- tz - lists.iana.org

DST and OS software
by Chris Carrier Nov. 17, 1996

Nov. 17, 1996

Which is why daylight saving time rules should NOT be incorporated into OS software -- because they continually change. I am wondering how many computers in the European Union (except UK/Ireland) went back to standard time September 29 when the EU didn't fall back until October 27, the former date being programmed into them when the code was written because the last Sunday in September was the rule then.

1 0

Only Microsoft could come up with this one...
by Monty Solomon Nov. 16, 1996

Nov. 16, 1996

Begin forwarded message: Newsgroups: rec.humor.funny From: GeraghtyR(a)scrha.health.qld.gov.au (Geraghty Rob) Subject: Only Microsoft could come up with this one... Keywords: chuckle, computers Approved: funny-request(a)clari.net Date: Sat, 16 Nov 96 4:30:02 EST Lines: 26 Just the other day, the eastern Australian States of New South Wales and Victoria went over to Daylight Saving time. Queensland did not. Windows 95 regards all three as being in one time zone. Consequently, all our email and schedule appointments have gone haywire, because Exchange is adjusting for a daylight saving difference we don't have. (please note that Queensland has only had daylight saving for one summer - about three years ago) I phoned Microsoft to ask whether they planned to fix the bug, but they told me: "It's not a bug." "Why not?" "Because Queensland had daylight saving when we wrote the code." -- Selected by Jim Griffith. MAIL your joke to funny(a)clari.net. Attribute the joke's source if at all possible. A Daemon will auto-reply. Remember: Always give your jokes a descriptive "Subject:" line. Don't use "joke" or "submission" or "joke submission," please. For the full submission guidelines, see http://comedy.clari.net/rhf/

1 0

tzcode1996m.tar.gz and tzdata1996l.tar.gz
by ado Nov. 5, 1996

Nov. 5, 1996

The files ftp://elsie.nci.nih.gov/pub/tzcode1996m.tar.gz and ftp://elsie.nci.nih.gov/pub/tzdata1996l.tar.gz are now available. No functional changes here; the files have simply been changed to make more use of ISO style dates in comments. The names of the above files now include the year in full. --ado

1 0

Re: Clock change orgy of 1996-10-27
by Peter Ilieve Nov. 2, 1996

Nov. 2, 1996

On Nov 2, 4:43am, Chris Carrier wrote: > [I wrote:] > >At least we are making progress here on the cooperation front, > >at least until the French upset things by suggesting they would scrap > >the whole idea. > > What is the status on that? I don't really know at present. There was a report in the Independent (a UK `quality' paper which is unfortunately only available on the Internet via AOL, and possibly only European AOL at that) on 25 October that at least acknowledged that France would have to consult with the rest of the EU about it. It said that `Paris would ask Brussels' meaning ask the EC Commission to be exempted from the Summer Time Directive as soon as possible, either next year or 1998. If it goes into 1998 it will be the eighth Directive which is not yet finalised. I have asked my usual contact at the Commission's Edinburgh office but she can't say much about the state of this new Directive. The Commission has produced a proposal which continues the current rules for 1998--2001. It is now the Council of Ministers' turn to reach a common postion, and then approve that as a Directive. They haven't shown much inclination so far and my contact points out that she can only speak for the Commission, not the Council. The new Directive is supposed to be approved by 1 Jan 97 but I think that is vanishingly unlikely even without this new French move. The Independent report said: `After three months of discussion and two parliamentary reports, it is clear that France intends to rid itself of a device that its experts say no longer saves fuel or manpower, upsets farmers, reduces milk yields by almost one-third and disrupts children's body-clocks.' Polls show 70% of people against changing the clocks, but an even split on the question of which time to use all the year round. Farmers and families with children favour permanent summer time (UTC+2), business and transport organisations favour winter time (UTC+1). The French Senate said winter time would be a good compromise with the rest of the EU, obviously assuming it continues to change its clocks, as France would have the same time as the UK in the Summer and the same as Germany in the winter. Can any French reader of this list give any references to the two official reports, possibly even a quick summary? Don't be too surprised about the importance given to milk yields. Cattle are an important determinant of EU policy, as witnessed by the current BSE row. :-) Peter Ilieve peter(a)aldie.co.uk

