At 18.45 +0100 01-02-24, Chris Rowley wrote:
>Can we agree that one question we need to ask is:
>  what is LaTeX's Internal Representation of all `math-mode token lists'?


With the current LaTeX character handling, there is an ambiguity on the
input side---hence my remarks about that---but the matter should rather be
delt with on the output side. I very much agree with:

>We should probably pin this down before venturing into Unicode for

Lars Hellström