- tz - lists.iana.org

Turkey to keep permanent DST
by Even Scharning Sept. 8, 2016

Sept. 8, 2016

Turkey will stay on DST, canceling October's transition back to UTC +2 standard time, according to Hürriyet, Turkey's "leading news source". http://www.hurriyet.com.tr/kis-saati-uygulamasi-yururlukten-kaldirildi-4021… Time in Turkey: http://time.is/Turkey Even Scharning Time.is - exact time for any time zone http://time.is/

3 4

Google search results displaying wrong time for Egypt
by Andrew Paprocki Sept. 7, 2016

Sept. 7, 2016

If you search for "current time in Egypt" (at least for me in NYC), it will display a UTC+3 time as if DST was applied (tzdata2016f not being used). In case there are any Googlers on the list want to raise awareness internally... -Andrew

7 11

[PROPOSED PATCH 1/5] zic now omits some mkdir and stat system calls
by Paul Eggert Sept. 6, 2016

Sept. 6, 2016

* NEWS: Document this. * zic.c (mkdirs): Do not mkdir a root directory. Avoid unnecessary stat call. --- NEWS | 2 ++ zic.c | 24 +++++++++++++----------- 2 files changed, 15 insertions(+), 11 deletions(-) diff --git a/NEWS b/NEWS index f5a0b13..192deb7 100644 --- a/NEWS +++ b/NEWS @@ -55,6 +55,8 @@ Unreleased, experimental changes stamps on the reference platform. (Thanks to Alexander Belopolsky for reporting the bug and suggesting a way forward.) + zic now avoids some unnecessary mkdir and stat system calls. + zdump has a new -i option to generate transitions in a more-compact but still human-readable format. This option is experimental, and the output format may change in future versions. diff --git a/zic.c b/zic.c index b8bd0a6..5ff78b4 100644 --- a/zic.c +++ b/zic.c @@ -3027,25 +3027,27 @@ mkdirs(char *argname) if (argname == NULL || *argname == '\0') return true; cp = name = ecpyalloc(argname); - while ((cp = strchr(cp + 1, '/')) != 0) { - *cp = '\0'; + + /* Do not mkdir a root directory, as it must exist. */ #ifdef HAVE_DOS_FILE_NAMES - /* - ** DOS drive specifier? - */ - if (is_alpha(name[0]) && name[1] == ':' && name[2] == '\0') { - *cp = '/'; - continue; - } + if (is_alpha(name[0]) && name[1] == ':') + cp += 2; #endif + while (*cp == '/') + cp++; + + for (; (cp = strchr(cp, '/')) != 0; cp++) { + *cp = '\0'; /* ** Try to create it. It's OK if creation fails because ** the directory already exists, perhaps because some - ** other process just created it. + ** other process just created it. For simplicity do + ** not check first whether it already exists, as that + ** is checked anyway if the mkdir fails. */ if (mkdir(name, MKDIR_UMASK) != 0) { int err = errno; - if (itsdir(name) <= 0) { + if (err != EEXIST && itsdir(name) <= 0) { char const *e = strerror(err); warning(_("%s: Can't create directory" " %s: %s"), -- 2.7.4

1 4

[PROPOSED PATCH 1/3] Call the new tarball experimental
by Paul Eggert Sept. 6, 2016

