I'm forwarding this message from Stephen Colebourne, who is not on the time zone mailing list. Those of you who are on the list, please direct replies appropriately. --ado -----Original Message----- From: Stephen Colebourne [mailto:scolebourne@joda.org] Sent: Monday, March 17, 2008 10:26 To: Olson, Arthur David (NIH/NCI) [E] Subject: Re: FW: Sryia
-----Original Message----- From: Stephen Colebourne [mailto:scolebourne@joda.org] For everyone's info, I saw an IATA time zone change for Sryia for this month (March 2008) in the last day or so. I don't have it to hand so sadly I can't confirm the details right now.
This is the data IATA are now using: Country Time Standard --- DST Start --- --- DST End --- DST Name Zone Variation Time Date Time Date Variation Syrian Arab Republic SY +0200 2200 03APR08 2100 30SEP08 +0300 2200 02APR09 2100 30SEP09 +0300 2200 01APR10 2100 30SEP10 +0300 Stephen
Country Time Standard --- DST Start --- --- DST End --- DST Name Zone Variation Time Date Time Date Variation Syrian Arab Republic SY +0200 2200 03APR08 2100 30SEP08 +0300 2200 02APR09 2100 30SEP09 +0300 2200 01APR10 2100 30SEP10 +0300
Here's a link to English-language coverage by the Syrian Arab News Agency (SANA)... http://www.sana.sy/eng/21/2008/03/11/165173.htm ...which reads (in part) "The Cabinet approved the suggestion of the Ministry of Electricity to begin daylight savings time on Friday April 4th, advancing clocks one hour ahead on midnight of Thursday April 3rd." Since Syria is two hours east of UTC, the 2200 and 2100 transition times shown above match up with midnight in Syria.
Between 1999 and 2006 Syria always changed to DST on April 1. In 2007 the change was at the start of Friday, March 30; this year it's set for the start of Friday, April 4. Absent better information, my guess is that the transition now happens at the start of the Friday that's nearest to April 1. Back in 2004 we changed the time zone compiler so that when presented with a rule such as... # Rule NAME FROM TO TYPE IN ON AT SAVE LETTER/S Rule Syria 2007 max - Mar Fri>=29 0:00 1:00 S ...the compiler creates a transition for March 29, March 30, March 31, April 1, April 2, April 3, or April 4 (it's prepared to roll over into the next month for the transition). Such rules break older versions zic, but it's a noisy break: you get a message such as... "asia", line 1392: no day in month matches rule... ...and the compiler gives up the ghost. Can we now comfortably use constructs such as "Fri>=29" in database files? --ado
participants (1)
-
Olson, Arthur David (NIH/NCI) [E]