Got it. 

Charles

-----Original Message-----
From: IBM Mainframe Discussion List [mailto:[email protected]] On
Behalf Of Jim Mulder
Sent: Wednesday, April 08, 2015 3:57 PM
To: [email protected]
Subject: Re: Leap Second

IBM Mainframe Discussion List <[email protected]> wrote on
04/08/2015 03:06:57 PM:

> From: Charles Mills <[email protected]>
> To: [email protected]
> Date: 04/08/2015 03:47 PM
> Subject: Re: Leap Second
> Sent by: IBM Mainframe Discussion List <[email protected]>
> 
> I was just searching MVS Data areas for CVTLSO and the first hit I got
was
> this (under ASASYMBP Map):
> 
> SYMBTTIMESTAMPISGMT
> "X'20'" The input timestamp is GMT-time, not Local. Note that the 
> value from STCK or STCKSYNC is not GMT. It needs to be corrected by 
> subtracting CVTLSO.
> 
> Does that not imply that leap seconds have to be considered where
converting
> from STCK to civil time?
> 
> I do an STCKCONV "by hand" in my code (not because I'm think I'm so
smart
> but because the code is in C and it is faster and simpler to do it
in-line
> than to call out to assembler) and I certainly factor in CVTLSO, and
seem to
> get correct results.
> 
> Charles

  Yes, if you want time zone and leap second adjustments to a TOD value, you
must make them yourself before converting (or before calling STCKCONV).
Assuming you are adjusting for the time zone of the z/OS on which you are
running, and the current number of leap seconds, you can add CVTLDTO and
subtract
CVTLSO.   If you are converting for a different time zone or a historical
time stamp when the number of leap seconds was different, then it is your
problem to figure out the adjustment values. 

----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN

Reply via email to