Commentary: ISO should kick OOXML off the standards bus
Posted by: Anonymous
on January 27, 2008 03:52 PM
A Stephane Rodriguez comment published here: http://www.noooxml.org/forum/t-35292/wishful-spinning: The XML proposed in OOXML is not a general purpose language for Office documents. The whole point of XML is to design it in such a way that it interoperates with applications out there, including those which don’t exist yet. Therefore that XML should be designed in such a way, and those in the trenches should be able to find clues about that. Problem : there is no such clue, because it’s really a poor XML. A good example I like to come up with is that there is a gazillion ways to describe text formattings (no less than 6 for Excel spreadsheets alone). Microsoft designed a bad XML that puts the burden on implementors who will spend their time on those details rather than productive time bridging it with applications across platforms. That’s why it should be rejected : it does not really benefit anyone except Microsoft. And even that remains (Microsoft is the sole benefitor) to be seen : from my own personal experience of Office file formats, it appears that nothing in OOXML will help create the next big things coming for Office (those that I see coming given what the workplace is moving towards).
Another thing that should be reminded is that having to spend so much time on details because OOXML is poorly designed is just the tip of the iceberg. The reality is that we already have two OOXML variants, and two more are coming when Office 2009 ships.
- OOXML 1.0 (i.e. ECMA 376 today)
- Office 2007 (i.e. OOXML 1.0 + all undocumented bits + all fixes)
- OOXML 1.1 (whatever is the outcome of Feb’s BRM)
- Office 2009 (OOXML 1.1 + undocumented bits).
Implementors will have to implement all 4 or will be unable to open an arbitrary document based on this thing called “OOXML”.
Among undocumented bits : macros, macro bindings, DRM, encryption, sharepoint metadata, …