On Fri, Oct 24, 2014 at 7:21 AM, <rob@xxxxxxxxx> wrote:
IBM i, being a real operating system, will automatically change the system
value QUTCOFFSET during the time change. Ours will go from -4:00 to
-5:00.
Windows, being an unmitigated piece of crap, has their own concept of the
system value QUTCOFFSET.
It's not quite fair to say they have "their own" concept of
QUTCOFFSET. What they have is a concept that is analogous to QTIMZON.
If you do WRKTIMZON, you will see that QN0500EST3 shows an offset of
-05:00, even now, during Daylight Time. And if you "display details"
for that time zone, there is no mention of -04:00 anywhere. What it
DOES say is that the Daylight Saving Time shift is 60.
You can argue that a geopolitical concept of time zones is not as
useful as a truly coordinated *universal* time, but for some people in
some situations, "time zone" is more useful than "current time offset
from GMT". To take one example (which you kind of alluded to later in
your message), it *is* convenient to be able to say "OK, I'm setting
up this machine, and I'm in New York, so I'll pick the New York time
zone" and not have to worry what time of year it is, or what the
difference is between *current* New York time and GMT. And once the
machine is set up with the right time zone, you don't *ever* have to
manually futz with the time again, as long as it physically stays in
that time zone and has an Internet connection.
At least in principle, that's the idea behind QTIMZON, and it's the
same idea that Windows is attempting to adhere to. QTIMZON is a more
user-friendly notion than QUTCOFFSET, and Windows does tend to value
user-friendliness over academic rigorousness.
John Y.
midrange.com
