Dear Frank,

I well understand your arguments, but there are also very strong
arguments against leap seconds. So, one has to decide which
applications are more important: the many technical applications
which would go better without leap seconds, or applications like
sundials or traditional navigation which are the fields of
enthusiasts.

> Well all diallists and anyone who uses astronomical tables
> care about UTC precisely because it is, currently, guaranteed
> the same as UT1 to within 0.9s and we can ignore that difference.
> 
> When UTC-UT1 is even a few seconds, never mind a large fraction
> of an hour, we won't be able to ignore the difference.

This is true. However, to get local time (true solar time), one
has to take into account much more than UTC-UT1. In the future,
one would just have to add one additional simple calculation.
UTC-UT1 will be more easily available in the future than it is
now, and if one needs it only with an error of about one second,
this can be predicted and tabulated in astronomical almanacs.
Just one additional simple table ... The difficulties with leap
seconds are much more complicated.

> I'll go further.  Everyone who ever sets a clock or a watch uses
> UTC or some very simple offset from UTC (usually an integral
> number of hours).  There are quite a lot of people who own watches!

True, but this is not the point. Why should UTC be close to UT1?
There is no reason besides the non-scientific astronomical applications
you mentioned. From a scientific/aesthetic point of view it would
be even better to use UT1 when one needs solar time instead of UTC
as an approximation.

> Moreover, most e-mail headers, certainly including yours, implicitly
> refer to UTC.  Your header said...
> 
>     Date: Tue, 12 Jul 2005 08:10:15 +0200 (MEST)
> 
> That +0200 refers to the offset from UTC.

It will be the same without leap seconds. The only difference
will be that it will no longer be also the offset from UT1.
So, I should have asked: Who cares about UT1?

Best regards,
Wolfgang
-

Reply via email to