Unix Group Takes UK Standards Body To Court Over OOXML

superglaze writes "Halfway through the two-month window of opportunity during which OOXML's ISO standardization can be derailed by a formal objection from a national standards body, the UK Unix Users Group is trying to force the British Standards Institution to do just that. According to the Unix Users Group, the BSI used a flawed decision-making process when they chose to approve OOXML in the ISO vote. 'The UKUUG is also folding in many other complaints about Office Open XML (OOXML), such as unresolved patent issues and a lack of completion in the specification's documentation, and is calling for the High Court of Justice to force a judicial review of the BSI's decision.' This is not the first time a country's ISO vote has been challenged."

Comments

Re:I am lost?

[DarkOx]

The real problem with ooXML is that its a bad standard. Its bad because it really fails to specify how to encode a document. It fails in essentially two major place. The first could be corrected, its simply that there are a number of ambiguities around the formating(display),percision(display), and storage(document) of non integer numeric values. A spread sheet should calculate and spit out the same results regardless of the software you open it in.

The second issue is ooXML allows large binary blobs of virtually any type to be encoded in the document. Binary in XML and in office documents is not all bad. Certainly for multi-media type things like pictures and sounds its appropriate. I would argue however that there should be limits on WHAT binary formats are allowed. Those should reference other standards. Being able to parse out the stucture of a document only to discover all the content is locked up in some binary format you have no idea what the stucture of is, is downright useless. The reason for standards is so that people can interoperate if you can't do that then the standard is broken.

Before people jump all over me about how being able to interoperate does not mean that you can display he document exactly as it is in Word or whatever let me say "I know that". The content should be accessible though. Rendering should be about how the user wants to display it. A blind person might want a text to speach enginge to read a document to them. The standard should allow them in all cases to dump out the text data from the document. It should be possible to run into the binary objects and have the software say "there is an image here" etc etc. That's usefull "Document, contains unkown data" is not.