1 0

Clock change orgy of 1996-10-27
by Chris Carrier Nov. 2, 1996

Nov. 2, 1996

More thoughts on the clock change orgy of 1996-10-27: >"Rob Jones 4-6597 MR Sys" > INTERNET:jones@mr.med.ge.com wrote: >Anyone willing to venture a guess as to the number of clocks involved ? >(including uwaves, vcrs, etc...) 11 clocks at my home, but only 3 people. In terms of number of people, the 1996-10-27 fall back probably was about equal to 1991-09-15, when China changed its clocks last; however, in terms of clocks changed, the 1996-10-27 event clearly set an all-time record given the higher development level and more consumer goods available in the US and EU; in my household there were five timing devices to be changed per person, and I imagine in China it would probably be closer to one device per five persons. Valery Aronov > INTERNET:valery@ind.tansu.com.au wrote: >My own personal view is that we all should move the start of the workday >and settings of our alarm-clocks if people really want to save energy or >to wake up with sun. No clock adjustment at all. The only inconvinience >I could imagine: two sets of working hours for offices - a summer set and >a winter set. So a DST adjustment will mean switching from one set to >another. Much less problems then now I think. Do you know if this >suggestion was ever considered? I would be very much in favor of this. One interesting thing in our office is that we have flextime and management is trying to get rid of it and to have everyone work 8-17. Most people's day begins around 7:45, and I would say about 15% of the office is on the equivalent of "Daylight Extra" with a workday beginning at 7. We have one woman who works 6-15 and who therefore could be considered on "Double Daylight Extra." I also have an interesting story about my former office, in Sacramento, the capital of California. When I worked there a decade ago, we had people in the office from 7:00 to 20:00. Full time employees were allowed to choose when they could come into work provided they worked an 8 hours within those limits. Sacramento is located in a valley in which summer days are VERY hot (average 35c and it's gone as high as 44c) but nights are quite pleasant, with an overnight low of 15c. I reacted to this by going on what might be called "Darkness Extra"; during the months when DST was in effect I set my work hours for 11:00-19:45, the equivalent of two hours of *reverse* DST to let me enjoy the cooler evenings. Alas, eventually the office went back to normal hours (first people in at 7:00; last out by 17:45) and I had to as well. Peter Hullah > INTERNET:Peter.Hullah@eurocontrol.fr wrote: >Isn't it about time the world came to realise that we don't live by the >sun anymore and that spreading our sunlight more-or-less evenly around >noon no longer has any sense? Strongly disagree; if you work an 8 hour day on a day shift, half of it usually is before noon so the work day starts at 8 which makes the average wakeup time on workdays between 6 and 7, which is when the sun ought to rise. In June, therefore, DST makes sense and in October it does not. >I reckon everyone should add two hours >to their current standard time and stick to it all year round. I've always believed that what can be done an hour before sunrise is much pleasanter to do an hour after sunrise. >In answer to those who say "Just get up and go to bed two hours earlier" >we all live in a community, and one family stepping out of line two >hours is not on. Actually I have heard of a few cases of families with pre-school children where the sole/or both adults work the night shift and the kids follow suit, although this has to change when the kids go to school. Garrett Wollman > INTERNET:wollman@lcs.mit.edu wrote: >...in a city like Boston ...retail sales in the major pedestrian >districts and downtown areas drops off about 25% when the sunset is >moved from 5:45 (after work) to 4:45 (before work lets out). I wonder how much of that activity is transferred to the lunch period, breaks, or before work? Probably most of it. >As far as I'm concerned, we would be better off observing Atlantic time >year-round. Had Boston (and all of New England) not been near the largest city and capital of the US, it might have. Consider there are many cities (Cincinnati, Louisville, Chatannooga, Tallahassee) which astronomically belong in the Central zone observe Eastern time, and large areas are on Central time when they astronomically belong on Mountain (most of the six states in a north- south row between North Dakota and Texas). Alex Livingston >INTERNET:alex@agsm.unsw.edu.au wrote: >Reply to: Re: did the Holiday Is observe DST after the rest of Queensland >stopped? >Australia's vying powers-that-be unfortunately have a very childish, >ingnorant and it seems almost purely politically motivated approach to >daylight saving. I have never tried to get to the bottom of how the dates >for the clock changes were chosen but they are *particularly* bad choices. >The switch made recently by the south-eastern mainland of the country comes >at a time when the sun is rising only about 20 minutes later in the day (in >absolute terms) than the earliest it ever rises, but about 2 hours earlier >than the latest it ever rises. This means that the earliest sunrise of the >year (in clock terms) occurs the day before clocks are advanced. Just like most places in the United States before 1987 when almost nobody sprang forward before the last Sunday in April, wartime and 1974-5 excepted. >On the other hand, when the switch back to so-called standard time occurs at >the end of March, the sun is rising about 1 hour and 20 minutes later (in >absolute terms) than the earliest sunrise of the year and only about an >hour earlier than the latest... I've always thought the logical time to begin and end DST was at some time near the equinox when the sun is at least 1 hour earlier than the latest it does by standard time, and at least 1 hour later than the earliest it does by daylight time. DST should not be used short of an emergency during the time near winter solstice when the sun is rising <1 hour earlier than the latest it ever does by standard time, and if there is <2 hours between the earliest and latest sunrise, you don't have enough daylight to save to warrant DST! Peter Ilieve > INTERNET:peter@aldie.co.uk wrote on the subject: did the Holiday Is observe DST after the rest of Queenslan stopped? >Of my trio of islands, Eigg and Canna are also privately owned (I don't >know about Muck). I fail to see why owning land that is bounded by sea >rather than land bounded by a more normal fence makes the owners any less >subject to the Summer Time Act 1972 than I am. >Don't you have laws in Australia? :-) Setting up private and extra-legal times has a long history in the US. I remember Block Island, RI, went on DST in November 1973 two weeks after they fell back. I know people who were in Indiana in the 1950s-60s who remember the post office being an hour ahead or behind of the rest of town, an unincorporated area of a county being on a different time than a county seat, and so on. When the Uniform Time Act of 1966 was passed, Indiana passed a law requiring one clock in each public building to observe UTAT and marked "Official Time" so localities could go about their merry way. >People seem reluctant to give up summer time, >as shown by popular support in the rest of Europe for changing from >September to October, but don't seem to think of demanding an earlier >start. I wonder if anyone took a poll on the popularity of summer time in the last week in October 1996, when Europe was actually on summer time, as opposed to earlier polls showing majorities in favor. >At least we are making progress here on the cooperation front, >at least until the French upset things by suggesting they would scrap >the whole idea. What is the status on that? Actually the situation in France (and Spain) is that they are already on year round daylight time since their 'standard' is GMT+1 instead of the GMT zone that (nearly all of) France and Spain are astronomically in. When they talk about 'summer time' what they really mean is DOUBLE summer time. It should be noted when the UK was on double DST during WW2 it always ended the second weekend in August or earlier except in 1944, and even then it ran only to mid-September. I can see objections to DDST on October 26th having more than a little merit.

