> Plain text is the smallest common denominator, and with a working text > importer for paragraph _and_ character styles, its use still makes sense > if you need to handle input from many sources with different degrees of > experience and different software installations.
These thoughts are very much like what I have, BUT working with technical texts is not possible with plain text (where are my micro, ohm, degree etc symbols in plain ascii??). In that case I would need unicode text, and unfortunately I don't think authors would handle that. My idea is to use HTML, but limited to a chosen number of tags. That way you could use HTML entities for special symbols. Although I haven't tried it out yet:-) /Peter
