[Discussion] New Daylight Savings Time Changes [Was: ECS- Eliminate excess Printers]

[Carl Gehr]

On Mon, 22 Jan 2007 22:17:01 -0800 (PST), David Azarewicz wrote:

>On Mon, 22 Jan 2007 21:02:31 -0500 (EST), Carl Gehr wrote:
>
>>On Mon, 22 Jan 2007 14:33:35 -0800 (PST), Jon Harrison wrote:
>>
>>>On Mon, 22 Jan 2007 14:42:27 -0500 (EST), Carl Gehr wrote:
>>>
>>>>
>>>>I believe the expanded SET TZ that you show is not correct.
>>>
>>>Carl:
>>>
>>>I agree with you, that is why I posted the comment.
>>>
>>>jon
>>
>>Sorry, I thought you had posted:
>>>SET TZ=PST8PDT,4,1,0,7200,10,-1,0,7200,3600
>>
>>I was reacting to the '...4,1,...' which is not correct.
>>It should be:         '...3,2,...'
>>
>>The rest was just background.
>
>
>The original issue was that anything other than the standard 
>TZ=PST8PDT (or whatever time zone you are in) breaks zip. It doesn't 
>really matter what is added on the end, correct or incorrect, it still 
>breaks zip (and probably other unix ports).
>
>I have always wondered why everyone is making such a fuss about the 
>system changing the time by itself.  I have always sync'd my system to 
>NIST and let the government tell me what time it is, including whether 
>it is daylight time or not.  It has worked for me for years and I don't ever 
>have to think about it.  Maybe in other countries it works differently, but 
>in the US it has always worked for me.
>
>David

And, WADR, did you think the whole Y2K effort was also a waste of time
and effort, because it could have been handled by a bunch of clerks and
operators at system consoles during the rollover?

I guess if you want to be sure that every system is manually attended
to at 2:00AM on the appropriate change over day, to and from DST, then
that's your right to do so.  I believe that most people would prefer
that systems work correctly without having to be manually tweaked
because the system had a logic problem [or bad specification, if you
prefer] and no one would take the time to fix it.

And, just for the record, it is possible that there are applications
[not system code or just changing the time by an hour at a particular
instant] that could have logic hard coded to the old rules for the DST
change.  If that application is calculating a time differential [for
whatever reason], it may get bad results.  Yes, it is somewhat of a
mini-Y2K all over again; very 'mini' but the exposure is there.

IBM has posted over 175 changes that it has had to make in its own
code.  I doubt that they would have spent that much time/money if it
were as simple as you claim.

But, your system; your choice.  Your right to do it how ever you want. 
Personally, I prefer a correct system; MY choice.

Carl