In article ,
Jeff Liebermann wrote:

On Sun, 16 Aug 2009 17:34:37 -0700, David Nebenzahl
wrote:

On 8/16/2009 2:51 PM Meat Plow spake thus:

I agree that for most a minute per month is reasonable but I would
expect the same accuracy as my $29.99 Timex wris****ch which is more
like a second a month.

So that kinda begs the question of why computer mfrs. can't (or won't)
include clocks that are at *least* as accurate as a Timex, no? Wouldn't
a computah be a more compelling reason for a more accurate clock? (I
know, $$$ bottom line, right?)

Because it's difficult. The right way to have done it would have been
to do a function call from an RTC (real time clock) every time some
application needs the actual time.

I don't agree. NO CLOCK, running alone, can be really accurate over the
long term. A much better way is to take the output from a crummy,
inaccurate *but low cost* clock and using an external time reference,
synthesize from it a local clock of simply amazing accuracy.

NTP solves the problem completely, and at a very low cost (processing
cycles instead of expen$ive hardware). NTP works even if the computer
it's running on has *no RTC* (in the hardware sense) at all. All it
needs is some sort of interrupt generated every N cycles of the
processor clock (N is any integer that produces regular interrupts a few
times a second; the actual interval is not important).

Isaac