Theory revision 1.4
1# $NetBSD: Theory,v 1.4 1999/11/10 20:32:31 kleink Exp $ 2@(#)Theory 7.9 3 4 5----- Outline ----- 6 7 Time and date functions 8 Names of time zone regions 9 Time zone abbreviations 10 Calendrical issues 11 12 13----- Time and date functions ----- 14 15These time and date functions are upwards compatible with POSIX.1, 16an international standard for Unix-like systems. 17As of this writing, the current edition of POSIX.1 is: 18 19 Information technology --Portable Operating System Interface (POSIX (R)) 20 -- Part 1: System Application Program Interface (API) [C Language] 21 ISO/IEC 9945-1:1996 22 ANSI/IEEE Std 1003.1, 1996 Edition 23 1996-07-12 24 25POSIX.1 has the following properties and limitations. 26 27* In POSIX.1, time display in a process is controlled by the 28 environment variable TZ. Unfortunately, the POSIX.1 TZ string takes 29 a form that is hard to describe and is error-prone in practice. 30 Also, POSIX.1 TZ strings can't deal with other (for example, Israeli) 31 daylight saving time rules, or situations where more than two 32 time zone abbreviations are used in an area. 33 34 The POSIX.1 TZ string takes the following form: 35 36 stdoffset[dst[offset],date[/time],date[/time]] 37 38 where: 39 40 std and dst 41 are 3 or more characters specifying the standard 42 and daylight saving time (DST) zone names. 43 offset 44 is of the form `[-]hh:[mm[:ss]]' and specifies the 45 offset west of UTC. The default DST offset is one hour 46 ahead of standard time. 47 date[/time],date[/time] 48 specifies the beginning and end of DST. If this is absent, 49 the system supplies its own rules for DST, and these can 50 differ from year to year; typically US DST rules are used. 51 time 52 takes the form `hh:[mm[:ss]]' and defaults to 02:00. 53 date 54 takes one of the following forms: 55 Jn (1<=n<=365) 56 origin-1 day number not counting February 29 57 n (0<=n<=365) 58 origin-0 day number counting February 29 if present 59 Mm.n.d (0[Sunday]<=d<=6[Saturday], 1<=n<=5, 1<=m<=12) 60 for the dth day of week n of month m of the year, 61 where week 1 is the first week in which day d appears, 62 and `5' stands for the last week in which day d appears 63 (which may be either the 4th or 5th week). 64 65* In POSIX.1, when a TZ value like "EST5EDT" is parsed, 66 typically the current US DST rules are used, 67 but this means that the US DST rules are compiled into each program 68 that does time conversion. This means that when US time conversion 69 rules change (as in the United States in 1987), all programs that 70 do time conversion must be recompiled to ensure proper results. 71 72* In POSIX.1, there's no tamper-proof way for a process to learn the 73 system's best idea of local wall clock. (This is important for 74 applications that an administrator wants used only at certain times-- 75 without regard to whether the user has fiddled the "TZ" environment 76 variable. While an administrator can "do everything in UTC" to get 77 around the problem, doing so is inconvenient and precludes handling 78 daylight saving time shifts--as might be required to limit phone 79 calls to off-peak hours.) 80 81* POSIX.1 requires that systems ignore leap seconds. 82 83These are the extensions that have been made to the POSIX.1 functions: 84 85* The "TZ" environment variable is used in generating the name of a file 86 from which time zone information is read (or is interpreted a la 87 POSIX); "TZ" is no longer constrained to be a three-letter time zone 88 name followed by a number of hours and an optional three-letter 89 daylight time zone name. The daylight saving time rules to be used 90 for a particular time zone are encoded in the time zone file; 91 the format of the file allows U.S., Australian, and other rules to be 92 encoded, and allows for situations where more than two time zone 93 abbreviations are used. 94 95 It was recognized that allowing the "TZ" environment variable to 96 take on values such as "America/New_York" might cause "old" programs 97 (that expect "TZ" to have a certain form) to operate incorrectly; 98 consideration was given to using some other environment variable 99 (for example, "TIMEZONE") to hold the string used to generate the 100 time zone information file name. In the end, however, it was decided 101 to continue using "TZ": it is widely used for time zone purposes; 102 separately maintaining both "TZ" and "TIMEZONE" seemed a nuisance; 103 and systems where "new" forms of "TZ" might cause problems can simply 104 use TZ values such as "EST5EDT" which can be used both by 105 "new" programs (a la POSIX) and "old" programs (as zone names and 106 offsets). 