Experimental Physics and
Industrial Control System

"Laznovsky, Michael" <[email protected]> · Mon, 5 Mar 2007 10:18:43 -0800

There's no way to say "there is no daylight savings time this year"... it's hardcoded 
into the UTC<->local conversions.  But that only affects "local time" on the IOC, which 
matters to e.g. timestamp records in the DB (which PEPII does use, for fault times for 
instance).

Mike

-----Original Message-----
From: Chestnut, Ronald P. [mailto:[email protected]] 
Sent: Monday, March 05, 2007 10:14 AM
To: Andrew Johnson
Cc: [email protected]
Subject: RE: 3.13.6 daylight-savings follies

This sounds great! - now a question about some older applications ...

If we don't want to really fix the problem, but work around in the least invasive way, (for 3.13.2, for example), is it possible to set the TS_MIN_WEST string to one hour ahead (Denver instead of Palo Alto), AND set it to say "there is no daylight savings time this year". We could then simply reboot on March 11, and forget any problems until the November date where we switch back.

And another strange question - is it necessary at all to change the IOC times, if one is only interested in timestamps on PVs? It seems to be the case that IOCs return time stamps in GMT, and that clients on Unix do the conversions to local time.

Ron Chestnut

Experimental Physics and Industrial Control System

Experimental Physics and
Industrial Control System