# Re: text chars in math fonts

>I like the idea of a second text complement'' font more than
>sticking text characters into the math fonts.  I'm sorry I didn't
>make that clear in my previous message.

How many complementary characters are there that wouldn't be required
in math mode?  I've seen \dagger and \ddagger used this way (as binary
functors in category theory, if anyone's interested) and I'm sure that
all the other glyphs from cmmi and cmsy have been used in mathematics
at some point.

In retrospect, I realized that small caps need to be in a separate
Cork-encoded font, since the large caps from c&sc may need to be
letterspaced differently from the caps from the text font, and c&sc
should have the full range of accents from the Cork encoding.

(As a side issue, would it be possible to include c&sc in the NFSS in
such a way that italic or oblique c&sc were selectable?  For example,

\section{The Warsaw Pact and {\sc nato}}

might generate bold c&sc NATO' in the A heading, and italic c&sc
NATO' in the running head.)

Any complementary glyphs that are used in math mode (and for upwards
compatibliity, that means all of cmmi, cmsy, msam and msbm) should be
in a math font.  If there aren't many other glyphs, we might as well
put them all in a math font and be done with it.  Messy, but probably
unavoidable because of memory limitations and upwards compatiblity.

>I wonder if there is some way to modify TeX so that it doesn't
>the characters on demand.

The problem with this is that it would require a disk access every
time the user requested a previously unseen character, which would
probably be unacceptably slow.