Group items tagged

A legthy discussion of the MSOffice bound MOOXML file format. that was triggered by Rob Weir's infamous blog, "How to Hire Guillame Portes".  Lots of comments, pro and con, as to whether or not the applications specific tags used so extensibvely by MOOXML are needed, or not.

Many argue that the application bound tags should have been fully described.  That every tag in the specification should also include the informaiton needed to implement it.  Agreed!  Otherwise, the specification does not qualify as a standard.  It's simply a vendor specific, and in this case highly proprietary and encumbered file format.

Others argue that the app bound tags are the only way for Microsoft to provide backwards compatibility with the billions of binary documents bound to MSOffice through proprieatry and secret binary file formats.  These people argue that embedding an application specific binary in the XML file format, instead of converting it to proper XML, is the only way to insure backwards compatibilty.  BS.  There is no technical reason not to convert it to proper XML.  But that would mean fully describing the binary objects, including the nature of their application dependency.  Something Microsoft is quite reluctant to do.

The truth of the matter is that if the binary object is to be part of a specification submitted to ISO for standards consideration, then it should be fully described, including how to implement it.  Otherwise, MOOXML is just a standard for one.  A standard for Micrsoft only since they are  the only ones with the secret blueprint as to how to implement these binary objects.

A legthy discussion of the MSOffice bound MOOXML file format. that was triggered by Rob Weir's infamous blog, "How to Hire Guillame Portes". Lots of comments, pro and con

A legthy discussion of the MSOffice bound MOOXML file format. that was triggered by Rob Weir's infamous blog, "How to Hire Guillame Portes". Lots of comments, pro and con

The primary difference between ODF and MOOXML is that ODF was designed to be a universal file format.  MOOXML was designed to be an XML file format for MSOffice, the Win32 API, and the Vista Information Processing Chain API (.NET 3.0). 

ODF is application and platform independent.  MOOXML is application and platform specific.  It's bound to the Windows - Vista platform. 

Microsoft's Brian Jones recently got caugh tup in a argument with the heavily armed WMD ODF expert and combatant Sam Hiser (WMD=Words of Massive Destruction).  In their exchange, Brian got confused over this very important distinction between ODF and MOOXML.  ODF allows specific applications to place their configurations and requirements in a settings file that is separate from the content, presentation and metadata containers.  MOOXML on the other hand makes no distinction whatsoever between application specific (MSOffice only) configuration, settings, processsing instructions and systemm dependencies and the rest of the file format contents.  Application settings are bound to content, presentation, and schema containers.  So bound that Brian is seemingly unaware of what ODF has achieved.  Sam caught him by surprise, as did many others posting comments:

Brian Jones on MOOXML support for older versions of MSOffice:  Coments by Sam the WMD Man are below.

Wow! Awesome summary of an exhaustive testing process comparing MSOffice 2007's compatibility with legacy MSOffice installations. Vital to MSOffice bound workgroups - workflow business processes.

This is the fifth of the six major iX - interoperability enhancement proposals submitted to the OASIS ODF TC - SC between July 2006 and February of 2007. This particular iX proposal lead to the "List Enhancement Proposal" donnybrook that consumed the OASIS ODF TC for the next six months, ending with the OpenDocument Foundation being booted out of OASIS in May of 2007. The six iX proposals were all different approaches to the same basic problem: ODF was not desinged to be interoperable with MSOffice documents, applications or bound processes. The proposals come out of the OpenDocument' Foundation's efforts to save ODF in Massachusetts. ODF iX repressents a subset of ODF designed to grealty improve compatibility with MS binary and XML formats. With the ODF iX subset, the da Vinci plug-in would be able to convert the billions of MSOffice binary and xml documents with a very high level of fidelity, and do so within the bounds of "round trip" business processes. The most basic iX approach was to add five generic elements to the existing ODF specification. The five generic elements would cover lists, tables, fields, sections, and page dynamics (breaks). It is a well known fact that these five areas of incompatibility between OpenOffice ODF and MSOffice binaries represent 95% of all conversion fidelity problems. MSOffice has one way of implementing lists, and, OpenOffice has another. These application specific implementation models are irreconcilably different. It's also true that the applicaiton specific implementation models are directly reflected in each file format. So applications implementing ODF must also implement the OpenOffice model for lists, fields, tables, sections and page dynamics-page positioning if they are to have any meaningful measure of exchange fidelity. Perhaps the best of the iX approaches was that based on the innovative use of metadata to describe presentation-layout attributes.

