John Weiss <[EMAIL PROTECTED]> writes:
| I second the, "Well done," regarding the whole shebang.
>
| One idea I wanted to toss to you, Lars: two parsers in the LyX core.
| We'd keep the old format's I/O in place while adding the classes
| needed to parse & write the XML. This gives us an "instant"
| regression test should we choose to change any of the core data
| structures.
I won't remove anyting from the lyx sources until we have something
that works just as well as what we have now.
But I am not sure that I want the two parsers to co-exists, compile
time switch does it for me.
| Note that the move to an XML-format for LyX raises a new possibility:
| making all accented chars into XML char-entities, translating them
| accordingly when exporting to LaTeX. I'm not suggesting we do this
| right away. However, we should think about it, especially once the
| XMl-format branch stabilizes (and yes, I suggest branching & working
| over there while we play around).
Accents should be no or problem when using UTF-8. I don't think we
will need entities for them.
--
Lgb
