handle everything MSOffice throws at it. Everything! Can you say "universal"?

The proof of this can be found with any ODF plugin running natively within MSOffice. The OpenDocument Foundation's daVinci plugin proves my statement. Within OpenOffice, ODF works perfectly. Within MSOffice, ODF works perfectly. Including the legacy binaries. And on all MSOffice versions from 1997 to 2007. No problemo.

We have yet to see how perfectly the Novel Translator Project Plugin works inside OpenOffice, (yuk yuk), but that would be proof of just how well EOOXML handles whatever the feature rich OpenOffice can throw at it. From what i've heard, the results are dismal, bordering on a 65% fidelity rate. (OpenOffice conversion filters routinely hit a magnificent 85% fidelity rate).

We'll assume for the sake for argument that EOOXML can handle MSOffice and the legacy binaries because the original EOOXML charter specifically stated that as the sole objective of the effort. But so can ODF! When running natively in MSOffice, ODF hits a 100% fidelity rate with all of the legacy binaries and evolving proprietary XML InfoSet binaries.

Maybe i'm just having a bad day, but it seems to me that the MNC Translator Project is of no help to Microsoft as they try to shuck, shuffle and jive their way through the ISO Contradiction Review gauntlet. If MNC Translator works, it's a clear clean contradiction. If the MNC Translator fails, it is further proof of an inept contradiction. Persoanlly, i would rather fail the ISO Contradiction Review based on a clear clean contradiction. The alternative is embarrassing. Both to ISO and, Microsoft - ECMA.

Some days you get the bear. And some days the bear gets you. Munch munch Microsoft.

~ge~

  • - post by garyedwards
  • Links to ISO JTSC1 National Body Representatives

    The voting "P" member national bodies of JTC1 are listed here.


The national bodies' contradictions must be received by the Secretariat of JTC1 by the deadline on February 5, 2007. The current secretariat is held by Mrs. Lisa Rajchel of ANSI. Her contact information is here.

Full reference to the documents Microsoft submitted to ISO can be found here, again with many thanks to the legendary Marbux :Microsoft/Ecma's submissions to ISO for Ecma Office Open XML

And this blurb on the MNC EOOXML/ODF Translator Project: Tectonic: Novell to boost document interoperability

"As far as anyone" can tell, Novell has yet to make any contributions to the MS-CleverAge Translator Project. They simply mounted the existing translator work as a Novel OpenOffice plugin..."

Post a Comment