"aimed at achieving an optimum degree of order .... and .... aimed at the promotion of optimum community benefits:. Uh, excuse me Mr. Robinson, tha tsecond part of your statement, the one concerning optimum community benefits - that would also disqualify ODF!! ODF was not designed to be compatible with the 550 million MSOffice desktops and their billions of binary docuemnts. Menaing, these 550 million users will suffer considerable loss of information if they try to convert their existing documents to ODF. It is also next to impossible for MSOffice applications to implement ODF as a fiel format due to this incompatbility. ODF was designed for OpenOffice, and directly reflects the way OpenOffice implements specific document structures. The problem areas involve large differences between how OpenOffice implments these structures and how MSOffice implements these same structures. The structures in question are lists, fields, tables, sections and page dynamics. It seems to me that "optimum community benefits" would include the conversion and exchange of docuemnts with some 550 million users!!!! And ODF was clearly not designed for that purpose!

I don't agree with this statement from Microsoft's Oliver Bell. As someone who served on the OASIS ODF Technical Committee from it's inception in November of 2002 through the next five years, i have to disagree. It's not that Microsoft wasn't welcome. They were. It's that the "welcome" came with some serious strings. Fo rMicrosoft to join OASIS would have meant strolling into the camp of their most erstwhile and determined competitors, and having to ammend an existing standard to accomodate the implementation needs of MSOffice. There is simply no way for the layout differences between OpenOffice and MSOffice to be negotiated short of putting both methodologies into the spec. Meaning, the spec would provide two ways of implementing lists, tables, fields, sections and page dynamics. A true welcome would have been for ODF to have been written to accomodate these diferences. Rather than writing ODF to meet the implementation model used by OepnOffice, it would have been infinitely better to wrtite ODF as a totally application independent file format using generic docuemnt structures tha tcould be adapted by any application. It turns out that this is exactly the way the W3C goes about the business of writing their fiel format specifications (HTML, XHTML, CSS, XFORMS, and CDF). The results are highly interoperable formats that any applciation can implement.

You can harmonize an application specific format with a generic, applicaiton independent format. But you can't harmonize two application specific formats!!!!
The easy way to solve the document exchange problem is to leave the legacy applications alone, and work on the conversion of OOXML and ODF docuemnts to a single, application independent generic format. The best candidate for this role is that of the W3C's CDF.
CDF is a desription of how to combine existing W3C format standards into a single container. It is meant to succeed HTML on the Web, but has been designed as a universal file format.
The most exciting combination is that of XHTML 2.0 and CSS in that it is capable of handling the complete range of desktop productivity office suite documents. Even though it's slightly outside the W3C reach, the most popular CDF compound is that of XHTML, CSS and JavaScript. A combination otherwise known as "AJAX".

The war is over?  When did Microsoft surrender?  And when did they sign the official terms of surrender?

The terms of surrender are simple. Microsoft must agree to fully support and implement ODF as a native file format in all versions of MSOffice qualified for the current OOXML compatibility kit. Furthermore, MSOffice must offer end users the choice of selecting ODF as the default MSOffice file format.

Those are the terms of surrender, and i for one don't see how the Microsoft or Novell Translator plugin's qualify?  These things are garbage!

What if an MSOffice user was to work on a document, save it to OOXML only to open it later to find a near totally useless and corrupted document with a conversion fidelity equal to that achieved by the hapless MCN Translator Plugins?

