Document Wars

Glynn Moody, who is quoted in the article, obviously does not understand what was involved in Massachusetts. Sam Hiser, also quoted, was deeply and personally involved. OOXML wasn't even on the horizon then. Microsoft Office 2003 wrote to a flat XML format that was irrelevant in any event. What Massachusetts wanted --- and deserved from the ODF community --- was an ability to integrate OpenOffice.org with business processes that were already thoroughly bound to Microsoft Office. The problem was in fact solved by the OpenDocument Foundation's ODF plug-ins for Microsoft Office, but the project had to be ditched because Sun would neither adapt OOo nor allow ODF to be adapted for the purpose. Gary Edwards and I wrote an in-depth and heavily-referenced article on why ODF failed in Massachusetts. http://www.linuxworld.com/news/2007/072307-opendocuments-grounded.html (.) Those who do not comprehend that integration of data created by end point solutions and stored in legacy data silos is a fundamental requirement in service oriented architectures will never comprehend why interoperability is so vitally important. They wind up being unwitting advocates of incredibly expensive rip out and replace solutions. Good luck, particularly in the current economic climate. The wonderment is why anyone believes that Microsoft Office can be toppled by ODF from its monopoly position without ultra-high fidelity interoperability with Microsoft Office. Too much "what works well enough for me works well enough for anyone" mindset, I suspect.

This article compliments the previous publication, "The better Office Alternative - Softmaker Office". Good stuff! Excerpt: "It's the question that vexes free open source software advocates and commercial competitors around the globe: Why is Microsoft Office so difficult to dislodge from its perch atop the IT heap? Is it the exclusive bundling deals? The deep Software Assurance entrenchment? Steve Ballmer's backroom deal with the devil?" "The answer, of course, is none of the above (though some evidence of a Microsoft-Hell alliance exists). Rather, it's the Office ecosystem -- the vast library of third-party add-ons and vertical solutions built (with copious encouragement from Microsoft) on Office's extensive programmatic model -- that makes Microsoft's suite so hard to kill."

Exactly! It's not that ODF is under-specified. It's that ODF can't be fully specified until OpenOffice is completely documented and capable of supporting expected features. There is this famous quote from Sun's Svante Schubert that pretty much says it all; "Nothing goes into the ODF specification unless it's supported by OpenOffice". The statement was made at a meeting of the OASIS ODF Metadata SC while discussing the controversial use of "XML ID". IBM's Elias Torres, of RDFa fame, was passionately arguing that use of the XML ID should be left open to all developers. Sun had taken the position that XML ID should be limited to only a handful of elements supported by OpenOffice. The discussion acutally got far worse than the quote would indicate. Elias compromised his arguments suggesting that we should allow developers to have at it with XML ID for at least one year, and then revisit the issue. This suggestion lead to a discussion of how developers implementing elements with metadata would notify the metadata sub committee of their use case. A discussion list was proposed. The idea being that developers would send in their use cases and the oligarchs on the sub c would approve or disprove. Incredibly, this suggestions was shot down by Bruce d'Arcus (OpenDocument Foundation). Bruce thought that any developer needing metadata support for particular elements should have to join the OASIS ODF Metadata SC like everyone else before their needs would be considered. ( i.e. - like the other oligarchs). At the next weeks meeting, Rob Weir showed up with a list of 14 spreadsheet related elements that IBM needed XML ID support for. Sun representatives Svante Schubert and Michael Brauer immediately approved the use, agreeing that OpenOffice would support that implementation. This how things work at OASIS ODF. Ever wonder why SVG or XForms support in ODF is so limited? It's because the specification directly reflects the limits of the OpenOffice implement

Sorry. Diigo cut my comment off about half way through. I've complained to Wade and Maggie about this problem, especially how it impacts and cripples the "Group Auto-Blog Post" feature . Months have gone by. Yet still the issue remains.

