On Friday, 04/08/2016 at 07:41 GMT, Rob van der Heij 
<robvdh...@vnet.ibm.com> wrote:
> I'm afraid I am not going to do multiple timezones. Understand that we
> don't even change the name of the timezone here when the offset changes.
> And I am not going to touch the leap seconds either.
>
> Yes, if you set your alarm for 7am 20 years ahead, it may well be that
> powers have ruled and changed the schedule for DST (or dropped it).

If there's a change in the law, then DST will occur at a different time, 
or never.  All that needs to happen is for WAKEUP and Pipelines to 
understand what to do when a zone-change happens.

In particular, they should
a) Register to hear about TZ changes
b) When TZ change occurs, query system to find new TZ and offset
c) Adjust intervals that represent absolute time value to account for new 
offset.
d) Wake up any task that was 'swamped' by a forward-motion TZ change

CMS functions shouldn't be worrying about the TZ definitions, per se, but 
should be querying CP about the name and the offset.

>
> To me local time is STCK plus the current offset. It sounds like it is
> wise to distinguish between interval and time of day, and then handle 
them
> differently around timezone change.

Yes.  In MVS, setting the timer in a program is specified as 
a) A delay in hundredths of seconds as defined by a 31-bit binary integer
b) A delay defined by a 64-bit TOD clock value, where bit 51 represents 1 
microsecond
c) A delay represented as HH:MM:SS:hh
d) A delay in time units represented by a 32-bit value, one timer unit is 
equal to 26.04122 microseconds
e) An absolute value expressed as HH:MM:SS:hh UTC
f) An absolute TOD clock value that is expressed in LOCAL TIME.

I think C and E are sufficient.  For convenience, the following could be 
added:
g) An absolute value expressed as HH:MM:SS:hh LOCAL TIME

Alan Altmark

Senior Managing z/VM and Linux Consultant
Lab Services System z Delivery Practice
IBM Systems & Technology Group
ibm.com/systems/services/labservices
office: 607.429.3323
mobile; 607.321.7556
alan_altm...@us.ibm.com
IBM Endicott

Reply via email to