[LEAPSECS] BBC radio Crowd Science

[Steve Summit]

Brooks Harris wrote:
> Yes, there is no uninterrupted incrementing numbering that represents UTC.
> Indeed, that's the whole point, isn't it?

Right.  This is one of the fundamental aspects of the whole
situation.  *Anything* you do that tries to work with UTC as if
it were, or tries to map UTC onto, a single, monotonic count of
seconds is suspect.  If you are using a monotonic count of seconds,
there's already a name for it, and that name is: TAI.  (And if
despite that, you're still trying to work with a *UTC* count of
monotonic seconds, there's a name for that, too, and it is: time_t.
And we know how well that works at handling leap seconds.)