I thought that a leap year was every year that could be exactly divided by 4 (multiple of 4).
But I was doing some research of the Doomsday Algorithm in this page and learned that the definition of a leap year was:
This means that 2000 is a leap year, while 1900 (multiple of 4) is not.
Is this the real definition of a leap year?
Almost. CKDexterHaven says
Yes. See the U.S. Naval Observatory on leap years.
This point is evidently debatable. See http://world.std.com/~dpbsmith/leapyearfaq.txt, which attributes the rule only to the Encyclopedia Britannica, amidst several other sources that do not follow it. Neither the National Institute of Standards & Technology nor the Royal Observatory Greenwich mentions the 4000-year exception, although the Royal Observatory does acknowledge an error in the 400-year exception that “amounts to about 3 days in 10,000 years.”
I will let CKDexterHaven argue his own case on this, as I was merely quoting his article.
Since this is so directly addressed by a Staff Report, I’ll shoot the thread to that forum.
That way, Dex’ll see it and can address any apparent discrepancies.
The last time I studied this, I could not find that any official body had set the proposed 4000-year rule. (Believe it or not, though, I have heard an IT executive from the Federal Reserve complain that a certain piece of software didn’t do it.)
Mathematically, it really should have been a 2000-year rule, and 2000 should not have been a leap year. If the 4000-year rule goes into place, two leap years will have to be skipped around the year 20,000.
Note that this assumes that the lengths of the solar day and the tropical year will not change significantly.
I don’t remember where I took that from, but I did some rudimentary math along with it, and it seemed to work out OK.
Obviously, a day every 4,000 years is nothing to lose sleep about.
I found it very disappointing that 2000 was a leap year; it would have been so cool to live through a century year that wasn’t a leap year, and hear all the explanations blah blah. Sigh.
Why so pessimistic? Only 97 years 4 days until the next opportunity!
As I understand it, the old Julian calendar (the one the Staff Report says Julius Caeser implemented) had a leap year every four years. The Gregorian calendar had two of the three EXCEPTs listed - it does not take into account the 4000-year rule. This seems to agree with the USNO link that brianmelendez gave:
I remember reading, where I know not, that the Russians did not switch from the Julian calendar until the 20th century, whereas most other people who switched did so in the 16th or 18th centuries. Thus, by the time the Russians got around to it, the 4000-year anomaly was known, and they did not adopt the Gregorian calendar, but rather one with all three EXCEPTs.
At least you got to hear all the arguments about whether the millenium began on January 1st, 2000 or 2001. 
[sub]Psssst…it was 2001.[/sub]
It’s certainly possible that the Russians could have done that. I don’t know of any non-RC country that has ever accepted the Gregorian calendar; they’ve all come up instead with “New” or “Reformed” calendars that just happen, by an amazing coincidence to work out the same as the system introduced by the icky, evil Pope.
Well, somewhere around the year 802,701 AD when the Morlocks have taken over, their calendar will be off by more than six months! hah! We’ll have the last laugh then!
And that’s why some people think that the October Revolution occurred in November.