> is that really going to work out? Somebody must have thought so:
TIME(2) TIME(2)
NAME
time, nsec - time in seconds and nanoseconds since epoch
...
Times from time should be stored in and treated as ulongs;
this extends the range of valid times into the year 2106.
