Not here at all. My clock is a strftime("%Y-%m-%d
%H:%M:%S"), so it's right
once every 68 or 136 years, given that Unixtime is the base.
Well, I've got a 64 bit system, which means that it's 4 billion times
time_t should still be only 32 bit. Currently.