107 108* To handle places where more than two time zone abbreviations are used, 109 the functions "localtime" and "gmtime" set tzname[tmp->tm_isdst] 110 (where "tmp" is the value the function returns) to the time zone 111 abbreviation to be used. This differs from POSIX.1, where the elements 112 of tzname are only changed as a result of calls to tzset. 113 114* Since the "TZ" environment variable can now be used to control time 115 conversion, the "daylight" and "timezone" variables are no longer 116 needed. (These variables are defined and set by "tzset"; however, their 117 values will not be used by "localtime.") 118 119* The "localtime" function has been set up to deliver correct results 120 for near-minimum or near-maximum time_t values. (A comment in the 121 source code tells how to get compatibly wrong results). 122 123* A function "tzsetwall" has been added to arrange for the system's 124 best approximation to local wall clock time to be delivered by 125 subsequent calls to "localtime." Source code for portable 126 applications that "must" run on local wall clock time should call 127 "tzsetwall();" if such code is moved to "old" systems that don't 128 provide tzsetwall, you won't be able to generate an executable program. 129 (These time zone functions also arrange for local wall clock time to be 130 used if tzset is called--directly or indirectly--and there's no "TZ" 131 environment variable; portable applications should not, however, rely 132 on this behavior since it's not the way SVR2 systems behave.) 133 134* These functions can account for leap seconds, thanks to Bradley White 135 (bww@k.cs.cmu.edu). 136 137Points of interest to folks with other systems: 138 139* This package is already part of many POSIX-compliant hosts, 140 including BSD, HP, Linux, Network Appliance, SCO, SGI, and Sun. 141 On such hosts, the primary use of this package 142 is to update obsolete time zone rule tables. 143 To do this, you may need to compile the time zone compiler 144 `zic' supplied with this package instead of using the system `zic', 145 since the format of zic's input changed slightly in late 1994, 146 and many vendors still do not support the new input format. 147 148* The Unix Version 7 "timezone" function is not present in this package; 149 it's impossible to reliably map timezone's arguments (a "minutes west 150 of GMT" value and a "daylight saving time in effect" flag) to a 151 time zone abbreviation, and we refuse to guess. 152 Programs that in the past used the timezone function may now examine 153 tzname[localtime(&clock)->tm_isdst] to learn the correct time 154 zone abbreviation to use. Alternatively, use 155 localtime(&clock)->tm_zone if this has been enabled. 156 157* The 4.2BSD gettimeofday function is not used in this package. 158 This formerly let users obtain the current UTC offset and DST flag, 159 but this functionality was removed in later versions of BSD. 160 161* In SVR2, time conversion fails for near-minimum or near-maximum 162 time_t values when doing conversions for places that don't use UTC. 163 This package takes care to do these conversions correctly. 164 165The functions that are conditionally compiled if STD_INSPIRED is defined 166should, at this point, be looked on primarily as food for thought. They are 167not in any sense "standard compatible"--some are not, in fact, specified in 168*any* standard. They do, however, represent responses of various authors to 169standardization proposals. 