# Re: More on subscripts and superscripts

• To: math-font-discuss@cogs.susx.ac.uk
• Subject: Re: More on subscripts and superscripts
• From: Michael Downes <MJD@MATH.AMS.ORG>
• Date: 11 Aug 1993 11:47:32 -0400 (EDT)

> True, but I can still see this breaking under unusual' conditions,
> for example _ inside math mode or inside a \write.

(?) _ in math mode can only be subscript OR \_, but not both. If _ is
made active' via mathcode "8000, it will work all right inside a
\write (or conversely, it will only break where it would have broken
before). But generally, I agree with your next words:

> ... This sort of thing is better left to macro
> packages rather than by hacking around with plain (IMHO).

Re mathchoice etc.:

> Agreed, the decision not to let the current style be accessable in
> math mode is one of the more... er... ideosyncratic design decisions
> in TeX!  Unfortunately, it would be a drastic change to TeX to allow
> the current style to be tracked, since it would require new syntax for
> ${...\over...}$ and friends.

Agreed, except I would have put it the other way around: the decision
to use syntax like {...\over...} for \over and friends was the design
flaw that necessitated all the jury-rigging around \mathchoice and
made it difficult to provide a current math style variable.