RE: Islamic Calendar Stuff

[Ali Abdin <aliabdin aucegypt edu>]

>The problem, as I see it, is that a calendar system which is actually based 
on observation, rather than algorithmic or astronomical calculation is almost 
inherently non-computer friendly.  Are there *any* applications which 
"correctly" handle this calendar system?  If so, I suggest you look at them 
and try to figure out how.

No - there isn't. The way microsoft handles it is the way that i suggested it 
ship with the current date and as you know the user can double click the time 
part and set his own 'current date'. Wait - i think Microsoft doesn't do it 
right - according to my memory you can set the ENGLISH date but can't set the 
islamic date - hmmmmmm

>Manually syncing things doesn't seem to be a particularly easy thing to have 
to remember to do, but it may be the best answer.  As you note, it makes 
working in the past or the future quite complicated.  (Side note: I don't 
understand why you only
>need to keep two months of past date counts...)

All the months are 30 days. There are two months where the date changes - 
Ramadan, and the current islamic month (the one that has eid al-adha). These 
two months are when they keep track of 'observing the moon' and enforce it 
(the other months they just let it have 30 days without checking moon 
observance). THis is for religious reasons. Hence, these two months are 
'dynamic' and can change from year to year (they can be 30 or 29 days - i 
think they could also be 31 days - not sure). so you keep track of those two 
months, and assume all the other months are 30 days. That is if you want an 
/accurate/ past. Of course it'll become slightly inaccurate as soon as the 
future occurs (i.e. the user has to adjust the date). At this point in time 
i'm inclined to say "screw the past" - it doesn't matter if its inaccurate or 
not (i.e. you're not gonna schedule something in the past are you?)

I guess my solution is the best since this calendar system is just not 
computer-friendly (heh - thats why they rarely publish islamic calendars - cos 
it could be inaccurate)

>Finally, even if you could figure out an astronomical algorithm for computing 
the probability of new moon sighting at a particular place, aren't you still 
just guessing?  Unless you can find a probability for cloudiness?
 the Subject.

True - but by using that astronomical algorithm thingy you get INCREASED 
accuracy on the calendar. and that mooncalc program is i guess how they do it.

You could of course be User-Interface friendly with respects to gnomecal and 
inform people "this date may be <X> days inaccurate" or whatever (inspired by 
a website that does islamic-greg conversions)