However @. is a bit buggy for PDF output: it seems that kerning is lost.
For instance, compare the space between the P and the period with:
@emph{OP}. Bar
@emph{OP}@. Bar
Thanks for the report. I'll see if I can fix it.
Best,
Karl
