I have never implemented a word processor, nor designed an XML language for expressing rich text. But somehow, I can’t help but feel that both ODF and OOXML are way too complex, and anyone wanting to ensure that documents are readable in the long run would be better off adopting XHTML, Docbook or something else that actually can be written with a simple text editor and understood by an average human being.

The best post I’ve read on this, though, comes from Steve Loughran.