CMOSTIME Version 1.0
		Copyright 1994 by David J. Arcand
  The Programmers Bulletin Board System of Western Massachusetts.
		 413-732-8323 14.4Kbps 24 hours/day

		      July 30, 1994  12:35 PM

	I've been running some kind of BBS since August of 1985.  One 
problem that has continually plagued me was the lack of a reliably 
accurate clock in TRS-80's and now the PC.  In the past it was always a 
matter of going in once a week and correcting the 40 or so minute 
difference between the computer that ran the BBS and the rest of the 
universe.

	That was until I started running Wildcat 3.9m and DESQview with 
three lines.  Now I was finding sometimes a 90 minute timewarp per day.  
It was not unusual for me to even miss several mail run windows and 
folks complaining about slow network mail.  Now I was having to manually 
reset the time every day I came home from work !   This state of affairs 
takes a lot of fun out of running a BBS.  Totally unacceptable.

	Now I had many solutions, most involve spending money that I 
didn't have.  Purchasing a faster motherboard, purchasing an RTC chip, 
or calling long distance to some national time standard.

	Now, the thing was, when I rebooted the system, the time was 
accurate again.  This meant that the HARDWARE REAL TIME CLOCK was still 
dead balls accurate, but the BIOS interrupt driven clock was loosing its 
mind getting interrupted all the time.

	What was going on was that with three COM ports, (two with high 
speed modems) there was not always a timely interrupt to increment the 
BIOS clock.  Sometimes it got missed and during high speed uploads, I 
could lose upwards of 5 minutes per session.  Well, now that I knew what 
the problem was, and a possible solution in mind, I called the Mustang 
HQ BBS and asked about something to help.  John Neafus and Samual Koh 
came back with the inspiration that I'd probably have to write this 
utility myself.

	What does CMOSTIME do ?  It reads the time from the hardware RTC 
and writes it to the BIOS clock.  I could have written it in about 10 
lines of code but there are some other stuff there to report on how 
badly the clock was off in the first place.

	I run this program every 8 hours as part of a routine batch event 
from within Wildcat.  There is absolutely no reason why this program 
would not work on other BBS software or usage.  For instance a SERVER 
that is constantly being access by other workstations is sure to lose 
time and could be fixed up by running this program.

	The program is freeware.  The source code is included with the 
distribution, and I reserve all rights.  But if you make changes, you 
should to forward me a copy and keep my credit in it.

	Whenever I write a program, I play this game called "GOOD, BUT 
COULD BE BETTER" and wonder what I've done to improve the program.  Best 
I can come up with is color characters for the final message report to 
warn of really wild ass time differences.  Command line options to just 
report the difference but not make any changes.  I mention these only 
for those folks who are learning to program and want to have something 
easy to learn with.

	David Arcand -- Springfield Massachusetts -- July 30, 1994