Sept. 6, 2016

* Makefile (tzdata$(VERSION).tar.gz): Omit $(TZS). * NEWS: Call the new format experimental, and say it has to2050.tzs. Do not imply a time limit for the old format. * README, tz-link.htm: Bring back traditional two-tarball recipe. * tz-how-to.html: Defer to tz-link.htm for tarball link. * tz-link.htm: Update some URLs. --- Makefile | 2 +- NEWS | 33 ++++++++++++++++----------------- README | 11 +++++++++++ tz-how-to.html | 3 +-- tz-link.htm | 52 ++++++++++++++++++++++++++++------------------------ 5 files changed, 57 insertions(+), 44 deletions(-) diff --git a/Makefile b/Makefile index 90363e9..31a2ec5 100644 --- a/Makefile +++ b/Makefile @@ -683,7 +683,7 @@ tzcode$(VERSION).tar.gz: set-timestamps.out tzdata$(VERSION).tar.gz: set-timestamps.out LC_ALL=C && export LC_ALL && \ - tar $(TARFLAGS) -cf - $(COMMON) $(DATA) $(MISC) $(TZS) | \ + tar $(TARFLAGS) -cf - $(COMMON) $(DATA) $(MISC) | \ gzip $(GZIPFLAGS) > $@ tzdb-$(VERSION).tar.lz: set-timestamps.out diff --git a/NEWS b/NEWS index e61a992..6cbdbd7 100644 --- a/NEWS +++ b/NEWS @@ -63,20 +63,18 @@ Unreleased, experimental changes Changes to build procedure - The tz distribution has a new file to2050.tzs that contains what - should be the output of 'zdump -i -c 2050' on primary zones. - 'make check' now checks that zdump generates this output. - - A new distribution format is available in the tarball - tzdb-VERSION.tar.lz and the signature tzdb-VERSION.tar.lz.asc. - The new tarball has the contents of tzcodeVERSION.tar.gz and - tzdataVERSION.tar.gz, in a single top-level directory 'tzdb' with - all other files under this directory, as is typical for software - distributions. The new format is intended to replace the old - tarball pair. The old format will continue to be distributed for - a while. The new format uses lzip compression, which is smaller - than gzip; see <http://www.nongnu.org/lzip/lzip.html>. - (Thanks to Antonio Diaz Diaz and Oscar van Vlijmen for comments.) + An experimental distribution format is available in the tarball + tzdb-VERSION.tar.lz with signature file tzdb-VERSION.tar.lz.asc. + It unpacks to a top-level directory containing the code and data + of the traditional two-tarball format, along with extra data that + may be useful. The traditional format will continue to be + distributed. (Thanks to Antonio Diaz Diaz, Oscar van Vlijmen, and + many others for comments about the experimental format.) + + The experimental distribution contains a file to2050.tzs that + contains what should be the output of 'zdump -i -c 2050' on + primary zones. If this file is available, 'make check' now checks + that zdump generates this output. 'make check_web' now works on Fedora-like distributions. @@ -3613,14 +3611,15 @@ This file contains copies of the part of each release announcement that talks about the changes in that release. The text has been adapted and reformatted for the purposes of this file. -Typically a release R consists of a pair of tarball files, +Traditionally a release R consists of a pair of tarball files, tzcodeR.tar.gz and tzdataR.tar.gz. However, some releases (e.g., code2010a, data2012c) consist of just one or the other tarball, and a few (e.g., code2012c-data2012d) have tarballs with mixed version -numbers. +numbers. Recent releases also come in an experimental format +consisting of a single tarball tzdb-R.tar.lz with extra data. Release time stamps are taken from the release's commit (for newer, -git releases), from the newest file in the tarball (for older +Git-based releases), from the newest file in the tarball (for older releases, where this info is available) or from the email announcing the release (if all else fails; these are marked with a time zone of -0000 and an "is missing!" comment). diff --git a/README b/README index ba0b790..8a09aa2 100644 --- a/README +++ b/README @@ -13,6 +13,17 @@ and daylight-saving rules. Here is a recipe for acquiring, building, installing, and testing the tz distribution on a GNU/Linux or similar host. +To acquire the distribution, run the following shell commands: + + mkdir tz + cd tz + wget --retr-symlinks 'ftp://ftp.iana.org/tz/tz*-latest.tar.gz' + gzip -dc tzcode-latest.tar.gz | tar -xf - + gzip -dc tzdata-latest.tar.gz | tar -xf - + +Alternatively, the following shell commands acquire the same +distribution, with extra data useful for regression testing: + wget --retr-symlinks 'ftp://ftp.iana.org/tz/tzdb-latest.tar.lz' lzip -dc tzdb-latest.tar.lz | tar -xf - diff --git a/tz-how-to.html b/tz-how-to.html index 7d9d521..a94759e 100644 --- a/tz-how-to.html +++ b/tz-how-to.html @@ -12,8 +12,7 @@ Database</a> Source Files</h2> times of day from the <a href="tz-link.htm">tz database</a> source files. It might be helpful, but not absolutely necessary, for the reader to have already downloaded the -<a href="http://www.iana.org/time-zones/repository/tzdb-latest.tar.lz">latest -release of the database</a> and become familiar with the basic layout +latest release of the database and become familiar with the basic layout of the data files. The format is explained in the “man page” for the zic compiler, <code>zic.8.txt</code>, in the <code>code</code> subdirectory.