[TRLog] Computer clocks

Paul Kirley pkirley@neonramp.com
Sun, 10 Nov 2002 12:01:11 -0600


There have been several posts recently about PC clocks losing time when running TRLog.  The problem probably originates with the fact that the PC CMOS hardware clock is not the same as the software clock.  The former is as accurate as a typical quartz clock, but the latter is notorious for inaccuracy.  I believe that I once read an article about the difference, but I can't find it now.  As I recall, the operating system consults the CMOS hardware clock once, when it is started.  It can also reset that CMOS hardware clock, through DOS TIME or DATE or via Windows Control Panel, for example.  But software normally consults the software clock and not the hardware clock.

A Google search indicated that the clock problem with TRLog dates at least to 1997, when a post complained:
***************************
Here's a question for the computer experts:  I exit WIN95 to DOS to run TR.
When I take a fairly long contest break my PC goes into an "energy saving"
mode, which means it basically shuts down. I re-activate TR the same way
that one exits a screen saver program, by hitting a key or moving the mouse.
The problem is that after the PC goes into the energy saving mode the
contest clock time is incorrect. It appears that the clock stops when the PC
is in the energy saving mode. What can I do to prevent this from happening?
Possible solutions include: (1) Changing something (I don't know what) to
prevent the PC from entering that mode; (2) Exiting the program and shutting
the PC off -- then restarting when the break ends.     Is there a better way?  
Thanks! 
***************************

Is there a reflector subscriber who knows enough about the difference between the CMOS hardware clock and the DOS/Windows software clock that he/she could explain possible things to check if a difference is encountered?

Fortunately, I don't have this problem with my current machine, but that might not last forever, so I'd like to learn the answer to the time problem.

73, Paul W8TM