Re: Time change questions !
- From: Ken Fairfield <my.full.name@xxxxxxxxx>
- Date: Tue, 31 Oct 2006 16:06:48 -0800
On 10/31/2006 11:46 AM, Richard B. Gilbert wrote:
JF Mezei wrote:Stephen Hoffman wrote:
There are applications around which do not react appropriately when
the time changes -
However, when you consider that competing modern OS have automated,
sudden time changes, new applications, often written for such OS and
later ported to VMS, would have been written to handle such a sudden
time change.
If you have an old application being installed onto a fresh/new VMS
instance, then perhspas the onus should be on the system manager to
disable to automated time change. Wouldn't the vast majnority of apps
now survive a sudden time change ?
If you do financial transactions and you really do not want to have
transactions "back dated", then you need to a a lot to handle those time
changes and would have to be very aware of everything and disable the
SYSGEN AUTO_DLIGHT_SAV parameter (and possibly use some app that tweaks
the clock rate to slowly drift it to lose/gain an hour).
Modern O/Ss, including VMS, normally keep time in UTC. The display of local time is merely a convenience for the user. Each user, if he wishes, can have times displayed using his own local time zone. Time is thus always monotonically increasing, it's only the DISPLAY of the time that bounces around; time stamps should be unambiguous.
I'm sorry, this last paragraph is NOT TRUE for VMS.
VMS system time *is* the displayed time, no matter the timezone and/or
TDF settings. While there are various system services/LIB$/CRTL
routines to convert between system time and UTC, it is entirely
incorrect state that VMS uses UTC as the internal (system) time.
You can reason this out for yourself: a VMS quadword time is the number
of centiseconds since 17-NOV-1858. There is NO TIMEZONE information
carried in that definition. Data files (SYSUAF for example, but
database records as well) that store the time often do so as the
binary VMS quadword. Given the late arrival of UTC on VMS, it would
break backward compatibility to change the definition of that quadword
to be defined as UTC (and VMS Engineering doesn't do that...).
This does not mean you can't set the local timezone to GMT and avoid
the problems... It is actually a human factors problem rather than
a technical one (all the really hard problems seem to involve humans,
not machines!).
-Ken
--
I don't speak for Intel, Intel doesn't speak for me...
Ken Fairfield
D1C Automation VMS System Support
who: kenneth dot h dot fairfield
where: intel dot com
.
- Follow-Ups:
- Re: Time change questions !
- From: Rich Alderson
- Re: Time change questions !
- From: Richard B. Gilbert
- Re: Time change questions !
- From: Paul Sture
- Re: Time change questions !
- From: Stephen Hoffman
- Re: Time change questions !
- References:
- Re: Time change questions !
- From: Richard B. Gilbert
- Re: Time change questions !
- Prev by Date: Re: XML Parser OpenVMS 7.3-2 OSU Web Server.
- Next by Date: Re: HP eyes top VMS people for cuts
- Previous by thread: Re: Time change questions !
- Next by thread: Re: Time change questions !
- Index(es):
Relevant Pages
|