Re: Time Problem

[Fraser Farrell <fraser@trilobytes.com.au>]

On Thu,  8 Nov 2001 12:39, Tom Droege wrote:
> This all worked through the DST switch on October 28.  In fact it worked
> through November 4th.  That is images taken before and after October 28 had
> WCS parameters that gave correct star positions.
>
> The data taken on November 5th is suddenly off by one hour.  I have to
> subtract 1 hour from the indicated star position as shown by DS9 to get the
> actual star position.

Tom,

A few things for you to check:

 - The BIOS settings (usually accessed by pressing the DEL key during 
bootup). Some of them have their own ideas about daylight saving...if yours 
does you may be able to toggle it. See under BIOS Advanced Settings or 
Chipset Settings.

 - Some other program you ran recently may have changed the clock. I'm 
imagining a scenario where you (somehow) told a program it's 11:00am in your 
local -standard- time zone; and the program logic has said "but today's date 
corresponds to -summer- so let's make it noon on the clock".  Hmmm....wait a 
minute you're in the northern hemisphere so your local clocks would have 
recently subtracted an hour. Anyway you get the idea  ;-)

 - Your operating system may have twiddled the time. No version of MS-DOS or 
DR-DOS does automatic daylight saving changes, but Windows 95 and later do 
so, and Linux has always had this capability.


There are bugs with this function in some versions of Windows: if the 
computer is on at the moment daylight saving ends, it will turn back the 
clock. One hour later it will turn back the clock again. One hour later it 
will turn back the clock again....a reboot after 3am fixes it.

There was a related bug when daylight saving began in Australia last year. 
For reasons which defy logic and commonsense the New South Wales state govt 
decided to begin daylight saving 7 weeks earlier than usual in 2000. 
Supposedly because it simplified the logistics for the Sydney Olympic Games.

Microsoft issued an update to the Windows daylight saving code for this 
(buried inside a secuity update as I recall); but mistakenly applied it to 
-every- Australian locale. This error bit several of my customers who run 
time-sensitive software, many of them thinking it was a belated Y2K bug.

While we're on the subject of time zones, I want to remind our TASS software 
gurus that some of us are afflicted with non-integer time zones. I continue 
to be amazed at the number of -new- astronomical programs and telescope 
controllers that cannot comprehend (or accept) my local standard time of 
UT+9h30m. Or for the next few months UT+10h30m. As to how we got +9h30m as 
our time zone; well that's the outcome of a bigger saga of political 
stupidity than the recent NSW example.

A certain well-known manufacturer's controllers can't comprehend a latitude 
south of the equator either (hint: their name has five letters, starting with 
T). I trust our code cutters won't make -that- sort of elementary mistake!


cheers,

-- 



Fraser Farrell
---------------------------------------------------
Buying Your First Telescope In Australia
www.trilobytes.com.au/astronomy/scope.htm

Binocular variables for Southern Observers
www.trilobytes.com.au/astronomy/bvstars.htm
---------------------------------------------------