Right.  What's good for the goose is good for the gander.  Until these idiotic MCN Translators can achieve a conversion fidelity between ODF and OOXML acceptable to MSOffice users - comparable to native documents use and expectations, they should be regarded for what they are: an experimentation proving conclusively that OOXML is not even close to being interoperable with ODF.

~ge~

Marbux has done some heroic work here, using the GrokDoc Wiki.  The Title is "EOOXML Objections", and it's primary purpose is to help ISO National Body Memebers evaluate the 0ver 6,000 pages of the Microsoft - ECMA Office Open XML Specification for MSOffice. 

On January 5th, 2007, Microsoft officially submitted EOOXML to ISO under the fast track rules.  Before EOOXML can hit the fast track though, ISO provides members with a 30 day "Contradition Review Phase".   During this brief phase, ISO NB's (national standards body members) muct evaluate the proposal and post their allegations concerning contradictions and inconsistencies with other ISO products - like ODF.

What Marbux is assembling here is a one stop shop for ISO NB's strugglign to understand the issues at stake.  It's incredible wha the has accomplished in such a short time.  But then, the clock is ticking.  February 5th is a hard and unmovable deadline. 

The basic contradiction is thatt EOOXML is a subset of ISO existing product, ODF.  Both attempt to do the exact same thing:  provide an XML file format for desktop productivity environments such as MSOffice, OpenOffice, and WordPerfect Office.  What seriously differentiates the two is that ODF was designed expressly to be a universal file format, application and platform independent, able to transition across many different information domains connecting the legacy of desktop productivity to near everything else.  MOOXML on the other hand was designed for MSOffice and the legacy of billions of binary documents that only Micrsoft knows the secrets to converting to XML.  As such, MOOXML is designed to be application and platform bound, with these proprietary dependencies written right into the specification.

One of the more important elements of the Marbux arguments is that the OpenDocument Foundation's daVinci Plugin and InfoSet Engine - API prove conclusively

A mus tread.  Carol Sliwa of ComputerWorld intervies Clark Kelso, California CIO.  ODF is the main issue, with clark casting all his answers in the context of business decisions.  Carol o fcourse is asking the best questions of any journalist alive.

Keep in mind that ComputerWorld and the Boston globe filed for the Freedom of Information Act to be invoked in Massachusetts.  They got access to all the eMail, documnetation, and conferencing notes concerning ODF  and Microsoft.  Carol's interview with Louis Gutierrez last week was filled with the same hard questions Clark Kelso fielded so deftly.

The "committee" Clark Kelso has set up to look at these issues is headed by Bill Welty, the CIO of the California Air Resources Board.  Bill is a long time opensource - Linux guy, but will be the firs tto admit that Microsoft is the only vendor providing a means of getting everything inot XML.  And that's the heart of any SOA strategy, "First, get everything into XML".

With a 500 million MSOffice desktop bound business process headstart, Microsoft has the extreme advantage in this much needed migration to XML. 

They now have their own proprietary application and platform bound version of XML; MOOXML (Microsoft OfficeOpenXML) heading for international standardization at ISO. 

They now have their XML Hub in place; the Exchange4/SharePoint Hub.  This is also an essential part of any SOA strategy.  You've got to have an XML Hub where the XML information streams and service connection to legacy black box systems can be piped into, managed and resolved.  The XML must also provide an end user interface to these information flows.  One that converges and integrates information, documents, data, and workflows into an easy to manage and participate in interface.  The E/S Hub excells at this because it covers the fundamentals of eMail, messaging, portal, calendar, scheduling, c

Great inteview, i'll comment as i make my way down the page.  Hopefully others will do the same.

The List Proposal donnybrook refuses to go away. In a recent argument concerning Rick Jellife's controversial post, "Harmonization by augmenting ODF with OOXML elements", ODF defender Bruce D'Arcus pointed to his own OASIS ODF TC message thread in an effort to defend Sun.

