- tz - lists.iana.org

Re: Zeller
by Paul Eggert Oct. 4, 2000

Oct. 4, 2000

Oscar van Vlijmen <o.van.vlijmen(a)tip.nl> writes: > Allow me to give a couple of key sources for Zeller's algorithm to find the > day of the week: I see that those source's formulas have diverged a bit from Zeller's original algorithm, which you can find in: Chr. Zeller, Kalender-Formeln, Acta Mathematica, Vol. 9, Nov. 1886 I haven't read the original article (I don't read German), but I have seen a translation whose location escapes me now. I think if you translate his original formulas unchanged into C, and assume nonnegative years, they would look like this: /* Calculcate weekday number for Gregorian date YYYY-MM-DD. */ int jan_or_feb = MM <= 2; int y = YYYY - jan_or_feb; int J = y / 100; // century number int K = y % 100; // year number within the same int m = MM + 12 * jan_or_feb; // month number int q = DD; // day number in the month int h = (q + ((m + 1) * 26) / 10 + K + K/4 + J/4 - 2*J) % 7; // weekday number (1 is Sunday) However, this ignores overflows and doesn't handle negative years correctly. If you want it to work more generally, you need to alter the formulas, e.g. to something like the following. Please note that I've written this code by hand just now without testing it, so it's quite possibly buggy. #define DIV(a, b) (((a)/(b)) - ((a)%(b) < 0)) int jan_or_feb = MM <= 2; int K1 = YYYY % 100; int K2 = K1 - (K1 == 0 & jan_or_feb); int J = YYYY / 100 - (K2 < 0); int K = (K2 + 100) % 100; int m = MM + 12 * jan_or_feb; int q = DD; int h1 = (q + ((m + 1) * 26) / 10 + K + DIV(K,4) + DIV(J,4) - 2*J) % 7; int h = (h1 + 6) % 7;

2 1

Alternative place names?
by Oscar van Vlijmen Oct. 4, 2000

Oct. 4, 2000

> Jesper Nørgaard / Vietnam: Saigon vs. Ha Noi. > David Keegel / Theory file. > Gwillim Law / Theory file. > Vadim Vygonets / Saigon vs. Ha Noi and Jerusalem vs. Tel Aviv. > Paul Eggert / Jerusalem vs. Tel Aviv. > Gwillim Law / I think it's fine to put > alternative place names into the comments, but I hope we'll resist making > frequent changes to the zone names. One of the TZ prescriptions is: "Use the most populous among locations in a country's time zone, e.g. prefer `Shanghai' to `Beijing'. Among locations with similar populations, pick the best-known location, e.g. prefer `Rome' to `Milan'." The rationale of the 11 Theory prescriptions is of course: The user has to be able to find the wanted timezone immediately. I am NOT proposing to change the Theory rules, but I wonder: population count, popularity, media exposure all change, a capital hardly or not. A capital is or should be known to everybody. Haven't we all had some kind of geography lessons, years or decades ago? If in a TZ application the end user is able to see the commented alternative names, then in some cases alternatives should be mentioned in the data files. If not, mentioning alternatives still could be useful to the developers of Windows, Macintosh and web derivatives of the TZ application. > Paul Eggert, Vadim Vygonets, Antoine Leca / China: Chongqing A recurring problem is that people inside and outside of China don't like the official pinyin transliteration. Most professional map makers however use pinyin. Officially (Peking University course) the q- sound is an aspirated voiceless affricate, with j- the unaspirated voiceless counterpart. In practice j- sounds more like the (voiced!) g- in English genuine and q- more like the ch- in English chew. How Chongqing is pronounced in Chongqing I do not know. What I do know is that Chongqing is the official pinyin transliteration. Chung is a British transliteration (Wade, 1867), indeed: tch'ong is the French EFEO (1902) transliteration, but k'ing is the French transliteration for pinyin qing, whereas ch'ing would be the British transliteration. I'll vote for pinyin Chongqing. > Antoine Leca / Cambodia: Phnum Penh There is a (Unicode) "Proposal for encoding the Khmer script", by Maurice Bauhahn, a proposal to ISO-IEC JTC1/SC2/WG2. This is not a Romanization proposal, but in this document I found a reference to: American Library Association - Library of Congress romanization tables: transliteration schemes for non-Roman scripts approved by the Library of Congress and the American Library Association, Randall K. Barry, ed. 1997, isbn 0844409405. Who has access to these tables? See also the ISO/TC46/SC2 working groups for international transliteration standards (http://www.elot.gr/tc46sc2/) I suggest: Phnom_Penh as zone name, Phnum Penh in comment. > Gwillim Law / Do not change established names unless they become ambiguous or > shockingly incorrect. Yes, but established by whom? And what would you call shockingly incorrect? To some people it's shockingly incorrect to say Kirgizstan instead of Kyrgyzstan, but a certain influential publication or whoever 'established' an alternative transliteration. Where to draw the line? > Vadim Vygonets / Kirgizstan BTW: Kyrgyz is not a Russian (Slavonic) language, but a Turkic language, and -i- and -y- indeed do sound different. > Paul Eggert / Hong Kong, Macao, Harbin, Urumqi, Kashgar. I'll vote for these as zone names; one could put more or less common or official alternatives in comments. Alternatives to Kashgar: Kaxgar, Kashi. Oscar van Vlijmen 2000-09-27

