[LEAPSECS] content of the NTP-aimed leap-seconds.list files

[Zefram]

Rob Seaman wrote:
>There was a missing perl JD module from the host I first tried.

"sudo cpan Date::JD" to install it, and similarly for any others you're
missing or need to upgrade.  I didn't restrict myself to using modules
that are bundled with the Perl core.

>Does not address DUT1.

Indeed.  What needs do you perceive here?  Do we need a table of
historical values for the decisecond-rounded projected DUT1 that's meant
to go into broadcasts?  What about historical tables of what DUT1 values
particular stations actually broadcast?

I'm having difficulty imagining use cases for historical projections.
There's a need for historical measurements of UT1-UTC, and for current
UT1-UTC projections, and for future schedules of the decisecond-rounded
DUT1 for broadcast.  The latter seems quite a separate need from the
others.  Unrounded UT1-UTC (measured or projected) is quite unlike TAI-UTC
in that there's no quantisation on either axis, and that projections
get replaced.  The rounded-for-broadcast DUT1 has quantisation on both
axes, like TAI-UTC, but the relative lack of use for historical values
is quite different.

>Would require an externally authenticated secure channel.

I imagine a GPG signature going into a separate file, as the first
implementation of signing for this data.  But there's a fair bit of
value even without that: just having a .lmtr file retrievable from IERS
over HTTP (with an expectation that it'll be kept up to date) would be
a significant advance on the status quo.

>No objection to naming this after Msgr. Lemaitre, but he had little
>direct connection with the issue, having died 6 years before UTC year
>0 ;-)

Sure.  I did toy with naming it for Louis Essen, but figured that
the choice of any namesake who actually was directly involved would
be politically loaded (for the narrow context of the debate over leap
seconds).  It's like the rules about naming asteroids for politicians:
they have to have been dead long enough to be disconnected from current
tensions.

-zefram