 |
 | Home | Contact Us | FAQ | Search & Site Map | Link to Us |
 | Sign In | Join | Other 45 Sites in Network |
Java Forum / General / January 2006Handling of DST in java
Hi, Does java handle DST automatically ? I'm using java Gregorian calendar to get the time/date. Does this automatically give the local time/date iirrespective of DST or special handling should be done ? > Does java handle DST automatically ? > I'm using java Gregorian calendar to get the time/date. Does this > automatically give the local time/date iirrespective of DST or special > handling should be done ? What do you really want to do? If you need to represent a point in time internally in your application, then a calendar is a bad choice. Use a Date() object, or the time (as a long, in milliseconds). Java keeps its date/time in milliseconds since Jan .1st, 1970 00:00 GMT. Only when you you do some I/O you should consider if you want to have GMT/UTC date/time, or local time (with DST) E.g. if you just need to print out some time, [Simple]DateFormat is often a much better choice than a calendar. It takes the local time-zone into account, and you an specify another zone. Time-zone information contains DST information. A calendar is useful for presenting(!) some point in time in, guess what, calender format, and doing some calculations based on this representation, but not more. The representations produced by GregorianCalender are somewhat imprecise. E.g. GregorianCalendar doesn't know about leap seconds. GregorianCalender does take time-zone information (and thus DST, if defined for a time-zone) into account. But if I remember correctly, does so inconsistently. If I remember correctly, then it sometimes assumes a change from/to DST happens at midnight local time, which is wrong for several time-zones. GregorianCalender also has a problem with the times which appear twice when the clock is turned back at the end of DST. You can't specify if you want to have the A (DST) or B (standard) time when you e.g. set the HOUR value, or when you move into that period with and add(). Doing timekeeping with a calender is usually a bad idea in Java. /Thomas  SignatureThe comp.lang.java.gui FAQ: ftp://ftp.cs.uu.nl/pub/NEWS.ANSWERS/computer-lang/java/gui/faq http://www.uni-giessen.de/faq/archiv/computer-lang.java.gui.faq/ >Does java handle DST automatically ? >I'm using java Gregorian calendar to get the time/date. Does this >automatically give the local time/date iirrespective of DST or special >handling should be done ? Yes. However, because the tables are so voluminous, and because people think it fun to change the rules, as if they were ordering the sun to change it rising time, Sun's tables will always be a little off. I think it is a nutty idea. It creates two goofy anomalies each year -- a missing hour and an hour lived twice. Leave the time alone and just declare different summer and winter business hours. Non-compliance would help spread rush hour. SignatureCanadian Mind Products, Roedy Green. http://mindprod.com Java custom programming, consulting and coaching. >I think it is a nutty idea. It creates two goofy anomalies each year >-- a missing hour and an hour lived twice. Leave the time alone and >just declare different summer and winter business hours. I agree. It just seems to be an attempt to show to power of the government, "look, we can change the run of the time!" or "Look, we can make completely idiotic rules, still you have to live by them." In 2005, US-president Bush signed a paper to extend DST by four weeks in the USA (starting in 2007), and to make life even more complicated by adding another rule modification ... There is an "international movement" against it, the German website is http://www.initiative-sonnenzeit.de/ > I agree. It just seems to be an attempt to show to power of > the government, "look, we can change the run of the time!" or > "Look, we can make completely idiotic rules, still you have to > live by them." Local time is a concept on its way out. With netmeetings across the globe, you have to arrange your meeting in some common global base time, be it UTC , EST or head office time. Similarly for satellite broadcasts across time zones. As travel speeds up, we may find even meetings between in and out of towners arranged in some common base -- NOT local time. When a politician meddles with the rules for time, it creates a headache for anyone who has a payroll system, or anything that runs 24 hours. I first became aware of the problem back in the 1970s. To my surprise, "war time" as it was originally called was quite chaotically employed, using an offset up to 5 hours, with every community separately making up their own rules about when to kick in and out each year. Railroads refused to play and published their schedules in standard time. I suppose with computers they have now relented. I like the idea of keeping correlation between time and position of the Sun in the sky. Pretending