Daylight Saving Shift

[Wojtek]

Roedy Green wrote :

The other possibility, at least for Internet communication, email and
international meetups is to use UTC.

And database dates.

 

[Mike Schilling]

My point was, you can always pin a meeting time to a datetime in *one*
given location. Phrasing it as a 9 AM Pacific *and* 5 PM British is
already a cause of problems - just phrase it as 5 PM British *or* 9 AM
Pacific, and let the other parties worry about the translation.

Which places the onus on them to figure out that DST has arrived in a remote
country. Not a recipe for success.

 

[Arved Sandstrom]

Which places the onus on them to figure out that DST has arrived in a remote
country. Not a recipe for success.

Not on _them_, leastways not if you mean human beings. That's the whole
point of software, to take care of tasks like this. Date and time math,
especially if we stick just to the Gregorian calendar, is not exactly
that mysterious, even with DST thrown in. The information is available -
it's not hidden away. If I'm in Nova Scotia and it's August 2009, and
someone in Brussels suggests a meeting at 3 PM their time on Oct 30th of
this year, it actually only takes me less than a minute to *Google* up a
website that tells me that I'll still be in DST and they won't be...I
figure a simple app that talks to a simple database should answer a
question like that in not much more time than it takes me to call it up
and dial in the info.

Because some changes are announced pretty close to the actual switch
it's probably most convenient to look up this kind of information from a
service.

AHS

 

[Arne Vajhøj]

Then when the evening starts depends on what job you do. A farmer's
evening starts at full dark by that definition.

Doesn't it make more sense to define evening in terms of where the sun
is than what one's profession is?

I define evening as when the sun is close to setting, i.e., when the
light begins to fade. It's a fuzzy concept, of course, but utterly not
dependent on what the clock says.

I define afternoon as when sun passes its zenith.

I find other definitions stupid, as indeed I find the whole concept of
Daylight Savings Time. Ptui! I spit on the practice!

Most people follow the clock not the sun.

Arne

 

[Arne Vajhøj]

I would favor abolishing DST altogether. If that means leaving the
clock set ahead of historic Standard Time settings, so be it, although
the last time that was tried in the U.S. (in the 1970s) it was a
failure. People objected to the children having to wait for morning
school buses in the dark, among other things.

The subject is politically controversial. Some claim energy savings due
to the use of DST. AFAIK there's no hard evidence to support this, at
least, not that takes into account the increase in costs due to air
conditioning and morning lighting. Certainly there are lots of claims
that DST saves energy, but for some reason no one ever seems to cite
studies or methodologies to support those claims. There is a vocal but
politically disadvantaged contingent that denies the validity of those
claims.

But our infinitely wise and benevolent governments say that it must be
so, though somehow jurisdictions that don't use DST don't seem to suffer
unduly thereby. As Arne points out, dates are a matter of
socio-political mandate and our software simply must reflect the reality.

There are plenty of research in this area.

References can be found in:
http://en.wikipedia.org/wiki/DST#Energy_use
http://en.wikipedia.org/wiki/DST#Economic_effects

Arne

 

[Arne Vajhøj]

The recent research summaries I've seen all seem to indicate that the
impact of DST on energy use is somewhere around ±0.2% (it's a number
that's rarely put into full context, so I'm not exactly sure what the
percentage is of--probably average daily summer energy usage). The sign
is naturally hotly debated in political circles whenever tweaking DST is
bandied about.

*Performs some searching to find research papers*

The literature review I just finished reading seems to suggest that most
of the conclusions about the energy-saving nature of DST were formulated
about 25 years ago, when lighting in particular was much less efficient
than now (the hypothetical best-case scenario for energy savings would
be equivalent to replacing about 15% of your incandescent light bulbs
with compact fluorescents) and also fails to take into account the
modern shifts in habits. Its primary conclusion was "the stuff out there
sucks, we need modern comprehensive research on this topic."

Other papers recently published seem to suggest that DST may no longer
be saving energy. That may just be a manifestation of confirmation or
perhaps publication bias--I'd expect that people finding energy savings
due to DST would be less likely to publish their results nowadays. I'm
also not a big fan of DST myself.

Oh well, if humanity ever discovers interplanetary or even interstellar
travel (as well as sufficiently speedy communication to make
chronological synchronization across disparate settlements necessary),
the mess resulting from DST will be the least of our worries.

Wikipedia has lots of references including newer material.

