Re: [Groff] Integral signs

[Miklos Somogyi]

Joerg,

This only needs a bit of compression, not to overwhelm an equation.
It is good looking in print and it is simple,  thank you.

Since PS was incorporated into troff (a very long time), I completely 
forgot
about pic, and was wondering why anyone wanted to use it.
Now I need to reconsider, pic can still be rather useful.

Thanks again,

Miklos



On 28/02/2007, at 7:06 PM, Joerg van den Hoff wrote:

> On Wed, Feb 28, 2007 at 01:42:38PM +1100, Miklos Somogyi wrote:
> > Dear Folks,
> >
> > I need some integral signs with circles/ellipses around them (e.g. for
> > control volumes).
> > I've experimented a lot but only got something that is just acceptable
> > in print,
> > but downright ughhly in Preview and Acrobat.
> >
> > How would you stretch the circle of the double- and triple integrals?
> > I've tried \(ci  instead of `o' but it looks very thin compared to
> > thick int signs.
> > What can you suggest (espec something scalable)?
> >
> > Thanks,
> >
> > Miklos
>
> I'm sure there is some low level way to do this, otherwise `pic' won't 
> work
> anyway, but maybe you can use some ugly construct like:
>
> .EQ
> delim $$
> .EN
> .PS
> .ps  24
> ps = 24
> eh = ps / 140
> ew = 4 * eh
> bw = 10 * eh
> ellipse height eh width ew "$int int int$"
> box invis width bw with .w at last ellipse.e "$f(x, y, z) dV$"
> .PE
>
> (and shoving this into a macro might even preserve readability of the 
> document
> source ...)
>
> this would work only for display equations, not for inline equations. 
> I'm not
> sure if one can use troff registers within `pic'. if so, one could 
> avoid to
> dublicate the point size definition (needed for the empirical 
> "scalability"),
> but `ps = \n[.s]' does not work, it seems.
>
> HTH
>
> joerg