170 171Other time conversion proposals, in particular the one developed by folks at 172Hewlett Packard, offer a wider selection of functions that provide capabilities 173beyond those provided here. The absence of such functions from this package 174is not meant to discourage the development, standardization, or use of such 175functions. Rather, their absence reflects the decision to make this package 176contain valid extensions to POSIX.1, to ensure its broad 177acceptability. If more powerful time conversion functions can be standardized, 178so much the better. 179 180 181----- Names of time zone rule files ----- 182 183The names of this package's installed time zone rule files are chosen to 184help minimize possible future incompatibilities due to political events. 185Ordinarily, names of countries are not used, to avoid incompatibilities 186when countries change their name (e.g. Zaire->Congo) or 187when locations change countries (e.g. Hong Kong from UK colony to China). 188 189Names normally have the form AREA/LOCATION, where AREA is the name 190of a continent or ocean, and LOCATION is the name of a specific 191location within that region. North and South America share the same 192area, `America'. Typical names are `Africa/Cairo', `America/New_York', 193and `Pacific/Honolulu'. 194 195Here are the general rules used for choosing location names, 196in decreasing order of importance: 197 198 Use only valid Posix file names. Use only Ascii letters, digits, `.', 199 `-' and `_'. Do not exceed 14 characters or start with `-'. 200 E.g. prefer `Brunei' to `Bandar_Seri_Begawan'. 201 Include at least one location per time zone rule set per country. 202 One such location is enough. 203 If all the clocks in a country's region have agreed since 1970, 204 don't bother to include more than one location 205 even if subregions' clocks disagreed before 1970. 206 Otherwise these tables would become annoyingly large. 207 If a name is ambiguous, use a less ambiguous alternative; 208 e.g. many cities are named San Jose and Georgetown, so 209 prefer `Costa_Rica' to `San_Jose' and `Guyana' to `Georgetown'. 210 Keep locations compact. Use cities or small islands, not countries 211 or regions, so that any future time zone changes do not split 212 locations into different time zones. E.g. prefer `Paris' 213 to `France', since France has had multiple time zones. 214 Use traditional English spelling, e.g. prefer `Rome' to `Roma', and 215 prefer `Athens' to the true name (which uses Greek letters). 216 The Posix file name restrictions encourage this rule. 217 Use the most populous among locations in a country's time zone, 218 e.g. prefer `Shanghai' to `Beijing'. Among locations with 219 similar populations, pick the best-known location, 220 e.g. prefer `Rome' to `Milan'. 221 Use the singular form, e.g. prefer `Canary' to `Canaries'. 222 Omit common suffixes like `_Islands' and `_City', unless that 223 would lead to ambiguity. E.g. prefer `Cayman' to 224 `Cayman_Islands' and `Guatemala' to `Guatemala_City', 225 but prefer `Mexico_City' to `Mexico' because the country 226 of Mexico has several time zones. 227 Use `_' to represent a space. 228 Omit `.' from abbreviations in names, e.g. prefer `St_Helena' 229 to `St._Helena'. 230 231The file `zone.tab' lists the geographical locations used to name 232time zone rule files. 233 234Older versions of this package used a different naming scheme, 235and these older names are still supported. 236See the file `backwards' for most of these older names 237(e.g. `US/Eastern' instead of `America/New_York'). 238The other old-fashioned names still supported are 239`WET', `CET', `MET', `EET' (see the file `europe'), 240and `Factory' (see the file `factory'). 241 242 243----- Time zone abbreviations ----- 244 245When this package is installed, it generates time zone abbreviations 246like `EST' to be compatible with human tradition and POSIX.1. 247Here are the general rules used for choosing time zone abbreviations, 248in decreasing order of importance: 249 250 Use abbreviations that consist of 3 or more upper-case Ascii letters, 251 except use "___" for locations while uninhabited. 