The issue is that Rick Jellife and others are wondering why it is that Sun opposes interoperability enhancements to ODF that would solve the problem of converting MS binary and XML documents to ODF, and back.

Bruce's reference introduces the incredibly acrimonious "List Proposal Vote Deadline" thread. He also brings up a really important problem. The ODF Charter does not reference compatibility with existing file formats as a key objective.

The consequences of this neglect in the ODF Charter is that every time the issue of compatibility with existing MS binary or xml documents comes up, Sun claims it's, "Outside the Charter and Out of Scope".

I've been hearing that excuse for the last five years!! Meanwhile, the world has discovered it's impossible to implement ODF without also having to totally rip out and replace MSOffice. And that means a costly re engineering all existing business processes, line of business integrated apps, and assistive technology add-ons.

ODF failed in Massachusetts because Sun and OASIS TC refused to recognize the importance of an ODF plugin for MSOffice offering the same high fidelity "round trip" conversion as the OOXML plugin for MSOffice.

For one reason or another, big ODF vendors are locked into a "rip out and replace - legislative mandate" strategy. A strategy to limit the interoperability of ODF with MS documents, applications, and processes has only one consequence. As Massachusetts proved to the world, ODF is impossible to implement if you have workgroups and business processes based

Yoon Kit brings up an interesting point about the ISO consideration of MSOOXML (Ecma 376);  ISO approval of MSOOXML would backdoor a good many MS proprietary technologies that compete directly with W3C XML standards.

YK gives the example of MS VML, which competes with the W3C SVG standard used by ODF.  He could have also cited that legacy versions of MSOffice (98-2003) make use of VML as the default graphic format, while MSOffice 2003 9with XML plugin) and MSOffice 2007 (by default) implements DrawingML as the replacement for VML. 

So, would ISO approval of Ecma 376 backdoor VML and DrawingML in as "standards"?  Or MSOffice MathML?   One has to wonder since they are essential to MSOOXML.

I've got to give this one extra points for creativity!  All anyone has to do is visit the W3C web sites for CDF WICD Full 1.0 to realize that there is in fact a CDf profile for desktops.  CDF WICD Mobile is the profile for devices.

My guess is that Chris Lilley is threading the needle here.  IBM, Groklaw, and the lawyer for OASIS have portrayed the Foundation's support for CDF WICD Full as a replacement for ODF - as in native file format for OpenOffice kind of replacement.  Mr. Lilley insists that CDF WiCD Full was not designed for that purpose.  It's for export only!  As in a conversion of native desktop file formats.

Which is exactly what the da Vinci group was doing with MSOffice.  The Foundation's immediate interest in CDF WICD was based on the assumption that a similar conversion would be possible between OpenOffice ODF and CDF WICD.

The Foundation's thinking was that if the da Vinci group could convert MSOffice documents and processes to CDF WICD Full, and, a similar conversion of OpenOffice ODF documents and processes to CDF WICD could be done, then near ALL desktop documents could be converted into a highly interoperable web platform ready format.

Web platform ready documents from OpenOffice?  What's not to like?  And because the conversion between ODF and CDF WICD Full is so comparatively clean, OpenOffice would in effect, (don't go native file format now) become ahighly integrated rich client end user interface to advancing web platforms.

The Foundation further reasoned that this conversion of OpenOffice ODF to CDF WICD Full would solve many of the extremely problematic interoperability problems that plague ODF.  Once the documents are in CDF WICD Full, they are cloud ready and portable at a level certain to diminish the effects of desktop applications specific feature sets and implementation models.