1 0

Re: did the Holiday Is observe DST after the rest of Queensland stopped?
by Nathan Myers Nov. 1, 1996

Nov. 1, 1996

> [Somebodies] unfortunately have a very childish, > ingnorant and it seems almost purely politically motivated approach to > daylight saving. ... But I digress... Daylight saving has always been a purely political phenomenon, an attempt to legislate reality. (Consider observance of summer time in tropical countries!) It is, and always will be, subject to all the foolishness that any purely political abstraction is. Tracking that foolishness in software is as thankless as it is necessary. Nathan Myers ncm@@cantrip.com

1 0

Re: did the Holiday Is observe DST after the rest of Queensland stopped?
by peter＠aldie.co.uk Nov. 1, 1996

Nov. 1, 1996

> >I know I am a long way away from the scene of the action, but from looking > >at the map this looks about as likely as Eigg, Muck or Canna > >having different DST rules from the rest of the UK; ie. not very. :-) > > > >My Times Concise Atlas lists Hayman and Lindeman as being on either > >side of Whitsunday Island. My Britannica Atlas doesn't list either > >but gives 20.17S 148.59E for Whitsunday. Neither has any trace of > >Hamilton. > > I have actually *been* to Hamilton Island, and a close friend of mine > recently had a holiday on Hayman Island. The three islands are probably > owned entirely by holiday (that's "vacation" to you North Americans) > operators and probably made autonomous decisions as to what time to keep on > their properties. I have heard or read more than once that some resort > islands off the coast of Queensland chose to keep observing daylight-saving > time even after Queensland ceased to. It seems that no only does each state > decide independantly if and when to observe daylight saving, but > commercially run holiday resorts can also do it their way. And then there's > Lord Howe Island (what a disaster!). I have reason to doubt, by the way, > that "Holiday Islands" is an officially gazetted name: it's not one I'm > familiar with; as far as I'm aware, all these islands are part of the > Whitsunday Islands group. Of my trio of islands, Eigg and Canna are also privately owned (I don't know about Muck). I fail to see why owning land that is bounded by sea rather than land bounded by a more normal fence makes the owners any less subject to the Summer Time Act 1972 than I am. Don't you have laws in Australia? :-) Seriously, should the tz database document legal time, or should it also document other sorts of `custom and practice' time? If the latter I'm sure I can promulgate an Aldie time to be observed on my patch that would break something in the existing code. :-) Another nearby island, Rum (Rhum if you believe the Ordnance Survey) is also owned by a single organisation and is largely given over to research into Red Deer. Maybe I should suggest to Scottish Natural Heritage that their scientific work justifies keeping UTC all year round. > Australia's vying powers-that-be unfortunately have a very childish, > ingnorant and it seems almost purely politically motivated approach to > daylight saving. ... But I digress... We have the same thing. People seem reluctant to give up summer time, as shown by popular support in the rest of Europe for changing from September to October, but don't seem to think of demanding an earlier start. At least we are making progress here on the cooperation front, at least until the French upset things by suggesting they would scrap the whole idea. Peter Ilieve peter(a)aldie.co.uk

2 1

Moscow DST correction
by Андрей Чернов Oct. 31, 1996

Oct. 31, 1996

Diffs related to europe 7.38 *** europe.orig Thu Jul 18 00:54:30 1996 --- europe Sun Sep 29 19:20:56 1996 *************** *** 852,858 **** Rule Russia 1921 only - Oct 1 0:00 0 K Rule Russia 1981 1984 - Apr 1 0:00 1:00 D Rule Russia 1981 1983 - Oct 1 0:00 0 K ! Rule Russia 1984 max - Sep lastSun 2:00s 0 K Rule Russia 1985 max - Mar lastSun 2:00s 1:00 D # These are for backward compatibility with older versions. --- 852,859 ---- Rule Russia 1921 only - Oct 1 0:00 0 K Rule Russia 1981 1984 - Apr 1 0:00 1:00 D Rule Russia 1981 1983 - Oct 1 0:00 0 K ! Rule Russia 1984 1995 - Sep lastSun 2:00s 0 K ! Rule Russia 1996 max - Oct lastSun 2:00s 0 K Rule Russia 1985 max - Mar lastSun 2:00s 1:00 D # These are for backward compatibility with older versions. -- Andrey A. Chernov <ache(a)nagual.ru> http://www.nagual.ru/~ache/

6 17

Re: did the Holiday Is observe DST after the rest of Queensland stopped?
by alex＠agsm.unsw.edu.au Oct. 31, 1996

