[LEAPSECS] Leap smear

[Rob Seaman]

On Sep 19, 2011, at 9:35 AM, Clive D.W. Feather wrote:


> You are (or appear to be) trying to sneak in the definition that the civil day is the synodic day, and then use that to "prove" that your approach is correct.


Requirements are not specifications.  Requirements are inherent in the problem space and are not supplied by humans.  They are discovered by humans.  Specifications are indeed "specified" by humans as part of proposing a solution to the problem.  One problem.  Many possible solutions.

If you don't believe that civil time is time-of-day where "day" means "synodic day", then assert an alternate definition for what the word day means.  Then we can debate the two alternatives head-to-head independent of the complexity of how different deployed solutions will address the requirements.

The ITU proposal, rather, sweeps the issue under the rug.

Convince the mailing list that the meaning of "day" is a specification, not a requirement.  That humans can choose day to mean 86390 or 86410 SI-seconds.  Or even 86399 or 86401 SI-seconds.  Similarly, it cannot mean a constant 86400 SI-seconds.


> THAT IS NOT HOW IT WORKS IN THE REAL WORLD.


The real world has a way of asserting "how it works".

Have to redirect my attention back to the talk in progress.  Maybe more later.

Rob