7 8

RE: week starts at Sunday or Monday??
by Rich Wales Oct. 4, 2000

Oct. 4, 2000

"Jake" wrote: > "and on the seventh day, ye shall rest" for Gregorian > calendar = Sunday, for Hebraic, Seventh Day Adventists > (and others) = Saturday, etc. In the mainline Christian tradition, the day of worship was changed to the first day of the week in order to commemorate Jesus' resurrection. So, even though Sunday is the generally accepted Christian day of rest, it is still considered to be the first (not seventh) day of the week. Rich Wales richw(a)webcom.com http://www.webcom.com/richw/

2 1

Capitals
by Oscar van Vlijmen Oct. 4, 2000

Oct. 4, 2000

> Garrett Wollman: > Many time-zone regions do not have a `capital' of any sort. To give > just one example, the time zone `America/Los_Angeles' has three > `capitals': Sacramento, Salem, and Olympia. None of these are obvious > choices. `America/Chicago' has even more possibilities. Dunno, the book "The world almanac and book of facts", World Almanac Books, 1998, gives on page 542 a list of US state capitals with Sacramento as the only capital of California, Springfield (not Chicago) as capital of Illinois, etcetera. Good maps will also show the one and only capital of states. For US states with 2 or more timezones, one could use the most populous county seats or courthouses. For instance, Indiana has 92 counties to choose from (ibid, page 424). The same information can be found (elsewhere) for Canada. But I think it's a bit futile to rework the whole tz database only for the sake of being administratively correct. Here and there a couple of obvious corrections won't hurt. But one may have a different opinion. Oscar van Vlijmen 2000-09-28

3 2

Time zone changes due to Brazilian Summer Time
by Paulo Alexandre Pinto Pires Oct. 4, 2000

Oct. 4, 2000

Dear sirs, Below are the patches (with full comments) for adjusting zone information for South America, due to Brazilian changing of beginning and ending dates for adoption of summer time. Patches can be applied to version 7.36 (as available from ftp://elsie.nci.nih.gov at 2000/10/04 00:00 GMT), and they are in unified diff format. I am also sending the full altered file, as a GNU-zip compressed attachment. Regards, Paulo A. P. Pires -- p(a)ppires.org 8<---------------------------------------------- --- /usr/src/share/zoneinfo/southamerica.orig Tue Oct 3 22:03:37 2000 +++ /usr/src/share/zoneinfo/southamerica Tue Oct 3 23:06:56 2000 @@ -367,8 +367,23 @@ # adds SE, AL, PE, PR, RN, CE, PI, MA and RR. # These give only one year's rules. After that, the rules are guesses # and are quite possibly wrong, but are more likely than no DST at all. -Rule Brazil 1999 max - Oct Sun>=1 0:00 1:00 S -Rule Brazil 2000 max - Feb lastSun 0:00 0 - +Rule Brazil 1999 only - Oct Sun>=1 0:00 1:00 S +Rule Brazil 2000 only - Feb lastSun 0:00 0 - +# +# From Paulo A. P. Pires (2000-10-03) +# Brazil changed rules once again. Late decree with poor announcement +# (even some telephone companies had their services one hour ahead of legal +# for a short time). +# According to sources in the Research Group on Energy, from the Institute of +# Economics of the Federal University of Rio de Janeiro, initial and final +# summer time dates were not originally supposed to change from previous year, +# but people in northern states don't like summer time (for the obvious reason +# that is not as effective, but rather uncomfortable, near the Equator line). +# Decree <a href="http://pcdsh01.on.br/DEC3592.htm">3592</A> (2000/09/06) lists +# the following states to obey summer time: RS, SC, PR, SP, RJ, ES, MG, GO, +# MT, MS, TO, BA, SE, AL, PE, RN, CE, PI, MA, RR and DF. +Rule Brazil 2000 max - Oct Sun>=8 0:00 1:00 S +Rule Brazil 2001 max - Feb Sun>=18 0:00 0 - # Zone NAME GMTOFF RULES FORMAT [UNTIL] ------------------------------------------->8 -- Pappires ... Qui habet aurem audiendi audiat quid Spiritus dicat ecclesiis.

