[LEAPSECS] Mean ... Orbits

[Rob Seaman]

Steve Allen wrote:


> The difference between now and then is that now we do have enough

> telecommunications and compute power to broadcast completely uniform

> time signals while still keeping something that tracks UT in a

> smoothed sense.  If we divorce the broadcast time signals from UTC

> then the kernel hackers can be instantly happy with their uniform

> underlying time and we can, at our leisure, decide just how

> predictably smooth UTC has to be to suit all applications.


You may want to link to one or more of the prior messages where you described your notion of how this would work.

I think it is worth pointing out that this infrastructure is independent of any changes to UTC.  What it does rely on is the continuing recognition that our fundamental interval timescale (currently TAI) and our fundamental Earth Orientation timescale (currently UTC) are separate parts (with separate names!) of the same system.

For instance, Steve's suggestion could be implemented in parallel to changes to increase the scheduling horizon of leap seconds.  In fact, these could be coordinated (but loosely) parts of the same infrastructure improvement project and with a cautious roll-out as advised by Warner.

Rob