SG> After testing the taskbar clock here, I tend to agree. I originally
SG> compared the RTC in my bbs software (in a window) with the countdown
SG> clock in Irex. My bbs RTC didn't falter so I assumed it was Irex. :)
  
 RN> What's really odd about this whole thing is that my mailer's clock
 RN> is accurate and exhibits no such anomaly.  In one segment of the
 RN> main screen, it shows the current time progressing and a countdown
 RN> clock in another section of the main screen counting down to the
 RN> next event without even so much as a hiccup.  

as a coder, i highly suspect what you are seeing is an artifact of the
differing methods of sending data to the video buffer(s)... i remember
coding stuff for DOS and DESQview... DESQview uses seperate buffers for
each task... additionally, DEQview also had virtualization settings that
can allow for programs to write directly to the assigned video buffers...
all my code performs similar majik and unless i do not, there are delays in
the screen writes such as you are seeing... in some circles, one might say
that BIOS writes are being used... however, with all the object oriented
code used today, i'd say that this is closer to the truth than anything
else...

)\/(ark