Karl, would that be the right kind of approach?
Yes, in general, this is the right track.
I know there has been talk in the past about a LaTeX backend to
solve some of the problems about Unicode charsets in the TeX
output. Is this is to avoid texinfo.tex having to code all the
special rules about different characters in different languages?
That's one of a number of reasons. Others (off the top of my head)
include correct hyphenation, font support (usually required for language
support), and, at the other end of the spectrum, user-tweakability of
formatting. None of that is going to happen in texinfo.tex, as far as
I'm concerned; it would be senseless to (try to) duplicate 30 years of
LaTeX infrastructure development.
K
