In five days it'll be a month since I posted the TIODCDTIMESTAMP patch
to com.c. Bill Studenmund pointed out the patch would catch both edges
of the once-per-second, on-the second pulse. I've fixed that in my own tree.

Since no-one has objected or offreed comment, and since Charles Hannum
has never bothered to offer any comment whatsoever on the proposal to
add PPS support (as opposed to other ideas for supporting obsolete
clocks), I propose to commit the TIODCDTIMESTAMP patch once the month
is up.