[TZInfo-users] one more...

Philip Ross phil.ross at gmail.com
Mon Oct 22 18:17:46 EDT 2007


On 22/10/2007, Sanjay Vakil <svakil at gmail.com> wrote:
> It appears, in looking at the zone data, that there are multiple
> "city" entries in what much of the world considers a single timezone.
>
> I'm assuming this is for historical reasons -- Detroit and New York
> were not always in the same timezone.

Yes, your assumption is correct. Detroit and New York have been
different in the past, but since 27 April 1975 have used identical
rules.

The source tz database (http://www.twinsun.com/tz/tz-link.htm) lists
all the timezones available for each country rather than just those
with different current and future rules.

> Is there a way to get a list of just those timezones which are
> actually different going forward?

There isn't a simple way to do this with TZInfo - you'd have to
compare each future TimezonePeriod for each pair of Timezones to
determine if they were the same. If they were the same, you'd need a
rule of some sort to determine which zone you will use, so this might
require some manual intervention.

> I guess this relates back to my earlier question about showing PST/EST/MST/CST.

For the US, there are a set of linked timezones defining the distinct
set of rules that are in force. These are:

US\Alaska
US\Aleutian
US\Arizona
US\Central
US\East - Indiana
US\Eastern
US\Hawaii
US\Indiana - Starke
US\Michigan
US\Mountain
US\Pacific
US\Samoa

Phil

-- 
Phil Ross
http://tzinfo.rubyforge.org/ -- DST-aware timezone library for Ruby


More information about the TZInfo-users mailing list