@Peter, thanks, interesting. I have tried to wrap my head around the exact
meaning to the system of "CET" and similar strings (as opposed to their meaning
as civil abbreviations).
@John, is that true? This is a "mainframe" behavior, not a UNIX behavior?
Charles
-----Original Message-----
From: IBM Mainframe Discussion List [mailto:[email protected]] On Behalf
Of John McKown
Sent: Friday, November 04, 2016 4:58 AM
To: [email protected]
Subject: Re: Question on TZ and European time change
On Fri, Nov 4, 2016 at 2:30 AM, Peter Hunkeler <[email protected]> wrote:
<snip>
> And from the same description you can see that the actual strings
> being used to name the standard and the DST periods (CET and CEST in
> the example
> above) have no meaning to the system. You can set them to anything you
> want without influencing the behaviour.
>
>
> The strings are only meant to show something meaningful to users.
>
>
> The first string is mandatory. The second string, in addition to be
> meant for the user, is also a flag to tell the system to apply DST
> according to the standard rules (mentione above), or by the rules specified.
>
>
> So, with
>
>
> TZ=WINTER-1SUMMER,M3.5.0/2:00,M10.5.0/3:00
> date '%a %b %e %T %Z %Y'
>
>
> will display
> Fr Nov 4 08:26:17 WINTER 2016
> today, but had displayed
> Fr Oct 28 08:26:17 SUMMER 2016
> last week.
>
>
> --Peter Hunkeler
>
Another indication that, at least at the time this was done, IBM was not
overly concerned with making z/OS UNIX be acceptable to the "real" UNIX
community. They just wanted to stamp "POSIX compliant" on z/OS for marketing
purposes. Aside: I really appreciate what Mr. Schoen of IBM has done to help
make z/OS UNIX more useful (bpxwunix & bpxwdyn come to mind).
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN
