On Mon, 3 Nov 2014 08:18:03 -0600, Kirk Wolf wrote: >I don't disagree that the TZ default for z/OS Unix should be derived from >the z/OS TIMEZONE offset. > >But the RFE that I submitted is about more than setting TZ - it is about >having dubbed processes inherit their environment variables from the init >process. Also, read the RFE and you will see that the suggestion is that >there be a system and user OMVS segment level option to control whether >blind dubbed processes should inherit their environment or not. > Oh, heck! I'll go further. Beyond that, the default TZ should be a setting in the user's OMVS segment, just like HOME and SHELL (umask, too.) FTP server should use this in formatting file timestamps for the LS command.
>Kirk Wolf >Dovetailed Technologies >http://dovetail.com > >On Mon, Nov 3, 2014 at 6:49 AM, John McKown <[email protected]> >wrote: > >> On Mon, Nov 3, 2014 at 6:25 AM, Elardus Engelbrecht < >> [email protected]> wrote: >> >> > Kirk Wolf wrote: >> > >> > >>Wouldn't it be great if you didn't have to configure default system >> > environment variables like TZ for each and every job? >> > >> > ... if your system is using LE, ... >> > What does "system is using" mean? LE is enabled for the LPAR? Any active job is using LE? Other? (JMcK:) >> This is what I have done. I am "lucky" in that my company is little and >> does not span time zones. But it does make me wonder what a multinational >> company would do for a default TZ which is not GMT. TZ of the "home >> office", or of the physical location of the data center, or the CEO's >> wife's favorite second cousin? <grin/> >> A good argument for a RACF segment setting. >> > >>Headline: Provide option for dubbed processes to inherit environment >> > from init process >> > >> > Question: add another requirement - that dubbed proc can either accept >> > inheretance or not. If not, use its own defaults if not setup (own TZ). >> > >> > >> I'm not understanding this. If the inheritance were done, the application >> could still have its own defaults set. Perhaps via LE parameters in the >> PARM=. Hum, and if a dubbed process "inherits" its environment variables, >> and is in an LE run unit, does it inherit from "init" or from the ENVAR >> parameter, if any, set via the PARM? I.e. who takes precedence when an LE >> run unit is dubbed. I vote for LE. And, other that in code, how are the >> application specific environment variables set? Via the SYSENV DD >> statement? Of course, if a request is made from IBM, that is _their_ design >> decision. I've often been told that IBM prefers to be told what you want >> the end results to be, not how you want them accomplished. >> . >> ... I don't >> really see the plus of /etc/localtime vs. just setting the z/OS TIMEZONE >> offset. They are both "global" to the z/OS image. I would hope that the >> UNIX default would be the same as the z/OS batch default. But that's just >> my personal take on it. >> Correct display of file timestamps from the other side of the Daylight Time boundary. (Today AZ and CO are the same; last week they were an hour apart.) Actually, the z/OS TIMEZONE should follow POSIX conventions in order to be DST aware. -- gil ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
