[LEAPSECS] All of this has happened before

[Rob Seaman]

We each wear multiple hats.  Two of mine are 1) to point out that physical reality trumps standards and software, and that 2) that there are precious few conversations here that haven’t occurred before:

  - Warner has explained his use case in the past.  There are likely no engineering use cases more varied than those related to timekeeping.  He’s simply asking for more notice of future leap second scheduling.  It appears possible now to lengthen this significantly beyond 6 months.  The standard in force would permit this.  (Although as Warner points out, LORAN has shut down since last we talked about it.)

(Whereas the “programmers stink so we need to dumb down the standards” thread has been flayed alive on the dissection table ;-)

We should be pleased to see lively debate continuing in this group after 15 years.  But the ancient wisdom of Usenet holds - better discussions result from reviewing the talking points contained in prior threads.  We don’t have just a few threads available to us on leapsecs, we have the Bayeux Tapestry: 

	https://pairlist6.pair.net/mailman/listinfo/leapsecs  (since 2007)
	http://www.ucolick.org/~sla/navyls/ (2000-2007)

In which you will find words similar to:

	Leapseconds are a means to an end.  It is simply true that Mean Solar Time and Atomic Time are two different things.  Civil timekeeping use cases descend from both, but especially from the fact that day means “synodic day” and thus pertains to solar time.  Seeking to pretend otherwise will inevitably cause engineering requirements to rear upwards and demand attention.  Systems engineering should be dealt with up front to avoid risks and repercussions (and unnecessary expense).

Rob
—
http://youtu.be/6bOy3RNyWME