In Massachusetts, the Foundation took

This quote from Chalres Schultz is ridiculous. Because Novell is not allowed to commit code to OpenOffice, they must maintain a separate code base of extensions and improvements. With each build of OpenOffice, Novell must reintegrate their changes into the code base, making for a managerial nightmare. When Novell does have improvements that Sun wants though, there is no end to the hoops of fire the Sun developers will jump through to get it. The Field Enhancement routine written by Novell's Florian Router is one of those improvements that Sun had to have. Sun even went so far as to arguing for changes in the way ODF implements fields to accomodate the Novell improvements! It's important to note however that Sun did not support the ODF Field Enhancements UNTIL Novell agreed to donate Florian's code to OpenOffice!!!!!! Proving conclusively what i have been arguing for years: Sun does not allow for any changes to ODF unless and until those changes can be implemented by OpenOffice. The ODF Field Enhancements needed by Florian's fix to OpenOffice were originally proposed on July 12th, 2006, when Florian was the CTO of the OpenDocument Foundation. These changes to the way ODF implements fields were needed by the da Vinci plug-in as part of our efforts to save ODF in Massachusetts. so here we have a rather direct example of Sun refusing improvements to ODF when needed by another application (da Vinci), but supporting those exact same changes when it is OpenOffice that can be improved!!! The arguments that the OpenOffice.org Community isn't open also apply to the OASIS ODF TC work!!!!!!

Good catch by Eric!
This link is to the infamous Sun statement of support for MS OOXML issued by Jon Bosak when ISO DIS 2900 was voted on by the US delegation to ISO.
The statement is important because it directly references the core issue: MS OOXML was written for MSOffice and the billions of binary docuemnts bound to that application suite. ODF on the other hand was written to OpenOffice.
Because ODF was not designed for the conversion of those billions of MSOffice documents, conversion is next to impossible. The implementation of ODF in MSOffice is next to impossible. The loss of information, especially the presentation-layout information, is so severe as to be intolerable in the real world.
This leaves the real world, where MSOffice dominates over 550 million desktops, unable to implement ODF. In light of this real world problem, Sun's Bosak urges support for MS OOXML as an ISO standard!!!
So we have this situation at OASIS ODF where Sun is in control of both ODF and OpenOffice, refusing in all cases to compromise the linkage or accomodate the much needed interoperability enhancemnts seeking to improve the conversion of billions of documents to ODF. And publicly supporting MS OOXML as the only pragmatic alternative to the situation Sun is responsible for!

Well, this is interesting.

What part of "Executive Board" makes you think they read 6,000 page XML specifications? <ge>

I think, in the best bureaucratic tradition, they argued definitions until they convinced themselves that they didn't need to do anything.  They decided that one standard contradicts another standard only if the proposed standard causes the existing standard not to work.  This is from analogy with the Chinese WAPI WiFi networking standard last year that was defeated because the protocol caused radio interference with existing 801.11 networks.  So they said that OOXML did not contradict ODF because both files could exist on the same disk without interfering with each other.   You will note that thiss argument can be used for every XML format, every programming language, every operating system, in fact every software standard, since software is ultimately data, and data can be segregated on disks.  So they essentially chose a definition so narrow that it nullified the concept of "contradiction" for most of what JTC1 has authority over.<!-- D(["mb","<div><br><span style\u003d\"color:rgb(0, 0, 153)\"><ge>  Wait a second.   You cannot have a OOXML document and a ODF document sitting on the same disk without having them interfer with each other.  We just proved that with our tests of both ACME 374 and ODF Da Vinci plugin on the latest release of MSOffice Word 2007.\n</span><br style\u003d\"color:rgb(0, 0, 153)\"><br style\u003d\"color:rgb(0, 0, 153)\"><span style\u003d\"color:rgb(0, 0, 153)\">OOXML clearly does interfere with the loading of an ODF file into MSWord 2007.  In prior versions of MSWord (98, 2000, XP, 2003

This is amazing stuff. With all the talk about OpenOffice ODF compatibility problems with existing MSOffice productivity environments and documents, this comparison is stunning. I stumbled across this Compatibility Comparison reading this article: ThinkFree Set to Launch The First Complete Android Office Suite. Documents To Go is currently the only provider of Word and Excel documents on Android. The ThinkFree Office comparisons to OpenOffice cover a number of familiar compatibility issues, with layout at the top of the list. ThinkFree Write 3.5 vs OpenOffice Writer 3.0 ".....When using a word processor to create documents, you really shouldn't have to worry about whether your client will be able to see the document as you intended." ".... However, if you use a low-cost solution like OpenOffice, you should be prepared for frustrations and disappointments....."

