[texhax] Correcting for math-mode kerning before commas and periods?

[Uwe Lück]

At 16:58 16.10.05, Oleg Katsitadze wrote:
>On Sun, Oct 16, 2005 at 02:34:14PM +0200, Uwe L?ck wrote:
> > (i) Maybe $f.$ is the solution!? If you try it, it doesn't conform
> > to that tradition as well as $f$.
>
>The problem with this is that the following space is not an
>inter-sentence space, it is just an interword space.

Indeed, well observed. However, this would not make a difference
with languages like French and German where usually
\frenchspacing is used.

> > (ii) When a sentence ends on a displayed equation, you cannot
> > avoid putting a dot at its end -- inside the math environment.
> > Now, on screen $f.$ and $$f.$$ seemed to differ with respect
> > to the distance between f and dot, but \showlists shows that
> > they are the same.
>
>They are exactly the same for me, under plain TeX (I was
>measuring at high zoom in xdvi).

They are exactly the same indeed, as I see with \showlists
and $f.$ vs $\displaystyle f.$.

> > (Or do something like $$f\,.$$)
>
>It would not be consistent -- the gap in $$f\,.$$ is bigger
>than in $f$.

You are right -- obsolete after the former observation, anyway.

> > However, `We arrive at $f=0.1.$'  isn't nice (I guess this is
> > the reason for that tradition).
>
>But again, I get the same result from both $f=0.1.$ and
>$f=0.1$. -- there is no italics correction for the roman
>digits.

You are right, confirmed it with \showlists (and maybe
Appendix G tells so.)

Best,
   Uwe Lueck.