Even if your clock loses time and the jumps are forward, you certainly
won't be able to rely on differences between timestamps to tell you the
difference between actual times if the timestamps span the jump point. Also
code that relies on internal short delays to timeout various operations may
behave erratically around the jump time.
If you care about time accuracy, don't fix it with multiple jumps!
DS
>There is one very easy way to fix this problem, set netdate to sync your
>clock every 2 hours or so. I do this on all of my networked systems so
>their logs are as close to real time as possible. I don't know why it
>would be doing this, as I don't have any real problems, except for the
>fact that whenever I set the date with netdate, and reboot my system, the
>time is just as wrong as it was before netdate (CMOS dosen't pickup the
>changes.
>
>RG gashalot@gnu.net