1 0

Patch for Brazil Summer Time 2000.
by Marcus Felipe Pereira Oct. 3, 2000

Oct. 3, 2000

Hi Arthur, Brazil's government changed the rules for the Daylight Summer Time - AGAIN! They changed the start date of it from "Fist Sunday of October" to "Second Sunday or October". So in 2000 the BRST will start on 8th of October an not in 1st of October. I'm sending a patch for the original "southamerica" timezone file from ftp://elsie.nci.nih.gov/pub/tzdata2000f.tar.gz Marcus Felipe Pereira marcus(a)task.com.br

1 0

Re: Alternative place names?
by Paul Eggert Oct. 3, 2000

Oct. 3, 2000

Oscar van Vlijmen <o.van.vlijmen(a)tip.nl> writes: > I am NOT proposing to change the Theory rules, but I wonder: population > count, popularity, media exposure all change, a capital hardly or not. Actually, capitals change suprisingly often. E.g. Kazakhstan changed its capital in 1997, and then changed the capital's name in 1998. (Good thing they didn't use new time zone rules too, or we'd have had to track that mess.) Even China, the most populous country in the world, changed its capital twice in the last century (in 1928 and again in 1949) and there was a period in which the location of its capital was in serious dispute. Capitals are political creatures, and we are trying to avoid politics here as much as possible, so I don't want to give the "capitalness" of a city much weight. > Chung is a British transliteration (Wade, 1867), indeed: tch'ong is the > French EFEO (1902) transliteration, but k'ing is the French transliteration > for pinyin qing, whereas ch'ing would be the British transliteration. The name "Chungking" is not a systematic transliteration; it is merely traditional. It's probably the weakest of the existing Chinese names, i.e. the one where there is the strongest reason to change, as I get the feeling that the Pinyin transliteration "Chongqing" is gaining wide popularity in ordinary English usage. Google reports 38k English-language hits for "Chongqing" and 21k for "Chungking", so if we were choosing now "Chongqing" might well be the winner. But I'm not sure that this is enough to change an established name. Oscar van Vlijmen <o.van.vlijmen(a)tip.nl> writes: Paul Hill writes: > Someone said there is the rule: every country gets at least one entry. > Is this implemented in practice? Every country including all small > countries that share zones (like the Caribbean)? Yes. In hindsight this rule was probably unwise, as countries are also political creatures. Not sure I want to change it now, though.

5 4

Re: week starts at Sunday or Monday??
by Janis.Papanagnou＠varetis.de Oct. 3, 2000

Oct. 3, 2000

>so my question is.... >now that I decided that days are niumbere 1 to 7, >what is day 1 ??? >is it Sunday, or Monday? >what is the Right Thing to do? Your choice is ok. One observation to add is about numbering/enumerating/counting in general: 1st, 2nd, 3rd, ... well - I've never heard about "the 0-st". Enumerating starting with zero seems to have come in vogue with C, due to the simpler address calculation (saves some micro seconds?). Whether week starts with Monday or Sunday depends on the culture; this is independent of any number assigned to it. Regards, Janis

5 5

Paraguay change
by Steffen Thorsen Oct. 2, 2000

Oct. 2, 2000

I have three independent reports that Paraguay (America/Asuncion) changed to DST this Sunday (1 Oct), so the "Para"-rules must be updated. Also, Brazil needs to be updated, like already reported by several people on this list. Best regards, Steffen -- Steffen Thorsen <steffen(a)thorsen.priv.no> <steffent(a)pvv.ntnu.no> http://www.thorsen.priv.no

1 0

Re: week starts at Sunday or Monday??
by Dik.Winter＠cwi.nl Oct. 2, 2000

Oct. 2, 2000

> > 1st, 2nd, 3rd, ... well - I've never heard about "the 0-st". > > Enumerating starting with zero seems to have come in vogue with C, > There's been a Zero Church Street in Cambridge, Mass., for a good deal > longer than there's been a C programming language. Blame the > mathematicians. The Indian calendars use a large number of counting years. Some of those start with 0, others with 1. This even predates Cambridge, Mass.. dik

1 0