[LEAPSECS] Historican timezones (Was: Re: Looking-glass, through)

[Magnus Danielson]

On 13/01/11 00:59, Warner Losh wrote:

> On 01/12/2011 15:41, Paul Sheer wrote:

>> The Olson time library has all historican timezone information.

>>

>> You can choose a zone, a time (back to the 1800's) and deduce

>> the precise zone offset in minutes and seconds.

>>

>> This is inclusive of wierd time zones that were based on the hour

>> since sunrise (or some similar wierd thing), as well as more basic

>> things like daylight savings that were tried one year, and revoked

>> the following year by that countries government.

>>

>> There are 3064 rules in the Olson time library zone files and 434

>> time zones, as well as GPS information for each zone.

>>

>> It would seem a number of maintainers do track time zone

>> announcements from all over the world and keep these

>> rules up to date. The most recent update was November 2010.

>>

>> Further, people updating the zone rules usually provide references

>> for the rule change, citing a government announcement.

>

> The Olson database is good for things since the 80's or 90's. Prior to

> that, the further that you go back in the past, the less reliable and

> complete the information gets. Since then the information has been

> updated in real time. Prior to that it was based on research for

> different areas and some of that data is still extant, while other data

> might not be.

>

> Don't get me wrong, or take this as me saying that the maintainers of

> this database haven't done a herculean job of making it the best we

> have. The problem is that it is hard to know for sure all the details

> you need to for historical times the further back into the past you get.

>

> There's even a nice disclaimer:

>

> # This data is by no means authoritative;....

> ...

> # I invented the abbreviations marked '*' in the following table

>

> etc


The Wikipedia article points out the historical examples when thing 
where different which the Olsen database does not handle, relating to 
the rules being chosen. In essence it focuses on UNIX time-life first 
and historical changes prior to that when possible. Again, this is not 
to discredit the database and the efforts to maintain it, but it is an 
effect of the prioritisations made in order to focus work on the primary 
use, namely recent and current years.

Cheers,
Magnus