Frank Mittelbach <[log in to unmask]> writes:

>  > *input* encoding, but deals with the output encoding.
>
> right and i think this is where the problem lies:

Duh...  I had a flash of revelation last night, but my home
computer is working only intermittently, so I couldn't recant.

I realized Frank's obvious point that the math-active trick doesn't
work when the input character does not match the output character.

Which Frank now repeats:

> what i meant is that the slot defined by inputenc for ä does not need to have
> anything to do with the slot defined by the current font encoding. however the
> trick you use makes use of such a relationship


Donald Arseneau                          [log in to unmask]