</p> diff --git a/tz-link.htm b/tz-link.htm index 8b05470..52bb492 100644 --- a/tz-link.htm +++ b/tz-link.htm @@ -10,7 +10,7 @@ content="Sources for time zone and daylight saving time data"> <meta name="DC.Creator" content="Eggert, Paul"> <meta name="DC.Contributor" content="Olson, Arthur David"> -<meta name="DC.Date" content="2016-08-30"> +<meta name="DC.Date" content="2016-09-05"> <meta name="DC.Description" content="Sources of information about time zones and daylight saving time"> <meta name="DC.Identifier" @@ -42,7 +42,7 @@ This database (often called <code>zoneinfo</code> or <code><abbr>tz</abbr></code>) is used by several implementations, including -<a href="http://www.gnu.org/software/libc/">the +<a href="https://www.gnu.org/software/libc/">the <abbr title="GNU's Not Unix">GNU</abbr> C Library</a> (used in <a href="https://en.wikipedia.org/wiki/Linux"><abbr>GNU</abbr>/Linux</a>), @@ -108,33 +108,35 @@ for Eastern Standard Time in the <abbr>US</abbr>.</p> <p> The following <a href="https://en.wikipedia.org/wiki/Unix_shell">shell</a> commands download -the latest release to a <abbr>GNU</abbr>/Linux or similar host; -see the downloaded -<code>README</code> file for what to do next.</p> -<pre style="margin-left: 2em"><code><a href="http://www.gnu.org/software/wget/">wget</a> --retr-symlinks 'ftp://ftp.iana.org/tz/tzdb-latest.tar.lz' -<a href="http://www.nongnu.org/lzip/">lzip</a> -dc tzdb-latest.tar.lz | <a href="http://www.gnu.org/software/tar/">tar</a> -xf - +the latest release's two +<a href="https://en.wikipedia.org/wiki/Tar_(computing)">tarballs</a> +to a <abbr>GNU</abbr>/Linux or similar host.</p> +<pre style="margin-left: 2em"><code>mkdir tzdb +cd tzdb +<a href="https://www.gnu.org/software/wget/">wget</a> --retr-symlinks 'ftp://ftp.iana.org/tz/tz*-latest.tar.gz' +<a href="https://www.gnu.org/software/gzip/">gzip</a> -dc tzcode-latest.tar.gz | <a href="https://www.gnu.org/software/tar/">tar</a> -xf - +gzip -dc tzdata-latest.tar.gz | tar -xf - </code></pre> -<p>This uses a convenience link to the -<a href="ftp://ftp.iana.org/tz/tzdb-latest.tar.lz">latest version</a> -archived in the +<p>Alternatively, the following shell commands download the same +release in a single-tarball format containing extra data +useful for regression testing:</p> +<pre style="margin-left: 2em"><code>wget --retr-symlinks 'ftp://ftp.iana.org/tz/tzdb-latest.tar.lz' +<a href="http://www.nongnu.org/lzip/">lzip</a> -dc tzdb-latest.tar.lz | tar -xf - +</code></pre> +<p>These commands use convenience links to the latest of the <a href="ftp://ftp.iana.org/tz/releases/"><abbr title="File Transfer Protocol">FTP</abbr> releases</a> of the <code><abbr>tz</abbr></code> database. -Archived versions are in files named <code>tzdb-<var>V</var>.tar.lz</code>, +Older releases are in files named +<code>tzcode<var>V</var>.tar.gz</code>, +<code>tzdata<var>V</var>.tar.gz</code>, and +<code>tzdb-<var>V</var>.tar.lz</code>, where <code><var>V</var></code> is the version. -For compatibility with releases before 2016g, a -<a href="https://www.gnu.org/software/gzip/">gzip</a>-compressed two-tarball -distribution is also available, where the code is in the -file <code>tzcode<var>C</var>.tar.gz</code>, -where <code><var>C</var></code> is the code's version; -similarly, the data entries are in <code>tzdata<var>D</var>.tar.gz</code>, -where <code><var>D</var></code> is the data's version. Since 1996, each version has been a four-digit year followed by lower-case letter (<samp>a</samp> through <samp>z</samp>, then <samp>za</samp> through <samp>zz</samp>, then <samp>zza</samp> -through <samp>zzz</samp>, and so on).</p> -<p> -The code and data files can also be obtained from the +through <samp>zzz</samp>, and so on). +The releases can also be obtained from the <a href="http://www.iana.org/time-zones">Time Zone Database website</a> of the <a href="http://www.iana.org">Internet Assigned Numbers Authority (IANA)</a>. @@ -144,6 +146,8 @@ in <a href="http://git-scm.com">Git</a> form from <a href="https://github.com">GitHub</a>; be careful, as this repository is less well tested and probably contains more errors. <p> +After extracting a distribution's files, see its +<code>README</code> file for what to do next. The code lets you compile the <code><abbr>tz</abbr></code> source files into machine-readable binary files, one for each location. It also lets you read a <code><abbr>tz</abbr></code> binary file and interpret time stamps for that @@ -295,7 +299,7 @@ program that compiles <code><abbr>tz</abbr></code> source into iCalendar-compatible VTIMEZONE files. Vzic is freely available under the <a -href="http://www.gnu.org/copyleft/gpl.html"><abbr>GNU</abbr> +href="https://www.gnu.org/copyleft/gpl.html"><abbr>GNU</abbr> General Public License (<abbr title="General Public License">GPL</abbr>)</a>.</li> <li><a href="https://sourceforge.net/projects/tzical/">tziCal – tz @@ -357,7 +361,7 @@ Advanced date, time and interval library for Java</a> contains a class <code>net.time4j.tool.TimezoneRepositoryCompiler</code> that compiles <code><abbr>tz</abbr></code> source into a binary format. Time4J is available under the <a -href="http://www.gnu.org/copyleft/lesser.html"><abbr>GNU</abbr> Lesser +href="https://www.gnu.org/copyleft/lesser.html"><abbr>GNU</abbr> Lesser General Public License (<abbr title="Lesser General Public License">LGPL</abbr>)</a>.</li> <li><abbr>ICU</abbr> (mentioned above) contains compilers and @@ -413,7 +417,7 @@ license.</li> <h2>Other <code><abbr>tz</abbr></code> binary file readers</h2> <ul> <li>The <a -href="http://www.gnu.org/software/libc/"><abbr>GNU</abbr> C +href="https://www.gnu.org/software/libc/"><abbr>GNU</abbr> C Library</a> has an independent, thread-safe implementation of a <code><abbr>tz</abbr></code> binary file reader. -- 2.7.4