the sun in irrelevant to man's activity is just arrogance. It enforces the idea that man is not an animal with the same environmental limitations that face all animals. Java wisely stores all dates in UTC and considers timezone merely a display quirk, not something inherent in the measurement. SignatureCanadian Mind Products, Roedy Green. http://mindprod.com Java custom programming, consulting and coaching. >> I agree. It just seems to be an attempt to show to power of >> the government, "look, we can change the run of the time!" or [quoted text clipped - 29 lines] > Java wisely stores all dates in UTC and considers timezone merely a > display quirk, not something inherent in the measurement. The sun is in fact a prime motivator for daylight saving time. Politicians are not simply meddling with time to display their power. It has been shown that daylight saving time results in less energy consumption. I'll leave it as an exercise to the reader why that is a good thing. :-) Ray SignatureThis signature intentionally left blank. >It has been shown that daylight saving time results in less >energy consumption. "less" in relation to what? Leaving the time zone unmodified, but shifting the start and end time of work might have the same effect. >>It has been shown that daylight saving time results in less >>energy consumption. > > "less" in relation to what? I think that it is clear that the statement above compares daylight saving time to the lack of daylight saving time. > Leaving the time zone unmodified, but shifting the start and > end time of work might have the same effect. Perhaps, but that is not something the government can easily legislate nor enforce. (Do you really want the time police checking to see if you are working an hour later?) Changing the wall clock time is the easiest way to get people to actually observe the change. There is nothing sacred about the time zone. It has always been under the purview of government. In China, there should be about five time zones by geographical standards, but there is only one. In the U.S., the Eastern time zone is about twice as big as it should be. Ray SignatureThis signature intentionally left blank. >Perhaps, but that is not something the government can easily >legislate nor enforce. (Do you really want the time police >checking to see if you are working an hour later?) Something like this is done here for the closing-time of stores. So, since the government already enforces opening-times and closing-times of stores, they could change these and they could change to times for schools, universities and their own agencies. Some other institutions then might follow. >There is nothing sacred about the time zone. It is only that, given two time stamps of the legal time for a location in Germany, like 2005-07-10T21:38:10 and 1960-07-10T21:38:13 it would be a typical programming task to write a routine that can calculate the difference between them in seconds. And this would be much more easy, without the need to have to find out all details of regulations for DST in Germany. After WWII, in Germany, there were several different zones controlled by different countries and some details for the DST differed between these zones. In one zone, there was a special DST with two hours added for the mid of summer. Then there are reports that there were years with DST even in winter. And so on. All these special regulations make it very difficult to write such a routine. >>Perhaps, but that is not something the government can easily >>legislate nor enforce. (Do you really want the time police [quoted text clipped - 6 lines] > and their own agencies. Some other institutions then might > follow. By "here", I assume you mean Germany, based on what you wrote below. There are no such laws in the US with the exception of some states that have "blue laws" that only apply on Sundays. Your suggestion sounds like a nightmare where you never know what is going to be available when. >>There is nothing sacred about the time zone. > [quoted text clipped - 19 lines] > write such a routine. > This is why any environment worth its salt has these routines available in a standard library. Asking people to forgo a substantial reduction in energy use for the (supposed) convenience of computer programmers is ridiculous. Remember that computers are not an end unto themselves, they are tools here to serve. Ray SignatureThis signature intentionally left blank. >>>It has been shown that daylight saving time results in less >>>energy consumption. [quoted text clipped - 3 lines] > I think that it is clear that the statement above compares daylight saving > time to the lack of daylight saving time. Why would DST result in energy savings? (Not rhetorical, I really don't know) - Oliver >>>>It has been shown that daylight saving time results in less >>>>energy consumption. [quoted text clipped - 6 lines] > Why would DST result in energy savings? (Not rhetorical, I