Microsoft Admits Trouble for Its Formats

I missed this a few days ago...

"Legitimate concerns' raised over Microsoft's Office formats" Matthew Aslett | Computer Business Review Online (7 Mar 2007).

The level of criticism targeted at Microsoft's XML-based office productivity file formats is significant, raising the potential that Open XML might not gain ISO approval...

Jason Matusow is the only one capable of a true statement, yet he still dissembles with sophistry and disinformation regarding Microsoft's intent behind using an open source project to develop its ODF translator...

Matusow said the company had chosen to contract third parties and an open source project to create its ODF translator to avoid suggestions that Microsoft was manipulating the standard. "The decision to do the translator was a matter of transparency," he said. "You want the transparency as well as the discipline of professional development."

Nonsense! Don't be silly, Jason.

The contrary is true: if you are Microsoft and you intend to deliver very bad document interoperability (which is, in fact, what we have), you need a 3rd party to blame for the poor performance. If Microsoft did the translator itself, the only passable performance would be perfect interoperability since Microsoft is the only body that knows the structure of its legacy binary document formats.

If you can't say something true, don't say anything at all.

Comments

[Jason Matusow]

Hi Sam - sorry to see you again apply the sophist title to me. I try to be as direct as possible in how I address things.

I disagree with your assessment of the reasoning behind the open source translator project we have funded. If we had built an internal, black-box, implmentation of the translator you would have been all over us for a lack of transparency. There are real technical differences between the products that use these formats, and between the formats themselves. It is critical that anyone can review the translation techniques, and assumptions made etc. so that they may have confidence in the translator technologies. Customers, on the other hand, have told us directly that they don't trust the testing quality of OSS projects like this. Thus, the financing of both general quality assurance, and specific complex document testing scenarios. If you are not satisfied with the results of the current translator, it is an OSS project. I encourage you to come join in the fun and make it better for everyone.

As for perfect translation - it is a goal for everyone. Yet, the technical reality behind all translation is that there are trade-offs in the assumptions you make, and in the differences in the features of the apps over time. The product capababilities get reflected in the format. That is why, if I am not mistaken, you see a difference in implementation of ODF in IBM Workplace vs. OpenOffice.

Translation is the key to a multi-format world. Thus, the announcements of translation between the Google app offerings and ODF. Bridging the gap to PDF. Working with legacy docs, and facilitating the move to future XML-based formats.

No matter what - an interesting time to be involved in the industry. Keep up the commentary, and be constructive in your criticsm.

Thanks -
Jason Matusow
Microsoft

[Sam]

Jason-

Thank you for your constructive response. It's better than I deserve in this case. I'll grant you that you would have been given a hard time no matter what...and your points about complexity and assumptions. Also true of IBM, OOo, et al...and we have our issues to work through on universality.

My issue is that we could have told you that XSLT would give poor translation, knowing up front that MOOXML keeps legacy data in binary form.

If you can out-transform the Foundation's Plugins, I'll cut you some slack. Even better, use ODF and I'll buy you a drink.

[zridling]

My question to Jason would be: Why will Microsoft not respond to the contradictions found in OXML? Who better to do so? Microsoft is its own best defense, but yet chooses silence? What does that say for OXML?

Also, I'd suggest that Jason can't have it both ways. You want OXML to be an "open standard" but then you write: "If we had built an internal, black-box, impl[e]mentation of the translator you would have been all over us for a lack of transparency."

That's a red herring. Jason, you seem to be saying that Microsoft 'could' build a perfect translation between the two formats but chose not to because you might have had to share the results publicly?

I'm confused. But I do know that OXML was one of Microsoft's two biggest mistakes. Microsoft could stem the tsunami of growing ODF adoption around the world if it adopted ODF as its native file format or if they went ahead and built that 'perfect' translator themselves. Instead, they will someday see that OXML is a dead-end street.

[Wesley Parish]

FWLIW, I have suggested to Bob Sutor that IBM - Microsoft's bete noir in this hunt - challenge Microsoft to come up with a working MS OO XML plugin for WASCE.

Well, Jason, you've just proved you read this blog - perhaps you could suggest it to the relevant people in the Microsoft Office division? I mean it seriously - there is no better way to get the bugs beaten out of a technical standard than to stretch it over a different framework - which is why I trust ODF, after all - it's now on KOffice as well as WASCE and OpenOffice.org.

[Sam]

Well put, Zaine-

"[Microsoft] will someday see that OXML is a dead-end street."

It already is, and they see it; but they have a sunk cost problem.

If you look at the reorgs around MSN/Live (whateveritis), this reflects that "sight"...that they need to shore up the financials to replace Vista/Office revenues because -- we are AT the end of the alley...where the hobos sleep over each other for warmth like a fresh litter of puppies and the cats scrounge in the dumpsters for scraps of food and you can hear Billy Holiday sing my man is gone on a transistor radio three flights up the rusty old fire escape.

[Wesley Parish]

Well, Sam, my expression for Microsoft's current whateveritis, is "Is Pan Am the Microsoft of today?"

Disentangling that question takes a bit of thought, and hopefully answers itself. But Pan Am did sink at least one of their competitors - Oceanic - on the California-Hawaii route, by threatening both Boeing and Douglas it would not buy their new jet airliners if they sold to Oceanic. Comparisons to Microsoft's OEM contracts are at your own risk ... ;)