Re: [hrpc] I-D Action: draft-irtf-hrpc-political-05.txt

[Eliot Lear <[e-mail]>]

Hi!

This is mostly geeky entertainment.

[Ted, I saw your note after putting most of this together.  Time has been explored… ad infinitum :-), but see below for some entertainment.]

> On 23 Sep 2019, at 11:44, Niels ten Oever <[e-mail]> wrote:
> 
> Time, and its standardization, is a very political project. Going back to Aristotle (Physics iv 10-14) there is the difference between Chronos and Kairos, where Chronos became the number to measure motion (kinesis). There have been many other perceptions and measurements of time, that were able to exist next to each other, but as part of the colonial project, the European concept of what time was, was made the universal standard. That is why Greenwich Mean Time (in Great Britain) is used as a standard for the whole world.


But interestingly enough the initial reason for standardization of time in Britain was not political, but a simple matter of practicality; and it wasn’t the government who standardized time per-se, but rather the railways in 1847, in order to manage train schedules, and this was known as Railway Time.  It wasn’t until 1880 that GMT was officially standardized across the island of Great Britain.  GMT was set as an international standard in 1925 as a practical way for astronomers to agree on when the day began.

This gives you some idea as to just how recent in our history there was any sort of standardization of what the time is at any given moment.  However, the much of the actual units of time that we use today date back to at least 1500 BC and the Egyptians.

As Ted points out, NTP and the TZ database are based on the Gregorian calendar scale.  This does indeed prove troublesome for some countries and cultures who are on a lunar calendar.  This was noted in the development of the calendaring standard, where a “calscale” can be specified, but to my knowledge nobody has ever used anything other than Gregorian.*  The folks Calconnect could correct me, tho.

> 
> There are also many stories about the development of timezones, and their consequences, the way in which time helped shape labor, life, and how it altered complete societies.


Time?  Sure, but what part of standardization?  Time zones?  Not as much.

The first official time zone difference was used by New Zealand in 1868 as a convenience to know what time it was in the home office, as it were.  This before there was a legal standard on Great Britain.

And then there are DST transitions.  These were introduced by Germany in 1916.  With Europe about to make a remarkable leap  (so to speak) into the unknown, nobody can possibly say that there isn’t a political influence on time zones.**  The political impact?  People debate them, so I suppose there must be some.

There are some interesting stories.  One involves a contract with a satellite operator with the Holy See in which the contract was written in local time in Brazil.  Apparently they booked the wrong hour, and so a law was passed to change time.  This is actually documented in the TZ file for South America.

> 
> I am not making a value judgment about this re-odering, but I am trying to show that NTP, which helps synchronize to a specific time regime, on which a lot of computing depends, helps promulgate a specific, and political, ordering.


Let us be precise:  NTP requires only two protocol invariants: the length of a second itself and a particular epoch.  The one field that has any real political aspect at this point is the leap second indicator.  There have been some, dare I say, highly entertaining debates in the ITU-R over leap seconds.  Indeed it’s those pesky astronomers who are once again concerned about dinking with leap seconds.[1]  And let’s not even get into whether Pluto is a planet!  Still, the protocol itself is neutral as to how and when the leap second bit is set, or who has the authority to set it.

Eliot

[*]  I gave at tictoc circa 2010 on a Brief History of Time(Zones) that shows Israel spiking; that’s because of a combination of time scale, and issues relating to the High Holidays.  This has mostly been kinked out, but not entirely.  If a transition would occur on Rosh Hashanah, it is moved to the 1st of October.  This cannot be accurately reflected in the TZ database in a programmatic fashion.
[**]. No, this is not an invitation to debate Europe’s decision.  Take it to facebook! ;-)
[***] If you would like references for the above, I highly recommend A Brief History of Time, and Wikipedia now does a very decent job of it.


[1] https://www.abc.net.au/science/articles/2005/09/23/1466849.htm