really don't > know) What happens is that people use more lights, etc. at night than in the morning. So moving an hour of daylight from the morning to the evening cuts down on energy use. There are more details at <http://www.energy.ca.gov/daylightsaving.html> if you are interested. Ray SignatureThis signature intentionally left blank. >>I think it is a nutty idea. It creates two goofy anomalies each year >>-- a missing hour and an hour lived twice. Leave the time alone and [quoted text clipped - 13 lines] > > http://www.initiative-sonnenzeit.de/ I really don't think Bush gives a sh.t what his people think let alone Germans. -- LTP :) Roedy Green wrote On 01/10/06 13:27,: >>Does java handle DST automatically ? >>I'm using java Gregorian calendar to get the time/date. Does this [quoted text clipped - 10 lines] > just declare different summer and winter business hours. > Non-compliance would help spread rush hour. Why stop there? Get rid of this silly business of time zones and just run the entire world on UTC. (That way I'll have a good excuse for lying in bed until noon and not returning there until tomorrow morning.) While we're at it, let's reform the calendar: Who needs this arbitrary assortment of irregularly-sized months, anyhow? "Thirty days hath September / The Ides come on the fifteenth day ..." Bosh! And look at all the confusion it's caused, what with the Merkins and Yoorops unable to tell whether 6/6 is June 6 or 6 June. A simple day-of-year count of 1..365 (0..365 in Leap Year) would do just fine. Knowing that today is the 10th and Christmas is the 359th would make it much easier to count those shopping days ... Come to think of it, who really needs a count of years any more? We'll just get all of civil society to switch to counting milliseconds since, since, well I guess we'll just call it Millisecond Zero. "You are charged with serving alcohol to a person under the age of 568036800000, for which you could be sent to prison for as long as 126230400000. How do you plead?" See how much simpler things are when you don't need to specify which of a multitude of confusing units is meant? Roedy, I think you've got something -- and if you'll let me smoke some, too, I won't rat to the narcs. SignatureEric.Sosman@sun.com > Come to think of it, who really needs a count of >years any more? At some point that will have to happen. Years are a very earth-centric measure, hardly suitable for a civilisation spanning many worlds. If someone discovers time is quantised, perhaps we could use time quanta since the big bang, with the high and low order bits deleted. But now there appears to be some doubt there was a big bang after all. We will have to pick an arbitrary starting point, perhaps the adoption of the new calendar. SignatureCanadian Mind Products, Roedy Green. http://mindprod.com Java custom programming, consulting and coaching. Roedy Green wrote On 01/10/06 15:16,: >> Come to think of it, who really needs a count of >>years any more? [quoted text clipped - 8 lines] > We will have to pick an arbitrary starting point, perhaps the adoption > of the new calendar. Don't invent Yet Another Zero; stick with existing precedent: http://www3.sympatico.ca/n.rieck/docs/calendar_time_y2k_etc.html#nov-17-1858 SignatureEric.Sosman@sun.com > Don't invent Yet Another Zero; stick with existing precedent: the old Julian astronomer's calendar epoch is a good one. It does not have religious connotations. Another than might just evolve out of Java is millis since 1970 since programmers learn it as a side effect of debugging date code. These probably is a defacto one that all the satellite and astronomical folk use for high precision work. SignatureCanadian Mind Products, Roedy Green. http://mindprod.com Java custom programming, consulting and coaching. > Roedy, I think you've got something -- and if you'll >let me smoke some, too, I won't rat to the narcs. I consider that sort of crack unacceptably rude, though I notice people using them quite casually without obvious malice. SignatureCanadian Mind Products, Roedy Green. http://mindprod.com Java custom programming, consulting and coaching. Roedy Green wrote On 01/10/06 15:20,: >> Roedy, I think you've got something -- and if you'll >>let me smoke some, too, I won't rat to the narcs. > > > I consider that sort of crack unacceptably rude, though I notice > people using them quite casually without obvious malice. My tongue was quite firmly in my cheek, as I thought would be clear from the rest of my message. If the resulting grimace offended you, I apologize. It's been observed more than once that although Usenet is a good channel for words, it's a poor communicator of "attitude." SignatureEric.Sosman@sun.com > My tongue was quite firmly in my cheek, as I thought > would be clear from the rest of my message. It was transparently clear. -- chris On Tue, 10 Jan 2006 20:20:47 GMT Roedy Green <my_email_is_posted_on_my_website@munged.invalid> waved