This is the report John Cody worked on.  I spent four months answering his questions but was unable to adequately explain to him the difference between an "Office Suite" and a workgroup-workflow centric "Productivity Environment".   John insists that it's entirely possible to rip-out-and-replace the MSOffice editors with the free OpenOffice Suite without disrupting important workflows and business processes.  I explained to him what happened in Massachussetts, including the 300 page pilot study report Sam wrote.  What he needs to do i think is pay close attention to the Burton Group coverage of what is now known as the SharePoint Foundation platform;  SharePoint 2010 having totally swallowed the MSOffice 2010, leaving the venerable desktop productivity office suite as an important end user interface into information rich business systems centered on the SharePoint "Unified Productivity" platform.

Marbux on metadata and the language of universal interoperability: Few people are aware of the raging debate that has pushed ODF to the edge. The OASIS ODF TC is split between those who support Universal Interoperability, and those who insist on continuing with limited ODF interoperability.

ODF (OpenDocument), formally known as Open Office XML, began it's standards life in the fall of 2002 when Sun submitted the OpenOffice file format to OASIS for consideration as a office suite XML fiel format standard. The work on ODF did not start off as a clean slate in that there were near 600 pages of application specific specification from day one of the standards work. The forces of universal interop have sought for years to separate ODF from the application specific features and implementation model of OpenOffice that began with those early specification volumes, and continues through the undue influence Sun continues to have over the ODF specification work.

Many mistakenly believed that submission of ODF to ISO and subsequent approval as an international standard would provide an effective separation, putting ODF on the track of a truly universal file format.

Marbux is one of those Universal Interop soldiers who has dug in his heels, cried to the heavens that enough is enough, and demanded the necessary changes to ODF interoperability language.

This post he recently submitted to the OASIS ODF Metadata SC is a devastating rebuttal to the arguments of those who support the status quo of limited interoperability.

In prior posts, marbux argues that ISO directives demand without compromise universal interoperability. This demand is also shared by the World Trade Organization directives regarding international trade laws and agreements. Here he brings those arguments together with the technical issues for achieving universal interop.

It's a devastating argument.

This is a follow up comment to a question cocerning my previous post, "commercialization of interoperability".  The question from "mosborne" is as follows:

A different viewI'm not on the ODF TC, but I have followed its evolution through the information publicly available at Oasis.

My outside view of some of the various interoperability discussions you mention is different than yours. I saw a resistance to adoption of features if the sole reason was because OOXML did it that way. The dissenting members wanted a more substantial reason, not simply to add OOXML "features" to ODF.

If the goal is to simply make ODF like OOXML, then what is the point? You would have conceded all control to Microsoft since they have effective control of OOXML.It's an interesting question, but not well informed.  The threads at OASIS ODF having to do with interoperability are focused on efforts to have our cake and eat it too. 

The List Enhancement Proposal thread played out over a six month period.  And yes, it is true that Sun fought the Novell proposal because they felt new and innovative features for OpenOffice/StarOffice were more important than the interoperability CIO's and IT departments are demanding.   But that misses the more important point that Novell was able to craft their interoperability proposal exactly so that the precious advanced feature sets of applications that command les sthan 1% marketshare would be accommodated.

What Sun and most others on the ODF TC don't get is that the markets have no use for these new and innovative feature sets unless and until they can transition their documents and business processes out of MSOffice.  If workgroup bound end users can't do that first, it won't matter how