EPICS Controls Argonne National Laboratory

Experimental Physics and
Industrial Control System

1994  1995  1996  1997  1998  1999  2000  2001  2002  2003  2004  2005  2006  <20072008  2009  2010  2011  2012  2013  2014  2015  2016  2017  2018  2019  2020  2021  2022  2023  2024  Index 1994  1995  1996  1997  1998  1999  2000  2001  2002  2003  2004  2005  2006  <20072008  2009  2010  2011  2012  2013  2014  2015  2016  2017  2018  2019  2020  2021  2022  2023  2024 
<== Date ==> <== Thread ==>

Subject: Re: Daylight savings time
From: Andrew Johnson <[email protected]>
To: Eric Norum <[email protected]>
Cc: tech-talk tech-talk <[email protected]>
Date: Tue, 13 Mar 2007 12:16:14 -0500
Eric Norum wrote:
I think that WIN32 does provide this routine -- at least my quick check with google seems to imply this.

Ok, I didn't look.


The vxWorks case is a little trickier. I suppose that someone out there actually uses the IOC shell rather than the vxWorks shell.

I don't see how that's relevent (the iocsh is moving into libCom and has to link on all architectures; we use it here on all our vxWorks IOCs). The change you presented just inserts a kludge into the epicsEnvSet() command. Just adding a tzset() command for those architectures that provide and need that routine would be preferable.


What's' your feeling on weak references?

Very wrong for this purpose. We know at compile time whether this architecture uses tzset(), and using a weak reference just postpones the check until runtime (apart from the fact that they're not portable - not all the binary formats we build even support weak references).


How about an epicsTzset() macro in some os-dependent header. This would do nothing on vxWorks and expand to tzset() on all others.

That still doesn't remove the fact that you'd be adding code that has nothing to do with actually setting an environment variable (other than the fact that the routine you're calling takes its arguments from the global environment) to the epicsEnvSet() command.


One solution to this would be to add a command that sets up the local timezone information. We could then make the same command work across all platforms, should that be seen to be desirable. I think that's a bit of an over-kill solution though, and I'd much prefer just adding the tzset() command - we could use an OSD header file to signal whether it should be included or not.

I'll let you know when I've committed my move of the iocsh files.

- Andrew
--
The right to be heard does not automatically include
the right to be taken seriously. -- Hubert H. Humphrey

References:
Daylight savings time Thompson, David H.
Re: Daylight savings time Lawrence T. Hoff
Re: Daylight savings time Andrew Johnson
Re: Daylight savings time Eric Norum
Re: Daylight savings time Andrew Johnson
Re: Daylight savings time Eric Norum

Navigate by Date:
Prev: Re: Daylight savings time Eric Norum
Next: RE: Darwin and EPICS_CA_AUTO_ADDR_LIST issue Jeff Hill
Index: 1994  1995  1996  1997  1998  1999  2000  2001  2002  2003  2004  2005  2006  <20072008  2009  2010  2011  2012  2013  2014  2015  2016  2017  2018  2019  2020  2021  2022  2023  2024 
Navigate by Thread:
Prev: Re: Daylight savings time Eric Norum
Next: Re: Daylight savings time Eric Norum
Index: 1994  1995  1996  1997  1998  1999  2000  2001  2002  2003  2004  2005  2006  <20072008  2009  2010  2011  2012  2013  2014  2015  2016  2017  2018  2019  2020  2021  2022  2023  2024 
ANJ, 10 Nov 2011 Valid HTML 4.01! · Home · News · About · Base · Modules · Extensions · Distributions · Download ·
· Search · EPICS V4 · IRMIS · Talk · Bugs · Documents · Links · Licensing ·