a wand and this message magically appeared: > > Roedy, I think you've got something -- and if you'll > >let me smoke some, too, I won't rat to the narcs. > > I consider that sort of crack unacceptably rude, though I notice > people using them quite casually without obvious malice. What's the matter, someone take away your sense of humour? Signaturehttp://www.munted.org.uk Anyone that thinks an imaginary deity is going to protect them against earthquakes and hurricanes needs psychiatric help. > Come to think of it, who really needs a count of > years any more? We'll just get all of civil society to [quoted text clipped - 6 lines] > specify which of a multitude of confusing units is > meant? I think it would make more sense to make the "second" the basic unit of time, just like the "meter" is the basic unit of length, or the "litre" is the basic unit of volume. And so the judge would probably not say "under the age of 568036800 seconds", but rather "under the age of 568 megaseconds". But then again, 568 megaseconds seems rather arbitrary, and would only exist for historical reasons related to the drinking age previously being based on the unit of year. So with these new units, probably for convenience, the drinking age would be amended to 500 megaseconds, or 600 megaseconds or some other round number (500 Ms is approximately 15 years and 10 months, 600 Ms is approximately 19 years). For a sense of scale with respect to life expectancy, 80 years is about 2.5 gigaseconds. - Oliver > > I think it is a nutty idea. It creates two goofy anomalies each year > > -- a missing hour and an hour lived twice. Leave the time alone and [quoted text clipped - 3 lines] > Why stop there? Get rid of this silly business of > time zones and just run the entire world on UTC. The reason for stopping there is that DST is an outdated hack that could easily /be/ dropped, whereas any the other options you mention would require that real people adapt to it (which ain't going to happen -- obviously). OTOH, if you are genuinely want to see peoples' local intuitions about times-of-day reflected in the clock, why don't we go back to the idea that the length of an hour/minute/second is dependent on the length of the local day/night (as used to be the case) ? That, very people-centric, idea was dropped several hundred years ago, but only because they didn't have the technology. Nor had they (I think) the sophistication to recognise that they could use two clocks, one for scientific/global purposes (such as astronomy and international trade) and another for local purposes. We do now (for both). So the hour from midnight to 1 AM in Aberdeen should be longer than the corresponding hour in London (at this time of year -- the reverse in summer). I admit that that would introduce some small problems (for instance the local speed limit would change with the time of day ;-), but nothing that couldn't easily be dealt with. -- chris On Wed, 11 Jan 2006 11:55:17 -0000, "Chris Uppal" <chris.uppal@metagnostic.REMOVE-THIS.org> wrote, quoted or indirectly quoted someone who said : >speed limit would change I wonder how soon it will be economical to replace all painted speed limit signs with electronic ones that can be controlled to give sensible numbers depending on visibility, slipperiness, nearby accidents and simply to optimally regulate traffic flow to prevent congestion. You could also have a sign an freeway on ramp to let you know when an open slot was coming, and to ramp up to speed to meet it, or to just stay off the freeway for a while since adding more cars would reduce the aggregate flow. It blows my mind how people insist on driving the speed limit in the slickest ice, on the first snow when people are caught without snow tires, or in pea-soup fog. You sometimes get multiple car pileups as a result. SignatureCanadian Mind Products, Roedy Green. http://mindprod.com Java custom programming, consulting and coaching. Hi , Thanks very much. This is right that JAVA gregorian calendar has got this bug. When DST occurs in Iran ( time 23:59:59 -> 01:00:00) the calendar returns wrong time. But it works when DST occurs in Egypt ( time 23:59:59 -> 01:00:00). > Hi, > Does java handle DST automatically ? > I'm using java Gregorian calendar to get the time/date. Does this > automatically give the local time/date iirrespective of DST or special > handling should be done ? Free MagazinesGet these publications absolutely FREE for up to 12 months. There are no hidden fees and no obligation. Simply choose a title, complete the application form and submit it. Read more ...
|
Reply to this Message
|
 Sign In
 Join
 My Latest Posts My Monitored Threads My Blog My Photo Gallery My Profile My Homepage Start New Thread Enable EMail Alerts Rate this Thread
|
©2008 Advenet LLC Privacy Policy - Terms of Use
This website includes both content owned or controlled by Advenet as well as content owned or controlled by third parties.