252 Posix.1 requires at least 3 characters, and the restriction to 253 upper-case Ascii letters follows most traditions. 254 Previous editions of this database also used characters like 255 ' ' and '?', but these characters have a special meaning to 256 the shell and cause commands like 257 set `date` 258 to have unexpected effects. In theory, the character set could 259 be !%./@A-Z^_a-z{}, but these tables use only upper-case 260 Ascii letters (and "___"). 261 Use abbreviations that are in common use among English-speakers, 262 e.g. `EST' for Eastern Standard Time in North America. 263 We assume that applications translate them to other languages 264 as part of the normal localization process; for example, 265 a French application might translate `EST' to `HNE'. 266 For zones whose times are taken from a city's longitude, use the 267 traditional xMT notation, e.g. `PMT' for Paris Mean Time. 268 The only name like this in current use is `GMT'. 269 If there is no common English abbreviation, abbreviate the English 270 translation of the usual phrase used by native speakers. 271 If this is not available or is a phrase mentioning the country 272 (e.g. ``Cape Verde Time''), then: 273 274 When a country has a single or principal time zone region, 275 append `T' to the country's ISO code, e.g. `CVT' for 276 Cape Verde Time. For summer time append `ST'; 277 for double summer time append `DST'; etc. 278 When a country has multiple time zones, take the first three 279 letters of an English place name identifying each zone 280 and then append `T', `ST', etc. as before; 281 e.g. `VLAST' for VLAdivostok Summer Time. 282 283Application writers should note that these abbreviations are ambiguous 284in practice: e.g. `EST' has a different meaning in Australia than 285it does in the United States. In new applications, it's often better 286to use numeric UTC offsets like `-0500' instead of time zone 287abbreviations like `EST'; this avoids the ambiguity. 288 289 290----- Calendrical issues ----- 291 292Calendrical issues are a bit out of scope for a time zone database, 293but they indicate the sort of problems that we would run into if we 294extended the time zone database further into the past. An excellent 295resource in this area is Nachum Dershowitz and Edward M. Reingold, 296<a href="http://emr.cs.uiuc.edu/home/reingold/calendar-book/index.shtml"> 297Calendrical Calculations 298</a>, Cambridge University Press (1997). Other information and 299sources are given below. They sometimes disagree. 300 301 302France 303 304Gregorian calendar adopted 1582-12-20. 305French Revolutionary calendar used 1793-11-24 through 1805-12-31, 306and (in Paris only) 1871-05-06 through 1871-05-23. 307 308 309Russia 310 311From Chris Carrier <72157.3334@CompuServe.COM> (1996-12-02): 312On 1929-10-01 the Soviet Union instituted an ``Eternal Calendar'' 313with 30-day months plus 5 holidays, with a 5-day week. 314On 1931-12-01 it changed to a 6-day week; in 1934 it reverted to the 315Gregorian calendar while retaining the 6-day week; on 1940-06-27 it 316reverted to the 7-day week. With the 6-day week the usual days 317off were the 6th, 12th, 18th, 24th and 30th of the month. 318(Source: Evitiar Zerubavel, _The Seven Day Circle_) 319 320 321Mark Brader reported a similar story in "The Book of Calendars", edited 322by Frank Parise (1982, Facts on File, ISBN 0-8719-6467-8), page 377. But: 323 324From: Petteri Sulonen (via Usenet) 325Date: 14 Jan 1999 00:00:00 GMT 326Message-ID: <Petteri.Sulonen-1401991626030001@lapin-kulta.in.helsinki.fi> 327 328If your source is correct, how come documents between 1929 -- 1940 were 329still dated using the conventional, Gregorian calendar? 330 331I can post a scan of a document dated December 1, 1934, signed by 332Yenukidze, the secretary, on behalf of Kalinin, the President of the 333Executive Committee of the Supreme Soviet, if you like. 334 335 336 337Sweden (and Finland) 338 339From: msb@sq.com (Mark Brader) 340<a href="news:1996Jul6.012937.29190@sq.com"> 341Subject: Re: Gregorian reform -- a part of locale? 