From: mike 

On Sun, 4 Mar 2007 12:52:35 +0200, "Antti Kurenniemi"
 wrote:

>"Geo."  wrote in message
news:45e9f626$3{at}w3.nls.net...
>> "Mike N."  wrote in message
>> news:l1dgu2957ubn3reo468k8h62qbuf9e0tu8{at}4ax.com...
>>
>>> Daylight Saving Time. Some may characterize DST as a
"mini-Y2K". No, it's
>>> not, it's worse than Y2K. At least with Y2K, everyone had the runs for
>>> the
>>> year leading up to it and were prepared. DST has hit American
IT over the
>>> head with a two-by-four. Here at Sunbelt, we have had buckets of work to
>>> do
>>> for the transition - updates to SalesLogix, Exchange, SQL 2005, Office,
>>> Java runtime engine, etc.
>>
>> WTF is wrong with these people, has the automatic time switch function
>> always worked perfectly for everyone or something, I mean bfd so a
>> computer is off by an hour until someone changes the clock. what like
>> we've never had that happen before?
>>
>> Twice every year I have to change half our computer's manually. That's why
>> I run an NST program on some of them. it corrects the clocks every 5000
>> seconds and cost me a whopping $0.
>
>A lot of apps need to calculate time / date internally, and when money is
>involved it's critical that the calculations are correct. So, if the rules
>are changed, i.e. if you have programmed your app to calculate time using
>the OS provided functionality, then this would break those apps. And yes, I
>know it's only a very small change, but I was involved in programming a
>delivery control system for a gasoline provider, and they tracked the price
>changes by 15 second intervals! I'm sure there are other industries where
>this would be significant change to make people not very happy - although I
>think mostly the problems seem overstated: if you have a system that is that
>critical, then don't run it on a system that is out of support.

The NTP software goes through great lengths to assure that system (not
necessarily display time) time only increases.

 /m

--- BBBS/NT v4.01 Flag-5