Re: Implementation Question -- accents

At 6:00 PM 7/2/96, rminer@geom.umn.edu wrote:
>Bruce's proposal doesn't deal with accents, math or otherwise.
>
>... It seems to me that the standard
>ones (tilde, bar, vec, dot, double dot, hat, etc) can be handled by
>the current display list machinery with the (moverscript ... ) schema,
>provided we introduce a schema like
>
>        (maccent "accent_name")
>
>to serve as the argument, e.g.
>
>        (moverscript
>                (mi "M")
>                (maccent "tilde")
>        )

That's part of it that I didn't have time to explain in the letter
(or to finish working out in full detail), but the proposal is intended
to handle accents in a really simple way:

just use an overscript, and have special characters for the accents
themselves. So your example can look something like

        (moverscript
                (mi "M")
                (mo "˜")
        )

(though I'm not sure if it's sensible to specify the use of mo there).
If necessary, the ˜ can instead be some character specific to
the use as an accent.

Are there problems with this way of handling accents?

(Nico says

> I believe you have to differentiate between accents
> (up, down or middle) and overline/underline

but doesn't explain why. I don't know anything about "accents in the
middle" -- maybe this is part of the reason overscript/underscript
are not adequate?)

>P.S. Should the comma in SL(2,C) be in an (mt ...) schema or something else?

I think it should be (mo ",")
since comma is essentially a low-precedence infix operator.