342</a> 343Date: 1996-07-06 344 345In 1700, Denmark made the transition from Julian to Gregorian. Sweden 346decided to *start* a transition in 1700 as well, but rather than have one of 347those unsightly calendar gaps :-), they simply decreed that the next leap 348year after 1696 would be in 1744 -- putting the whole country on a calendar 349different from both Julian and Gregorian for a period of 40 years. 350 351However, in 1704 something went wrong and the plan was not carried through; 352they did, after all, have a leap year that year. And one in 1708. In 1712 353they gave it up and went back to Julian, putting 30 days in February that 354year!... 355 356Then in 1753, Sweden made the transition to Gregorian in the usual manner, 357getting there only 13 years behind the original schedule. 358 359(A previous posting of this story was challenged, and Swedish readers 360produced the following references to support it: "Tiderakning och historia" 361by Natanael Beckman (1924) and "Tid, en bok om tiderakning och 362kalendervasen" by Lars-Olof Lode'n (no date was given).) 363 364 365Grotefend's data 366 367From: "Michael Palmer" <mpalmer@netcom.com> [with one obvious typo fixed] 368Subject: Re: Gregorian Calendar (was Re: Another FHC related question 369Newsgroups: soc.genealogy.german 370Date: Tue, 9 Feb 1999 02:32:48 -800 371Message-ID: <199902091032.CAA09644@netcom10.netcom.com> 372 373The following is a(n incomplete) listing, arranged chronologically, of 374European states, with the date they converted from the Julian to the 375Gregorian calendar: 376 37704/15 Oct 1582 - Italy (with exceptions), Spain, Portugal, Poland (Roman 378 Catholics and Danzig only) 37909/20 Dec 1582 - France, Lorraine 380 38121 Dec 1582/ 382 01 Jan 1583 - Holland, Brabant, Flanders, Hennegau 38310/21 Feb 1583 - bishopric of Liege (L"uttich) 38413/24 Feb 1583 - bishopric of Augsburg 38504/15 Oct 1583 - electorate of Trier 38605/16 Oct 1583 - Bavaria, bishoprics of Freising, Eichstedt, Regensburg, 387 Salzburg, Brixen 38813/24 Oct 1583 - Austrian Oberelsass and Breisgau 38920/31 Oct 1583 - bishopric of Basel 39002/13 Nov 1583 - duchy of J"ulich-Berg 39102/13 Nov 1583 - electorate and city of K"oln 39204/15 Nov 1583 - bishopric of W"urzburg 39311/22 Nov 1583 - electorate of Mainz 39416/27 Nov 1583 - bishopric of Strassburg and the margraviate of Baden 39517/28 Nov 1583 - bishopric of M"unster and duchy of Cleve 39614/25 Dec 1583 - Steiermark 397 39806/17 Jan 1584 - Austria and Bohemia 39911/22 Jan 1584 - Luzern, Uri, Schwyz, Zug, Freiburg, Solothurn 40012/23 Jan 1584 - Silesia and the Lausitz 40122 Jan/ 402 02 Feb 1584 - Hungary (legally on 21 Oct 1587) 403 Jun 1584 - Unterwalden 40401/12 Jul 1584 - duchy of Westfalen 405 40616/27 Jun 1585 - bishopric of Paderborn 407 40814/25 Dec 1590 - Transylvania 409 41022 Aug/ 411 02 Sep 1612 - duchy of Prussia 412 41313/24 Dec 1614 - Pfalz-Neuburg 414 415 1617 - duchy of Kurland (reverted to the Julian calendar in 416 1796) 417 418 1624 - bishopric of Osnabr"uck 419 420 1630 - bishopric of Minden 421 42215/26 Mar 1631 - bishopric of Hildesheim 423 424 1655 - Kanton Wallis 425 42605/16 Feb 1682 - city of Strassburg 427 42818 Feb/ 429 01 Mar 1700 - Protestant Germany (including Swedish possessions in 430 Germany), Denmark, Norway 43130 Jun/ 432 12 Jul 1700 - Gelderland, Zutphen 43310 Nov/ 434 12 Dec 1700 - Utrecht, Overijssel 435 43631 Dec 1700/ 437 12 Jan 1701 - Friesland, Groningen, Z"urich, Bern, Basel, Geneva, 438 Turgau, and Schaffhausen 439 440 1724 - Glarus, Appenzell, and the city of St. Gallen 441 44201 Jan 1750 - Pisa and Florence 443 44402/14 Sep 1752 - Great Britain 445 44617 Feb/ 447 01 Mar 1753 - Sweden 448 4491760-1812 - Graub"unden 450 451The Russian empire (including Finland and the Baltic states) did not 452convert to the Gregorian calendar until the Soviet revolution of 1917. 453 454Source: H. Grotefend, _Taschenbuch der Zeitrechnung des deutschen 455Mittelalters und der Neuzeit_, herausgegeben von Dr. O. Grotefend 456(Hannover: Hahnsche Buchhandlung, 1941), pp. 26-28. 457