Rick Jellife weighs in on the OpenOffice ODF- MSOffice OpenXML interop embroglio. His take is to focus on Classes of Fidelity, providing us with a comparative table of fidelity categories. I wonder though if this über document processing approach is anywhere near consistent with the common sense meaning of interoperability to average end-users? IMHO, end-users interpret "interoperability" to mean that compliant applications can exchange documents without loss of information. "..... In my blog last year Is ODF the new RTF or the new .DOC? Can it be both? Do we need either? I raised the question of whether ODF would replace RTF or DOC. I think this issue has come back with a bang with the release of Office 2007 SP2, and I'd like to give another pointer to it for readers who missed it first time around.... "...... OASIS ODF TC has some kind of conformance and testing wing at work, but it is not at all clear that they will deliver anything in this kind of area. Without targetting these classes, ODF's breezy conformance requirements means that ODF conforment software can deliver vastly different kinds of fidelity, yet still accord to the letter of the law (and, indeed, to the spirit of the ODF spec, which allows so many holes) which will cause frustration all-around....." Ouch!

ODF is important. So What Went Wrong? Response to Jeremy Allison: Having participated in a number of government pilot studies, I must say that you are right; government officials do care about ODF. They really want it to work. But they also had expectations that ODF simply wasn't designed for. What they expected ODF to be was an open technology based on highly-structured XML markup that was application, platform, and vendor-independent, backward compatible, universally interoperable, and importantly, Web ready. That is not ODF nor is it OOXML. In fact, the closest thing we have for meeting these expectations is an ajax-webkit style HTML+ (HTML5, CSS4, SVG/Canvas, JS jQuery, etc.). ODF is highly structured, but it is not application-independent. .....

Is there nothing that can cool the flames of this document war? Interesting coverage of the recent OOXML Interoperability event in London (Monday). Real stuff not talk. Since Florian, Jason and i are working on an OpenWeb ready HTML+ layer riding over OOXML there are some things mentioned that look very interesting. ..."An Opera browser plug-in for Open XML Document Viewer v1.0 was released at the meeting; the tool provides direct translation for Open XML documents (.DOCX) to HTML, enabling access to Open XML documents from any platform with a Web browser, including mobile devices. The document-viewing software already includes a plug-in for Firefox, Internet Explorer 7 and Internet Explorer 8...." ..."Microsoft and the other participants in Monday's forum also made available a beta of Apache POI 3.5, a Java API (application programming interface) to access information in the Open XML Format....."

Thanks to Paul for digging this up. Who would have guessed that years later, these same issues hang like a dark shroud on the future of ODF? Note also that June 1st of 2007 was the cut off date for ODF 1.2 proposals and recommendations. The OpenFormula and Metadata SC's were rushing to make the cutoff. The List Enhancement proposal itself was just one of many enhancements submitted by Florian Reuter in November of 2006, designed to greatly improve ODF compatibility with MSOffice "ODF". By November of 2006, thanks largely to the Massachusetts Pilot Study, there were a number of ODF plug-ins for MSOffice. All were capable of producing perfectly compliant ISO 26300 ODF, but falling far short of public expectations of high fidelity interop with OpenOffice ODF. Sound familiar? Everyone knew that it was only a matter of time before Microsoft was pressed into providing MSOffice ODF support. There was no doubt that they would face the exact same interop challenges as the many independent plug-in efforts. Hence the stepped up efforts by many at the OASIS ODF to "fix" ISO 26300! At the time of the List Enhancement Proposal, we had increasing evidence from the many pilot studies that ODF was impossible to implement in business and workgroup environments where the MSOffice productivity environment was the defining platform. ODF was not designed to be compatible with MSOffice or the binary documents so critical to business processes bound to this environment. OpenXML was designed exactly to be compatible with these environments. Unless ODF fixed it's compatibility/interoperability problems there was no way for the independent plug-ins to provide a reasonable ODF implementation alternative to OpenXML. And even if Microsoft did produce an MSOffice ODF compliant with ISO 26300, these productivity environments would remain entirely locked. The world expected ODF to be compatible, interoperable, Web ready, and fully capable of cracking open the iron grip Mic

Also includes stats in table form indicating that according to Google Search OOXML documents now outnumber ODF documents on the Web, for word processor documents, spreadsheet documents, and presentation documents.

.... "I recently described how Google's Wave, a collaboration tool based on the new HTML 5 standard, demonstrated the potential for Web applications to unglue Microsoft's hold on customers. My post quoted Gary Edwards, the former president of the Open Document Foundation, a first-hand witness to the failed attempt by Massachusetts to dump Microsoft and as experienced a hand at Microsoft-tilting as anyone I know......"