Poul-Henning Kamp wrote (among other things, including doubting the sobriety of Europeans):
I cannot test leapsecond handling once per month.
A variety of proposals have been discussed over the past five years. Somebody even made the suggestion that a positive or negative leap second be issued every day. In that case an adjustment would be made through the omission of a leap second. There are any number of possibilities for improving (or, at least, changing) the underlying technologies. The facts are two-fold:
1) Interested parties such as on this mailing list are left to talk to themselves. Nobody on the inside of this rather fundamental decision is apparently seeking to open the process up.
2) The technical details of improving the distribution of global time signals has been deemed irrelevant to this decision. Rather, as PHK continues to suggest, the only proposal on the table keys on tricking a planet-full of blissfully ignorant people into disconnecting their clocks from the rotation of the Earth.
There are two main classes of time in general use, interval time (e.g., TAI) and time-of-day (UTC). Both are needed for a variety of purposes. UTC is one way to capture both in a single civil time system. Perhaps there are others. I personally happen to think that time-of-day is logically and legally and historically the natural basis for civil time. Perhaps you disagree. But I am completely unmoved by arguments based on the incompetence of our fellow programmers or on the pseudo-requirements of badly conceived engineering projects.
By all means, discuss the future of Coordinated Universal Time, International Atomic Time and worldwide civil time. But carry on the discussions publicly, invite interested parties from outside the narrow confines of the precision timing community, and focus the discussions on improving services, not castrating them.
Stop looking for a backdoor way to make this decision.
Rob Seaman National Optical Astronomy Observatory