Oct. 31, 1996

>> According to the current tz database, Queensland stopped observing DST >> after summer 1991/1992. But the IATA SSIM (which I'm currently >> slogging through, thanks to Gwillim Law) says that the Holiday Islands >> (Hayman, Lindeman, Hamilton) observed DST through 1993/1994. Does >> anyone know the true story here? > >I know I am a long way away from the scene of the action, but from looking >at the map this looks about as likely as Eigg, Muck or Canna >having different DST rules from the rest of the UK; ie. not very. :-) > >My Times Concise Atlas lists Hayman and Lindeman as being on either >side of Whitsunday Island. My Britannica Atlas doesn't list either >but gives 20.17S 148.59E for Whitsunday. Neither has any trace of >Hamilton. I have actually *been* to Hamilton Island, and a close friend of mine recently had a holiday on Hayman Island. The three islands are probably owned entirely by holiday (that's "vacation" to you North Americans) operators and probably made autonomous decisions as to what time to keep on their properties. I have heard or read more than once that some resort islands off the coast of Queensland chose to keep observing daylight-saving time even after Queensland ceased to. It seems that no only does each state decide independantly if and when to observe daylight saving, but commercially run holiday resorts can also do it their way. And then there's Lord Howe Island (what a disaster!). I have reason to doubt, by the way, that "Holiday Islands" is an officially gazetted name: it's not one I'm familiar with; as far as I'm aware, all these islands are part of the Whitsunday Islands group. Australia's vying powers-that-be unfortunately have a very childish, ingnorant and it seems almost purely politically motivated approach to daylight saving. I have never tried to get to the bottom of how the dates for the clock changes were chosen but they are *particularly* bad choices. The switch made recently by the south-eastern mainland of the country comes at a time when the sun is rising only about 20 minutes later in the day (in absolute terms) than the earliest it ever rises, but about 2 hours earlier than the latest it ever rises. This means that the earliest sunrise of the year (in clock terms) occurs the day before clocks are advanced. On the other hand, when the switch back to so-called standard time occurs at the end of March, the sun is rising about 1 hour and 20 minutes later (in absolute terms) than the earliest sunrise of the year and only about an hour earlier than the latest. (The figures quoted apply roughly to the latitude of Adelaide or Sydney but will not differ hugely within the range of latitudes where daylight saving is now observed in Australia.) If the populace can live with the sun rising as late at it does at the end of the daylight-saving period then it should surely be happy to live with it at the beginning as well, in which case it could be started a full 7 weeks earlier (around September 9). And in all of this the possibility of coordinating with other southern-hemisphere countries which also observe daylight saving (even neighbours like New Zealand and Vanuatu) doesn't seem to have flickered into any law-maker's consciousness. (That would involve acknowledging that someone else thought of daylight saving first and rob them of the kudos of coming up with such a great "original" idea.) After all, even the states can't coordinate. But I digress... _______________ Alex LIVINGSTON Macintosh Support Computing & IT (C&IT) Australian Graduate School of Management (AGSM) The University of New South Wales (UNSW) Sydney NSW 2052 AUSTRALIA E-mail : alex(a)agsm.unsw.edu.au; cit(a)agsm.unsw.edu.au (C&IT) Facsimile: +61 2 9313-7279 Telephone: +61 2 9931-9264

1 0

Re: did the Holiday Is observe DST after the rest of Queensland stopped?
by Peter Ilieve Oct. 31, 1996

Oct. 31, 1996

> According to the current tz database, Queensland stopped observing DST > after summer 1991/1992. But the IATA SSIM (which I'm currently > slogging through, thanks to Gwillim Law) says that the Holiday Islands > (Hayman, Lindeman, Hamilton) observed DST through 1993/1994. Does > anyone know the true story here? I know I am a long way away from the scene of the action, but from looking at the map this looks about as likely as Eigg, Muck or Canna having different DST rules from the rest of the UK; ie. not very. :-) My Times Concise Atlas lists Hayman and Lindeman as being on either side of Whitsunday Island. My Britannica Atlas doesn't list either but gives 20.17S 148.59E for Whitsunday. Neither has any trace of Hamilton. The whole of that coastline is peppered with islands just off the coast. Peter Ilieve peter(a)aldie.co.uk

1 0

did the Holiday Is observe DST after the rest of Queensland stopped?
by Paul Eggert Oct. 31, 1996

Oct. 31, 1996

According to the current tz database, Queensland stopped observing DST after summer 1991/1992. But the IATA SSIM (which I'm currently slogging through, thanks to Gwillim Law) says that the Holiday Islands (Hayman, Lindeman, Hamilton) observed DST through 1993/1994. Does anyone know the true story here? If the IATA is right, we'll need a new entry Australia/Hamilton to handle the anomaly. I know that those islands exist (they're mentioned in <URL:http://www.lonelyplanet.com/dest/aust/gbreef.htm>, which says Hamilton has the most people) but none of them are in my Rand McNally Goode's World Atlas or in Shanks's International Atlas, so I'll need further info to find Hamilton's latitude and longitude.

