On Fri, Jun 20, 2008 at 09:08:32PM -0600, Idris Samawi Hamid wrote:
> On Fri, 20 Jun 2008 20:51:16 -0600, Khaled Hosny <khaledhosny@eglug.org>  
> wrote:
> 
> >> > Good point. I'll try to find a conversion algorithm...
> >>
> >> http://www.codeproject.com/KB/datetime/hijrigregorianclass.aspx
> >
> > The guy here claims to have a better algorithm than others
> > http://ojuba.org/wiki/doku.php/hijra/, it is written completely in
> > python. I didn't compare it with others myself, though.
> 
> Well, since we're not using it for long-term historical ranges -- and  
> we'll need to be able to advance/reverse it a couple of days no matter the  
> algorithm, then I say let's go with whatever is easiest for hans to read  
> ;-)

I think Umm al-Qura algorithm is more tested and is the one officially
used in Saudi Arabia, but I can't find any technical details about it,
though Arabeyes's ITL implements it and is giving me correct results so
far (unlike the Hijra algorithm). EMACS one seems to be widely used too,
but its ITL implementation gives me incorrect dates some times

http://www.arabeyes.org/project.php?proj=ITL

Regards,
 Khaled

-- 
 Khaled Hosny
 Arabic localizer and member of Arabeyes.org team