<div dir="ltr"><div><div><div>> The tz database has always eschewed sub-second offsets.<br><br></div><div>This was a result of the early work being focused on supporting non-sub-second functions such as "localtime;"<br></div><div>absence of sub-second timestamps in early versions of UNIX also played a role.<br></div><br></div>We might expect TZDBv6 to have sub-second offsets, as well as Julian calendar support, Persian calendar support, and noon typing (solar, local mean, civil:-)<br><br></div> @dashdashado<br></div><div class="gmail_extra"><br><div class="gmail_quote">On Wed, Nov 5, 2014 at 10:47 AM, Steve Allen <span dir="ltr"><<a href="mailto:sla@ucolick.org" target="_blank">sla@ucolick.org</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">From their inception the International Time Bureau (BIH) produced<br>
records of the deviations between various time sources. I have a<br>
sampling of those BIH publications from 1960 in images here<br>
<a href="https://plus.google.com/photos/112320138481375234766/albums/6078225731350227361" target="_blank">https://plus.google.com/photos/112320138481375234766/albums/6078225731350227361</a><br>
<br>
1960 is after the availability of cesium atomic chronometers and<br>
during the era when Heure Definitive was UT2. These pages include<br>
part of the initial attempt to "coordinate" radio broadcasts between<br>
the US NBS station WWV and the UK NPL station MSF. They include time<br>
offset plots for various observatories for the entire year of 1960,<br>
and a couple of pages of the radio broadcast offsets for the end of<br>
the year. Also note that these data were not published until as much<br>
as a year after the measurements, and only then could one know what<br>
time it should have been.<br>
<br>
>From the turn of the 20th century it was agreed that everyone would<br>
use the same the technical basis for determining civil time, but the<br>
quality of time available in any particular place depended on<br>
available resources and training, so the implementation was imperfect.<br>
In many countries the observatory time was the legal time, and the<br>
broadcast time signals were, practically, the official time.<br>
Plowing through these historic records of time offsets only makes<br>
sense for projects such as re-reducing occultation data.<br>
<br>
The tz database has always eschewed sub-second offsets.<br>
The data on these pages show part of why that is a good thing.<br>
<br>
--<br>
Steve Allen <<a href="mailto:sla@ucolick.org">sla@ucolick.org</a>> WGS-84 (GPS)<br>
UCO/Lick Observatory--ISB Natural Sciences II, Room 165 Lat +36.99855<br>
1156 High Street Voice: <a href="tel:%2B1%20831%20459%203046" value="+18314593046">+1 831 459 3046</a> Lng -122.06015<br>
Santa Cruz, CA 95064 <a href="http://www.ucolick.org/~sla/" target="_blank">http://www.ucolick.org/~sla/</a> Hgt +250 m<br>
</blockquote></div><br></div>