1 0

Re: Clock change orgy of 1996-10-27
by Rob Jones 4-6597 MR Sys Oct. 30, 1996

Oct. 30, 1996

Anyone willing to venture a guess as to the number of clocks involved ? (including uwaves, vcrs, etc...) 11 clocks at my home, but only 3 people. -- Rob Jones On Oct 30, 12:13am, Chris Carrier wrote: > Subject: Clock change orgy of 1996-10-27 > On 1996-10-28 05:04 GMT I posted: > > more people changed time today, 1996 October 27, than on any other date > in history. > > To which Paul Eggert commented: > > >Could be. 1991-09-15 is another candidate; it's the last time that > >China changed its clocks. > > Good point. The magnitude of the clock changing orgy, in terms of number of > people, would have been about the same. Speaking in very broad general terms, > the population of the US (ex. eastern IN, HI, and AZ) is about 250 million, > Mexico 100 million, EU 350 million, Russia 150 million, totalling 850 million, > plus or minus 100 M. The population of China is somewhere in the range of > 1100-1200 million, so it would take another 300 million people, in non-EU > Europe and the Central Asian republics of the CIS, changing time on 1996 > October 27 to exceed the Chinese 1991 record. Sounds like a close thing. > > I furthermore posted an: > > Editorial comment on the above: it should have been on the last Sunday in > September not October! >

1 0

Clock change orgy of 1996-10-27
by Chris Carrier Oct. 30, 1996

Oct. 30, 1996

On 1996-10-28 05:04 GMT I posted: more people changed time today, 1996 October 27, than on any other date in history. To which Paul Eggert commented: >Could be. 1991-09-15 is another candidate; it's the last time that >China changed its clocks. Good point. The magnitude of the clock changing orgy, in terms of number of people, would have been about the same. Speaking in very broad general terms, the population of the US (ex. eastern IN, HI, and AZ) is about 250 million, Mexico 100 million, EU 350 million, Russia 150 million, totalling 850 million, plus or minus 100 M. The population of China is somewhere in the range of 1100-1200 million, so it would take another 300 million people, in non-EU Europe and the Central Asian republics of the CIS, changing time on 1996 October 27 to exceed the Chinese 1991 record. Sounds like a close thing. I furthermore posted an: Editorial comment on the above: it should have been on the last Sunday in September not October! To which jhuggins(a)comsys-inc.com wanted to know: >I am a rather new subscriber to this newsgroup. Could you explain your >editorial remark about September versus October. I believe DST should end at the end of September because in October (in the North Temperate Zone, where most but not all of the people on this list live) there is simply not enough daylight in the morning when DST is used. The following examples for sunrise on the last day of DST this year (October 26) in certain parts of the United States are used for illustration: Dallas, TX = 7:45 Seattle, WA = 7:45 Oklahoma City = 7:46 Atlanta, GA = 7:51 Salt Lake City = 7:51 Russell, KS = 7:55 Cincinnati = 8:00 Midland, TX = 8:00 N. Platte, NE = 8:06 Louisville, KY = 8:02 Boise, ID = 8:12 Kalispell, MT = 8:13 Marquette, MI = 8:22 Williston, ND = 8:30 And since the Europeans and Russians have adopted our bad habit: Bonn.de = 8:12 Dublin.ie = 8:11 London.uk = 7:42 Madrid.es = 8:37 Moscow.ru = 8:22 Paris.fr = 8:27 St. Pete.ru = 9:04 Seville.es = 8:42 Going back on Standard Time on the last Sunday in September would make each of these sunrises an hour earlier; clearly a much better situation for public convenience and the safety of children on their way to school; in fact a DST ending of Labor Day (the American day, Mon => Sep 1) would serve schoolchildren even better. From: Garrett Wollman > INTERNET:wollman@lcs.mit.edu had a: >Comment on the editorial comment: who cares when the sun rises? Comment on the comment on the comment: Me, and anyone else who is sick and tired of getting up in the dark. Generally more DST is favored when it's not being used; I remember polls in December 1973 in the US in which a 3:1 majority was in favor of year-round DST. Then on 1974-01-06 we actually did go on DST and the 3:1 majority in favor quickly became a 3:1 majority against. The original advocates of DST did not favor DST in October. When Franklin semi-seriously suggested DST in the 18th century, he suggested that DST be used to advance the sunrise to when most people wake up. (He must have been thinking of city people; the majority lived on farms and woke up with the sun during his lifetime, whether sunrise was 3:00 or 9:00). When William Willett proposed it in 1905, he suggested advancing time 80m (20m on each of the first four Sundays) in April and ending it be turning the clock back 20m on each of the first four Sundays in SEPTEMBER, which would have meant reversion to standard time not later than September 28. My own personal view is that I don't really care (most of the time) if sunset is 14:00 or 23:59, I just want it light when I have to get up to go to work! Chris Carrier

2 1

Moscow time & public transit in CIS
by Chris Carrier Oct. 29, 1996

Oct. 29, 1996

According to a friend of mine who rode the Trans-Siberian Railroad from Moscow to Irkutsk in 1995, public air and rail transport in Russia after the USSR breakup still follows Moscow time, no matter where in Russia it is located; he mentioned getting off the train in Irkutsk at 03:00 railway time to see the sun up and people going to work at 08:00 local time, and when flying back having to wait an unexpected 5 hours in the airport for a plane which he thought was leaving on local time when it was really leaving on Moscow time. In those parts of the USSR which were in Europe and outside the Russian Federation, public transport runs on local time (which has a standard of GMT+2, moving back with Moscow in 1991 when they were still Soviet but staying on GMT+2 when Moscow standard time moved back to GMT on 1992-01-18, right after the breakup). Don't know what the Central Asian CIS republics are doing about transport time; anyone know? Chris Carrier

