Move to Standardize MS XML an Uphill Climb

For a revealing look at the internals of MS XML versus OpenDocument, see this useful post by the gentlemen at the OpenDocument Fellowship (Daniel Carrera, Bruce D'Arcus, J. David Eisenberg and Alex Hudson).

MS XML was designed for implementation by one office suite while OpenDocument was designed for implementation by any and all office suites. The internals tell the weather.

The article's implied argument is strongest on reuse of existing standards: OpenDocument scores well in its implementation of existing standards -- on XLink and with Metadata, for example -- while MS XML betrays a distinct lack of respect for extant standards, working with data in ways to which developers are not accustomed.

(What bleeds through the lines here is hubris: that the Microsoft engineers made the error of assuming their work would become the referenced implementation.)

UPDATE 11/28/2005

There is a relevant comment on Brian Jones' blog about this topic.

If Microsoft wishes to satisfy customers' requirements, the ones on which they'll need to work the hardest are interoperability and open collaboration by multiple parties in formulating the standard. Moreover, the sloppy internals of MS XML indicate the need for a re-think of their exploitation of the XML rules if they have sincere ambitions for international standardization and acceptance by developers of non-Microsoft office suite products.

MS XML as an "open" standard is 12 to 18 months behind OpenDocument in organizing within a standards body and about 5 years behind in testing and usage. As a reflection of innovation, OpenDocument makes it look positively antique. The subtext of this cautionary tale is that the best ideas are those that survive the white-hot indignities of wide scrutiny, open testing and public humiliation. Opening up is a cultural journey, a state of mind and not a checklist of customer requirements.

The road ahead for Microsoft on the file format looks Sisyphaean; on that the view is clear.