On Sun, 8 Feb 2009 10:31:46 -0500, David Andrews wrote:

>On Sun, 2009-02-08 at 09:41 -0500, Lizette Koehler wrote:
>> If you have a problem with UNIX time in z/OS vs. System time, check your
>> UNIX time values in TIMEZONE.   Both LE parms and Unix parms (/etc/profile
>> and /etc/init.options (?))
>
>We have a little program that we run weekly, during a brief sysmaint
>window, which interrogates a time server and sets the local time via
>operator command.  Works a treat in our little monoplex, and keeps us
>more-or-less synchronized with the outside world.
>
>But it doesn't modify those /etc files (used to set environment
>variables, and which aren't all that dynamic).  I wish the Unix time
>routines were able to accept something like TZ=CVTLDTO.
>
No!  What you propose would cause times displayed for file
timestamps to fluctuate incorrectly at daylight saving time
boundaries, and fails to support users' local adjustment
of the TZ variable.

If the timezone variable is set properly in (clearly too many)
/etc files, it should never be necessary to change it.  (Well,
only rarely, when legislation changes the daylight saving time
conventions.)

TOD clock drift is a different matter.  Has anyone tried to
compensate for this by fudging CVTLSO?  What would be the
consequences?

-- gil

----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Reply via email to