Arne

 

[Mike Schilling]

Not on _them_, leastways not if you mean human beings. That's the
whole point of software, to take care of tasks like this. Date and
time math, especially if we stick just to the Gregorian calendar, is
not exactly that mysterious, even with DST thrown in. The
information
is available - it's not hidden away. If I'm in Nova Scotia and it's
August 2009, and someone in Brussels suggests a meeting at 3 PM
their
time on Oct 30th of this year, it actually only takes me less than a
minute to *Google* up a website that tells me that I'll still be in
DST and they won't be...I figure a simple app that talks to a simple
database should answer a question like that in not much more time
than it takes me to call it up and dial in the info.

But if it's a standing 9:00 AM meeting that moves to 8:00 AM for just
a week or two [1], the attendees have to know to check. They
generally won't. (Yes, calendar software should let you set a
meeting to a different Timezone/locale and warn you when repeated
meetings change in local time. Absolutely. it should. I don't know
of any that does.)

1. Real-life example.

 

[Mike Schilling]

I don't know what Outlook does, but it does not always work well

Very true.

for a reoccurring meeting scheduled over DST changes.

Arne, there was some uneeded text at the end of your post.

 

[Arved Sandstrom]

Most people follow the clock not the sun.

Arne

Indeed. When I got on the bus after work yesterday (Monday), it was
pitch dark, yet I was able to correctly wish the bus driver Good Afternoon.

AHS

 

[Dr J R Stockton]

In comp.lang.java.programmer message <[email protected]
september.org>, Mon, 2 Nov 2009 14:45:58, Mike Schilling

Which places the onus on them to figure out that DST has arrived in a remote
country. Not a recipe for success.

Clearly the Southern Hemisphere will never put its clocks forward and
back on the same days as the Northern; and, as it seems to be generally
agreed that Summer Time should be longer than Winter Time (starting near
the Spring Equinox but finishing over a month later than the Autumn
one), they'll not change in the opposite direction on the same dates
either. There is therefore an essential disagreement between those who
have Summer Time in July, those who do not have it, and those who have
it in January; and the Lines of Disagreement follow, very approximately,
the Tropics of Capricorn and Cancer, with some regions being anomalous.

In the Northern Hemisphere, the vast majority of places which change
their clocks have chosen to do so on the last Sundays of March and
October : In Europe, all but the Far East do it at 01:00 UTC, and in
Asia all outside the Middle East change on the same dates but at a fixed
local time.

The only[1] Northern country which has chosen to act otherwise and use
different dates is the USA.

The problem that you[2] see is therefore of your own creation, and
sympathy from outside should not be expected. You, through your
"democratic" system, should have chosen to move to use the same dates as
the vast majority of Northern Hemisphere clock-changing countries with a
substantial majority of Northern Hemisphere clock-changing people. Then
you would only have date difficulty when dealing with the Deep South.


[1] Canada made no choice; there, the matter is left to the Provinces.
The Provinces all made the same choice of date, but they did not make it
simultaneously. Mexican rules also depend on location. The rest of
clock-changing North America is in practice a minor detail.

[2] Those who do not choose to indicate location or nationality are
deemed to be American.

 

[Mike Schilling]

Dr J R Stockton wrote:
..

In the Northern Hemisphere, the vast majority of places which change
their clocks have chosen to do so on the last Sundays of March and
October : In Europe, all but the Far East do it at 01:00 UTC, and in
Asia all outside the Middle East change on the same dates but at a
fixed local time.

Many don't change their times at all, which causes a discrepency with those
that do.

The problem that you[2] see is therefore of your own creation, and
sympathy from outside should not be expected.

Oh, and you're a lunatic.

 

[Arne Vajhøj]

Very true.


Arne, there was some uneeded text at the end of your post.

For this particular topic that is the relevant occasion.

BTW, in general I am not unhappy with it - it mostly works.

Arne

 

[Mike Schilling]

For this particular topic that is the relevant occasion.

BTW, in general I am not unhappy with it - it mostly works.

For the use I make of it (to add calendar item manually), it's fine. A few
years ago, when I was briefly working at a place that would mail out meeting
invitations with calendar attachments that would automatically enter
themselves, any meeting change was invariably followed up by five or six
more changes, each attempting to undo the previous change and make the
desired change properly. Since the person making the changes wasn't an
idiot, I concluded that it doesn't work very well.

 

[Dr J R Stockton]

In comp.lang.java.programmer message <[email protected]
september.org>, Tue, 3 Nov 2009 13:33:56, Mike Schilling

Dr J R Stockton wrote:

Many don't change their times at all, which causes a discrepency with those
that do.

I am there referring, as can easily be seen, only to places which do not
change their clocks. Those which do not change have a fixed offset from
GMT. You should not be confused about when they change their clocks.

You should know when you change your own clocks, and you should know the
difference between your time(s) and GMT. So it should be perfectly easy
for you to work out the difference between your time and that of, say,
Reykjavik (BTW, their rules are likely to change soon, but they may not
yet have realised it).