Microsoft Watch Finally Gets it - It's the Business Applications!- Obla De OBA Da - 0 views
-
To be fair, Microsoft seeks to solve real world problems with respect to helping customers glean more value from their information. But the approach depends on enterprises adopting an end-to-end Microsoft stack—vertically from desktop to server and horizontally across desktop and server products. The development glue is .NET Framework, while the informational glue is OOXML.
-
Microsoft seeks to create sales pull along the vertical stack between the desktop and server.
-
The vertical stack is actually desktop - server - device - web based. The idea of a portable XML document is that it must be able to transition across the converged application space of this sweeping stack model.
Note that ODF is intentionally limited to the desktop by it's OASIS Charter statement. One of the primary failings of ODF is that it is not able to be fully implemented in this converged space. OOXML on the other hand was created exactly for this purpose!
So ODF is limited to the desktop, and remains tightly bound to OpenOffice feature sets. OOXML differs in that it is tightly bound to the Vista Stack.
So where is an Open Stack model to turn to?
Good question, and one that will come to haunt us for years to come. Because ODF cannot move into the converged space of desktop to server to device to the web information systems connected through portable docuemnt/data transport, it is unfit as a candidate for Universal File Format.
OOXML is unfi as a UFF becuase it is application - platform and vendor bound.
For those of us who believe in an open and unencumbered universal file format, it's back to the drawing board.
XHTML+ (XHTML + CSS3 + RDF) is looking very good. The challenge is proving that we can build plugins for MSOffice and OpenOffice that can fully implement XHTML+. Can we conver the billions of binary legacy documents and existing MSOffice bound business processes to XHTML+?
I think so. But we can't be sure until the da Vinci proves this conclusively.
One thign to keep in mind though. The internal plugins have already shown that it is possible to do multiple file formats. OOXML, ODF, and XML encoded RTF all have been shown to work, and do so with a level of two way conversion fidelity demanded by existing business processes.
So why not try it with XHTML+, or ODEF (the eXtended version of ODF en
-
-
Microsoft's major XML-based format development priority was backward compatibility with its proprietary Office binary file formats.
-
This backwards compatibility with the existing binary file formats isn't the big deal Micrsoft makes it out to be. ODF 1.0 includes a "Conformance Clause", (Section 1.5) that was designed and included in the specification exactly so that the billions of binary legacy documents could be converted into ODF XML.
The problem with the ODF Conformance Clause is that the leading ODF application, OpenOffice, does not fully support and implement the Conformance Clause.
The only foreign elements supported by OpenOffice are paragraphs and text spans. Critically important structural document characteristics such as lists, fields, tables, sections and page breaks are not supported!
This leads to a serious drop in conversion fidelity wherever MS binaries are converted to OpenOffice ODF.
Note that OpenOffice ODF is very different from MSOffice ODF, as implemented by internal conversion plugins like da Vinci. KOffice ODF and Googel Docs ODF are all different ODF implementations. Because there are so many different ways to implement ODF, and still have "conforming" ODF documents, there is much truth to the statement that ODF has zero interoperabiltiy.
It's also true that OOXML has optional implementation areas. With ODF we call these "optional" implementation areas "interoperabiltiy break points" because this is exactly where the document exchange presentation fidelity breaks down, leaving the dominant market ODF applicaiton as the only means of sustaining interoperabiltiy.
With OOXML, the entire Vista Stack - Win32 dependency layer is "optional". No doubt, all MSOffice - Exchange/SharePoint Hub applications will implement the full sweep of proprietary dependencies. This includes the legacy Win32 API dependencies (like VML, EMF, EMF +), and the emerging Vista Stack dependencies that include Smart Documents, XAML, .NET 3.0 Libraries, and DrawingML.
MSOffice 2007 i
-
- ...6 more annotations...
The binding model for OOXML is "Smart Documents", and it is proprietary!
Smart Documents is how data, streaming media, scripting-routing-workflow intelligence and metadata is added to any document object.
Think of the ODF binding model using XForms, XML/RDF and RDFA metadata. One could even use Jabber XMP as a binding model, which is how we did the Comcast SOA based Sales and Inventory Management System prototype.
Interestingly, Smart Documents is based on pre written widgets that can simply be dragged, dropped and bound to any document object. The Infopath applicaiton provides a highly visual means for end users to build intelligent self routing forms. But Visual Studio .NET, which was released with MSOffice 2007 in December of 2006. makes it very easy for application and line of business integration developers to implement very advanced data binding using the Smart Document widgets.
I would also go as far to say that what separates MSOOXML from Ecma 376 is going to be primarily Smart Documents.
Yes, there are .NET Framework Libraries and Vista Stack dependencies like XAML that will also provide a proprietary "Vista Stack" only barrier to interoperability, but Smart Documents is a killer.
One company that will be particularly hurt by Smart Documents is Google. The reason is that the business value of Google Search is based on using advanced and closely held proprietary algorithms to provide metadata structure for unstrucutred documents.
This was great for a world awash in unstructured documents. By moving the "XML" structuring of documents down to the author - workgroup - workflow application level though, the world will soon enough be awash in highly structured documents that have end user metadata defining document objects and