1 0

problem with tzcode96l
by Todd C. Miller Oct. 29, 1996

Oct. 29, 1996

Hi, we're trying to get OpenBSD to pass the NIST POSIX test suite and it tries to do a mktime() with TZ=dst0std,J057,J059 with a time of Tue Feb 26 01:59:59 1991 (which should return 667533600). mktime() from tzcode96l (and earlier versions) returns -1 with that TZ setting. There seems to be a problem with the end of daylight savings time specification. If you comment out the getrule() on line 751 it at least keeps mktime from returning -1 :-) I really don't understand the code in question. Any clues would be greatly appreciated. thanks, todd --- Todd C. Miller Sysadmin--University of Colorado Todd.Miller(a)cs.colorado.edu

2 7

Re: Non-EU Europe, record set today
by peter＠aldie.co.uk Oct. 29, 1996

Oct. 29, 1996

> I would like to ask: given that the European Union has extended summer time > until October 27, and that Russia did the same, can we assume that all > European countries that are *not* in the EU and those former republics of the > ex-Soviet Union that observed DST also delayed their ending date this year > until the end of October? No. Although it won't affect the overall numbers much, it appears that Latvia at least didn't take part in this historic mass orgy of clock changing. I was checking with a relative in Estonia about the IATA manual, which gives a different time of change for Estonia to that which he had told me before, and he said unprompted that he knew Latvia had changed on the last Sunday in September. On the IATA manual and its claim that Estonia changes at 23:00 UTC he said: `I do not [know] exactly but there are some little different (confusing) rules for International Air and Railway Transport Schedules conversion in Sunday connected with end of summer time in Estonia. Anyway Estonia will change 03:00 local time.' This may be a more general problem. Does this IATA manual specifically say that it gives times of change for civil time, or does it describe some special IATA rules? Peter Ilieve peter(a)aldie.co.uk

2 1

Non-EU Europe, record set today
by Chris Carrier Oct. 27, 1996

Oct. 27, 1996

I would like to ask: given that the European Union has extended summer time until October 27, and that Russia did the same, can we assume that all European countries that are *not* in the EU and those former republics of the ex-Soviet Union that observed DST also delayed their ending date this year until the end of October? Another thought. Given that most of the USA, Mexico, the EU and Russia all changed back on the same date, I believe that some sort of record was set today: more people changed time today, 1996 October 27, than on any other date in history. (The combined population of the four nations/unions mentioned above being in the range 800-900 million). Editorial comment on the above: it should have been on the last Sunday in September not October! Chris Carrier

2 1

Re: multithreading and timezones
by Nathan Myers Oct. 22, 1996

Oct. 22, 1996

Valery Aronov <valery(a)ind.tansu.com.au> wrote: > How is it possible to manage current zone information in MT-safe fashion? From: Paul Eggert <eggert(a)twinsun.com> > Write and use a thread-safe version of the tz library. > > This requires changing the interfaces to localtime, gmtime, tzset, etc. > (localtime_r is not enough.) One must also make the locale a > parameter, for strftime at least. And I'm afraid that this will > require going into the guts of the C library, which will make it hard > to do portably. > > On my (long) list of things to do is to make a thread-safe version of > the glibc version of the tz library. Once you know you're dealing > with that particular library, the job becomes tractable. The non-re-entrancy, and the globalness, of the time functions has long bothered me. With the C locale itself global, there's not much to be done about that. The (draft) standard C++ locale is not (necessarily) global, and supports user extensions such as time zone handling. Now would be a good time to start defining a standard time zone facet for the C++ locale, so that it can be made official at the next round, in a few years. The glibc locale functions being worked on by Ulrich Drepper are designed to support a re-entrant (C) interface. We should co-ordinate with him to get the time zone code integrated into his architecture. Please, let's get the interface right this time. Nathan Myers ncm(a)cantrip.org

1 0

Solaris multithreading and timezones
by Valery Aronov Oct. 21, 1996

Oct. 21, 1996

We are moving our software towards multithreading (SunOS 5.5.1). One of the many concerns is the MT-unsafety of tzset() function. It is said to be MT-unsafe in man pages. It is partly due to the keeping references to TZ in the environment variable, I guess. Our software needs to deal with more than one timezone in one process. The only way I know to implement it is to change the value of environment variable TZ and call tzset() every time I want to switch to a new timezone. External variables tzname and others also makes it problematic to maintain access to them: other vendors' software we are using might use their own policy of access. How is it possible to manage current zone information in MT-safe fashion? I would appreciate any help. Valery

3 3

c-common.c
by ado Oct. 18, 1996

Oct. 18, 1996

The file ftp://elsie.nci.nih.gov/pub/c-common.c is now available; this is a replacement for c-common.c in gcc (either 2.7.2 or 2.7.2.1) that does strftime format checking (with special warning for formats that generate two digits worth of years and thus may be especially subject to year 2000 challenges). If anyone has feedback before I ship it off to the GNU folk, I'd love it! --ado

