[LEAPSECS] Leap smear

[Rob Seaman]

Recent arguments are getting even more surreal than usual.

On Sep 20, 2011, at 9:44 AM, Tony Finch wrote:


> This is why daylight saving time exists, because astronomical time isn't convenient.


The whole point of DST is an adjustment from a norm.  If there was no norm there would be no "fall back" for each "spring forward".  That norm is mean solar time.  The whole thrust of the ITU proposal is that TAI is "close enough" to mean solar time.  If this is true, then it should be trivial to expand the proposal to discuss the implications of what "close enough" means.  There are indeed implications that astronomers would have to accommodate.

Folks, google "use cases" and "engineering requirements".  Maybe "system engineering management plan".  Compare and contrast accuracy and precision.  If mean solar time is not relevant, what underlying cadence is?  Ask yourself if your position is really that clocks could tick at any randomly chosen rate picked out of a hat?

Back to listening to talks on explosive celestial transient events.  Things do go bump in the night, whether convenient or not.  Building a system that seeks to tweak reality requires that reality be understood and modeled in some fashion in your system.

Forget about leap seconds and UTC.  The ITU proposal is a bad example of engineering.  It could and should be improved.

Rob