On 2014-01-18 03:28 AM, Poul-Henning Kamp wrote:
I think it is cute you lay all these plans, but how are you going to
sell your new timescale ?
I'm certainly not going to do that alone. It will take a concerted
effort by a lot of people with more credibility in the field than I.
I think its going to take some sort of due-process proposal from some
internationally recognized standards body. The AAS looks like a good
place because there are many experts and it seems there is some
political and personal connections to ITU and other bodies.
[ I figure, well, it was the astronomers that got us into the calendar
mess to begin with, so its up to them to fix it :-) ]
I'm new to this list, but I see a collection of people that have lived
and breathed this topic for years. The many excellent articles and
presentations at the various AAS conferences display a level of
understanding and expertise unmatched in any other forum I know of.
LEAP_SECS list provides a unique forum for discussion. It might provide
a good place to hash out some informal agreement. If that happened,
maybe that could move to the AAS (and/or, perhaps some other body) for a
more formal due-process, and, from there, maybe, to ITU or other
appropriate body for consideration. I think other people on this list
know better how that political process might need to work.
As I understand it, "kill Leap Seconds" will be reconsidered in 2015.
Can an agreement like this be completed in a year to leave time for the
ITU to digest it?
I'm a little encouraged that my initial suggestion found some response.
I knew what I first proposed was incomplete and vulnerable to
misinterpretation. And sure enough, even amongst experts there's a
complete disconnect.
The discussion I'm having with Andew illustrates one part of the
challenge - that there is a lack of common understanding of the terms
and definitions. "One man's timescale is another man's epoch". By
casually outlining some features of my suggested "timescale" (CCT) I
opened the can of worms.
This lack of common terminology pervades the field of time-keeping. I've
been frustrated by it for at least a decade in standards body
discussions. And here it is again. The terms are the result of the
tortured history of time-keeping going back to (at least) Stonehenge. [
You see why I blame the astronomers? :-) ]
Any good standard is going to include a "terms and definitions" section.
This is a challenge in this field because of the history. But I see it
as central, critical, to get that part right if possible, because
without it the overall design, details, and subtleties of any
"timescale" or "time-keeping framework" will be misinterpreted.
How will you get EU to change UTC to $whatever in all their regulations ?
I'd hope the design of a proposal would not require them to do that
until they wanted to.
If you can honestly tell them "It's just a renaming, there is no semantic
difference", you *might* be able to persuade them to do it under an
administrative ruling by the uanimous commisioners.
Well, yeah, that's part of the idea. The standards that define TAI and
UTC are fractured. By well defining terms and consolidating the
description you go a long way toward unifying the understanding.
Further, there are missing pieces to the puzzle in the standards,
especially where "computer time-keeping" in concerned. A standard or
recommendation could include clear descriptions of how this should be done.
The idea is that the portion of the new "timescale" (maybe its a
"framework" or some other term) from 1972-01-01T00:00:00Z (UTC) onward
would map directly to the current standards of TAI/UTC. That way anybody
could adopt the new "standard" in their own time - the definitions of
TAI/UTC remain in force.
Its a "template" that overlays existing standards, clarifying terms and
formulas. It could also, perhaps, define refinements that improve
accuracy, but that should be done as *additions* to the underlying
TAI/UTC standards so as not to invalidate them somehow.
I belive there have been several hundred attempts at these and only
a handfull of successes during the lifetime of EU.
Yeah, its a long shot. But if something is not done the field will be a
mess forever.
But as far as I can see, your proposed timescale will not even
allow you contemplate that route.
Since there are semantic changes for the past, it will not apply
to the past in existing regulation under any circumstances ("no
post hoc legislation") so bothering about the definition for the
past is mostly a waste of your time.
As I've tried to explain, I'm not suggesting any changes to the past.
The past is past.
For the future, a directive originating in the Commission and
approved in the European Parliament would be able to say:
"Starting YYYY-MM-DD, in all EU documents for 'UTC' read 'CCT'"
Only taking many, many more words to do so.
The task will be slightly complicated by the fact that EU uses
a number of different words for "UTC" already, amongst these
"GMT", "Weltzeit" and so on.
The directive once approved, would then go to national legislatures
for implementation, where somebody will have to go through all laws
and come up with the necessary ammendments for parlimentary approval.
It's not an impossible process to push through, we do it all the
time, but there needs to a reason for people to spend the effort.
Ohh, and do mind the political minefields.
No way I'm qualified to even begin to anticipate the international
complexity. I speak English and c/c++.
For instance I doubt you'll find any UK politician willing to push
a s/GMT/$whatever/ legislation since that will just feed the UKIP
trolls and become a factor in the Scottish independence referendum.
So exactly how do you propose to sell this idea ?
I'd need all your help.
Remember, the your "competition", changing the definition of UTC
without retaining the name does not require any laws to be changed.
That's the idea, I think.
-Brooks
_______________________________________________
LEAPSECS mailing list
LEAPSECS@leapsecond.com
http://six.pairlist.net/mailman/listinfo/leapsecs