3 3

[Karst Koymans: bin/1740: Setting TZ to unusual values c]
by Garrett Wollman Oct. 9, 1996

Oct. 9, 1996

I don't know whether this is a BSD-specific problem or exists in the current tzcode, but I figured I might forward it here in case it does. -GAWollman ------- start of forwarded message (RFC 934 encapsulation) ------- Message-Id: <199610082133.OAA28831(a)freefall.freebsd.org> X-Send-Pr-Version: www-1.0 X-Loop: FreeBSD.org Precedence: bulk From: Karst.Koymans(a)phil.ruu.nl Sender: owner-bugs(a)freebsd.org To: freebsd-gnats-submit(a)freebsd.org Subject: bin/1740: Setting TZ to unusual values crashes applications using localtime Date: Tue, 8 Oct 1996 14:33:13 -0700 (PDT) >Number: 1740 >Category: bin >Synopsis: Setting TZ to unusual values crashes applications using localtime >Confidential: no >Severity: non-critical >Priority: low >Responsible: freebsd-bugs >State: open >Class: sw-bug >Submitter-Id: current-users >Arrival-Date: Tue Oct 8 14:40:02 PDT 1996 >Last-Modified: >Originator: Karst Koymans >Organization: Utrecht University >Release: 960501 snapshot >Environment: >Description: Setting TZ to unusual values like "Europe" crashes all applications using the localtime routine. This is because the tzload routine happily opens any file, even directories, as valid timezone files. "Europe" happens to be a directory under /usr/share/zoneinfo. >How-To-Repeat: TZ=Europe date >Fix: >Audit-Trail: >Unformatted: ------- end -------

1 0

patch to zic.8 to document format of UNTIL field
by Paul Eggert Oct. 8, 1996

Oct. 8, 1996

From: "Rob Jones 4-6597 MR Sys" <jones(a)mr.med.ge.com> Date: Tue, 8 Oct 1996 12:23:18 -0500 In the tzdata files when the UNTIL field is just a year (without a month day) specified, does that mean up to and including that year, or just up to ? Somehow that didn't get documented. Here's a draft patch. In practice I use vague dates like `1967' when I don't know exactly when the rules changed, and a more precise time stamp like `1967 Jan 1 00:00' if I know that's the exact time -- even though both inputs mean the same thing to `zic'. =================================================================== RCS file: RCS/zic.8,v retrieving revision 1995.7 retrieving revision 1995.7.1.1 diff -c -r1995.7 -r1995.7.1.1 *** zic.8 1995/10/30 14:58:15 1995.7 --- zic.8 1996/10/08 18:31:38 1995.7.1.1 *************** *** 304,309 **** --- 304,312 ---- If this is specified, the time zone information is generated from the given GMT offset and rule change until the time specified. + The month, day, and time of day have the same format as the IN, ON, and AT + columns of a rule; trailing columns can be omitted, and default to the + earliest possible value for the missing columns. .IP The next line must be a .q continuation

1 0

Antarctic update
by Chris Carrier Oct. 7, 1996

Oct. 7, 1996

Ethan Dicks > INTERNET:erd@mcmsun5.mcmurdo.gov has informed me: >BTW, the data for the British Base Rothera is out of date. Last >year (1995) was the last winter-over season. It's now a summer only station.

