Re: glib time functions

[Robert Brady <robert suse co uk>]

On Thu, 30 Nov 2000, Ali Abdin wrote:

> No I couldn't.

Exactly. Makeing a generic interface for this will be hard.

datetime_to_islamic_date(&datetime, &islamic_date);

islamic_data = 
  { { G_DATE_PROBABLY,
      { G_Ramadan, 1, 1000 }
    },
    { G_DATE_PERHAPS,
      { G_Ramadan, 2, 1000 }
    }
    { G_DATE_ONCE_IN_A_BLUE_MOON,
      { G_Ramadan, 3, 1000 }
    }
    { G_DATE_IF_THERES_JUST_BEEN_A_BIG_VOLCANIC_ERUPTION,
      { G_Ramadan, 10, 1000 }
    }
  };

?
   

> But this "date" issue has been happening ever since 1421 AD. I was surprised
> that the algorithm I have implemented didn't take in the "off by one" errors
> that could happen each year, this tells me there is something "funny" going on
> wiht the Islamic calendar, that makes this calendar somewhat accurate.

Yeah, it always keep sync with years, it just drifts back and forth a day 
or two...

One other problem, is that until the algorithm's data table gets updated
whenever the moon is seen, this breaks for the current month.

Like timezones, but worse.

-- 
Robert Brady
robert suse co uk