[tz] Timezones Morocco 2013 incorrect

[Patrice Scattolin]

Keep in mind that new moon crescent visibility is not an accurate 
science. Essentially it depends of the observer' skills and his location 
and atmospheric conditions. Usually a new moon of less than 24h of age 
is considered very difficult and essentially not observable although the 
record is with younger moons http://aa.usno.navy.mil/faq/docs/crescent.php

Since we have no idea of emacs' algorithm or how well it correlate with 
the observer's abilities (and it's possible the Saudi Arabian observer 
not be able to see it while a Marocain one could dues to the difference 
in longitude) I don't think that there are really reliable source other 
than government policy. The variety of sources Wikipedia cites for the 
start of Ramadan should warn us to be cautious 
http://en.wikipedia.org/wiki/Ramadan_%28calendar_month%29

For the record, anyone who would want to do to the math themselves is 
encouraged to look at Astronomical Algorithms 
http://www.amazon.com/Astronomical-Algorithms-Jean-Meeus/dp/0943396611 
which is essentially the base book giving all the formulas used by many 
astronomy programs. It's more work than it looks but for predicting the 
moon's age you hopefully won't have to get to more than the first 1/3 of 
the book.


On 05/03/2013 10:09 AM, Brian Inglis wrote:
> On 2013-03-04 02:18, Paul Eggert wrote:
>> On 03/04/2013 12:52 AM, Erik Homoet wrote:
>>> An official link for the announcement for this year I cannot find on
>>> official local websites but on all timezone and worldtime websites and
>>> airlines booking engines are using it.
>>
>> Thanks.  Last year it looks like the official announcement
>> by the Head of Government of the Kingdom of Morocco wasn't until
>> July 11 (for July 20 and August 20 transitions), so if we wait for
>> something official it'll be uncomfortably close
>> to the actual event.  Here's last year's announcement (in French):
>>
>> http://www.mmsp.gov.ma/fr/actualites.aspx?id=288
>>
>> I'm thinking of using the GNU Emacs predictions for Ramadan for 
>> 2013-2037,
>> whatever they are, and slapping them into the Morocco table by hand,
>> as predictions.  Any better suggestions are welcome.
>
> GNU Emacs predictions may be usable if they take the location 
> dependent visibility of the new moon crescent into account (see 
> references to Yallop's method) but if this becomes more widespread, it 
> may be better to develop (or borrow) a script that will predict the 
> start and end of Ramadan for any tz location.