1 0

Re: Recent changes to time zones reported by IATA
by Paul Eggert Oct. 6, 1996

Oct. 6, 1996

Date: Fri, 04 Oct 1996 19:34:39 -0500 (EST) From: Gwillim Law 919-852-4294 <LAW(a)encmail.encompass.com> I thought the tz mailing list would want to see these changes, and compare them to those reported by other sources. That is a very helpful contribution; thanks. I compared the IATA changes with the changes reported by other tz contributors, and have the following comments. Later, I'll send out a draft patch reflecting these comments, as well as all the other comments received so far. * Armenia stopped using DST this year, according to Edgar Der-Danieliantz <edd(a)AIC.NET> (1996-05-04). Did the IATA catch this change? * Azerbaijan data from the IATA disagrees with both Shanks and USNO. I guess that the IATA is right, and that we're missing some transitions between 1991 and 1996. Azerbaijan apparently enjoys the distinction of being the first Asian country to use the European Union's DST rules (including switching at 0100 UTC). * Brazil - You wrote: three states (Alagoas, Sergipe, and Tocantins) will begin using DST, probably on October 13, 1996. Could you give more details about this? The entries that you sent made it look like only the DST dates were adjusted. As far as I can tell from other sources, DST is observed uniformly in Brazil, even in the far north where the hemisphere is wrong. Perhaps these states used to be 2 hours behind UTC, and are now 3? This would agree with the map in <URL:http://www.bancobrasil.com.br/bbtur.ing/bbttime.htm>, but so far I've discounted that map, not only because no other sources have UTC-2 Brazilian mainland, but also because its line between UTC-2 and UTC-3 is suspiciously straight. Also, some of the beginning and end dates for DST were adjusted by a week. You gave the change only for BR1; were the dates also changed for other zones in Brazil? It would be odd if the different parts of Brazil had different DST dates. All the Brazilian transitions that we know of since 1989 are consistent with the rules Oct Sun>=10 00:00 and Feb Sun>=10 00:00; I'll propose a change to the tz rules accordingly. * Iran The tz database had Iran not using DST since 1980. For lack of better info I'll assume they readopted DST in 1996. Iran now apparently moves its clocks approximately on the equinoxes -- I think it's the only country to do this. * Kazakhstan - you wrote: Kazakhstan, formerly unified on UTC+6 with DST, will divide into three zones. They will observe UTC+4, 5, and 6, all with DST. The western zone (UTC+4) consists of Aktau, Atyrau, and Uralsk oblasts, and the central zone is Aktyubinsk oblast. Andrew Evtichov <evti(a)chevron.com> (1996-04-13) wrote that Kazakhstan has had multiple time zones for quite some time: at least 2 ``always'' and 3 now. I'd guess that the IATA SSIM is just catching up with this fact. Unfortunately, I misparsed his earlier message and thought that Kazakhstan had just 2 time zones now; he actually wrote that it has 3. I'll add a new zone for Aqtobe (formerly Aktyubinsk) and rename Asia/Aktau to Asia/Aqtau since `Aqtau' name seems to be the usual English rendering of the Kazakh name. * Lebanon. Lebanon returned from DST to Standard Time on September 28, 1996, instead of October 26 as previously announced. Shanks has Lebanon switching on May 1 and Oct 16 every year. Evidently that has changed more than once since 1991. For lack of better info I'll propose a patch so that Lebanon switches on the last Sunday in March and the last Sunday in September, starting 1996. * Libya if I'm right, the first actual time change was on September 30, 1996, when clocks were set back an hour. That sounds right: since 1959, Libya has switched back and forth between straight UTC+2 and UTC+1 with DST. * Mexico - the IATA and tz agree. * Mongolia The IATA says that changes occur at local midnight, whereas Shanks says they occur at 02:00 standard time. Also, the IATA says the fall transition is the 4th Friday (!) in October, whereas tz said it was the last Sunday in September. I'll assume that the IATA is right, and that the rules changed in 1996, but we're undoubtedly missing some other changes between 1985 and 1995. * Paraguay Paraguay will return from DST to Standard Time on March 1, 1997, instead of February 23 as previously announced. Shanks has this transition occuring on April 1. For lack of better info, I'll assume they changed the rules from April 1 to March 1 in 1996. * Portugal Portugal (excluding Azores and Madeira) has switched from UTC+1 to UTC (both using DST). There's no indication of when the change occurred. Martin Bruckmann <martin(a)ua.pt> (1996-02-29) reported that the change occurred this spring, by Portugal not changing its clocks. * Russia Russia's fall transition used to be the last Sunday in September, but it's changed to the last Sunday in October (at 02:00 standard time, as before). Did the IATA catch this change? Presumably this change also applies to Kazakhstan (I just checked the ricc.kz time server in Alma-Ata, and they're still on DST) as well as Kirgizstan. While we're on the subject, in the absence of any better info, the current tz database assumes that Belarus, Estonia, Latvia, Lithuania, and Moldova are moving their clocks on the last Sunday in October at 02:00 standard time, starting this year (in previous years they used September). Since the EU, Russia, and Ukraine are all known to be making a similar switch, it wouldn't be surprising if the other countries are switching too; but we have no info. * Sri Lanka Sri Lanka has changed from UTC+5:30 to UTC+6:30 year-round. There's no indication of when the change occurred. It occurred 1996-05-25 00:00 local time, according to LankaWeb <URL:http://www.virtual-pc.com/lankaweb/news/items/240596-2.html>. * Ukraine Ukraine will begin and end DST at 3:00 am local time, rather than midnight as it used to. (The data make it clear that this is 3:00 local standard time.) For most of Ukraine, this makes sense, since it's the EU rules. However, for Crimea, it doesn't make sense, since it's neither the EU rules (which operate at 4:00 local standard time) nor the Moscow rules (which operate at 2:00 local standard time). Apparently the Ukrainians and the Crimeans are having a little tussle; there was a news story about this by Itar-Tass <URL:http://www.itar-tass.com/tassnews.htm> at 09-18 00:43 entitled ``Crimea ponders over problem of time zone'' but unfortunately I don't have a copy of the story. For now I'll assume Crimea was using Moscow rules until this year, and is as IATA reports afterwards. Shanks says all of Ukraine switches at 2:00 standard time, not midnight. For lack of better info I'll assume that this was true through 1995.

1 0

Spring Forward, Fall Back -- but not just yet
by Monty Solomon Oct. 6, 1996

Oct. 6, 1996

Excerpt from RISKS DIGEST 18.50 ------------------------------ Date: Tue, 1 Oct 1996 12:02:10 +0200 From: minow(a)apple.com Subject: Spring Forward, Fall Back -- but not just yet The biannual daylight savings time confusion began this weekend in Sweden. When Sweden joined the EU (Common Market), it changed the fall changeover from the last weekend in September to the last weekend in October to conform with the rest of Europe. Unfortunately, a few hundred thousand Windows '95 machines were not informed of the changeover and, followed pre-programmed instructions, switched on the old schedule. This is, of course, one small example of a much more difficult problem: there is no obvious way to pre-program daylight savings time changeover in a way that is sufficiently robust to withstand government intervention. (My favorite example is Arizona, where federal land changes, but state land remains on mountain standard time year around.) Martin Minow minow(a)apple.com ------------------------------

1 0