I use the Merriam-Webster Online Dictionary all the time, but sometimes it lets me down.  Consider its article (or “usage note”) “Centuries and How to Refer to Them“.  The article is headed as follows:

Usage Notes
“Centuries and How to Refer to Them
“Is it the 1600s or the 16th century?

Right away we start worrying.  “The 1600s” and “the 16th century” are not alternatives for referring to any century.  The sixteen-hundreds were (except for 1600) the Seventeenth Century.  The article itself straightens that out, I’m glad to say.  But it gets worse:

Yep, that’s what’s coming next: the 22nd century. Its years will all* start with 21, proceeding up to the distant 2199.

Notice the asterisk in that fallacious claim?  It doesn’t link to anything (a bad mistake on a web page), but if you scroll down you’ll come to a note (almost as long as the article) in which the writer cravenly declines to take a position on the question, whether a century contains ninety-nine or a hundred years?

The note claims that the matter is unsettled, and cites the London Times as authority for the contention that a millennium has only 999 years:

The world has voted with its cheque book in the debate on precisely when the millennium ends. While pedants continue to pit December 31, 1999, against the end of the year 2000, everyone who is anyone, it seems, has opted for the earlier date as the time to organise what they hope will be the mother of all parties.
— The Times (London), 2 Apr. 1991

The London Times being a Murdoch rag, it naturally defers to the market on all matters, even on matters of objective fact.  If the market decides that 1000 = 999, then there is no more to be said.

Those of us who believe in counting, however, (we pedants) think different.

I’m not the first person to point out that, during the past century, “the world” (meaning, those of us who matter to the Times) grew accustomed to seeing odometers roll over, and to feeling that something momentous has happened when we see a string of zeroes.  Indeed, when the odometer turns over to 2000, our car has travelled two thousand miles.  On the other hand, we hadn’t completed the second millennium until the year-counter read 2001.  The odometer starts with zero, but the year-counter starts with one.  We count a mile when we finish it, but a year when we start it.

All this should be clear to the people at Merriam-Webster, and it troubles me that it’s not.