1 2

tzdata for 2016 leap seconds
by Jimmy Wang (jcwang) Sept. 5, 2016

Sept. 5, 2016

Hi Paul et al, We heard that the tzdata for 2016 leap seconds would be published around Oct. Is there anyway that this can be published sooner? Many of our upstream vendors won't release their tzdata update until the official tzdata is published. Thanks, Jimmy

7 11

[PROPOSED PATCH] port 'environ' decl to Solaris 10
by Paul Eggert Sept. 5, 2016

Sept. 5, 2016

POSIX says apps must declare 'environ' themselves. Although this is not necessary on GNU/Linux with _GNU_SOURCE defined, it is needed with stricter environments like Solaris 10. * date.c, zdump.c (environ): Declare even if HAVE_POSIX_DECLS. --- date.c | 3 ++- zdump.c | 3 ++- 2 files changed, 4 insertions(+), 2 deletions(-) diff --git a/date.c b/date.c index 4c11f61..512d1ef 100644 --- a/date.c +++ b/date.c @@ -42,8 +42,9 @@ #define SECSPERMIN 60 #endif /* !defined SECSPERMIN */ -#if !HAVE_POSIX_DECLS extern char ** environ; + +#if !HAVE_POSIX_DECLS extern char * optarg; extern int optind; extern char * tzname[]; diff --git a/zdump.c b/zdump.c index 694b04f..f68ba3d 100644 --- a/zdump.c +++ b/zdump.c @@ -258,8 +258,9 @@ enum { SECSPER400YEARS_FITS = SECSPERLYEAR <= INTMAX_MAX / 400 }; # define timezone_t char ** #endif -#if !HAVE_POSIX_DECLS extern char ** environ; + +#if !HAVE_POSIX_DECLS extern int getopt(int argc, char * const argv[], const char * options); extern char * optarg; -- 2.7.4

2 2

[PROPOSED PATCH] * zic.c (dolink): Fix typo (missing newline) in diagnostic.
by Paul Eggert Sept. 4, 2016

Sept. 4, 2016

--- zic.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/zic.c b/zic.c index 9c696c6..b8bd0a6 100644 --- a/zic.c +++ b/zic.c @@ -772,7 +772,7 @@ dolink(char const *fromfield, char const *tofield) fromisdir = itsdir(fromname); if (fromisdir) { char const *e = strerror(fromisdir < 0 ? errno : EPERM); - fprintf(stderr, _("%s: link from %s failed: %s"), + fprintf(stderr, _("%s: link from %s failed: %s\n"), progname, fromname, e); exit(EXIT_FAILURE); } -- 2.7.4

1 0

[PROPOSED PATCH] zic fix storage-allocation bug with realloc
by Paul Eggert Sept. 4, 2016

Sept. 4, 2016

This bug was introduced in last month's Casablanca-in-2037 fix. Problem found by valgrind. * zic.c (outzone): Don't access freed storage. --- zic.c | 11 ++++++----- 1 file changed, 6 insertions(+), 5 deletions(-) diff --git a/zic.c b/zic.c index 1c91d5a..9c696c6 100644 --- a/zic.c +++ b/zic.c @@ -2290,7 +2290,7 @@ outzone(const struct zone *zpfirst, int zonecount) register int compat; register bool do_extend; register char version; - struct attype *lastatmax = NULL; + int lastatmax = -1; max_abbr_len = 2 + max_format_len + max_abbrvar_len; max_envvar_len = 2 * max_abbr_len + 5 * 9; @@ -2524,8 +2524,9 @@ outzone(const struct zone *zpfirst, int zonecount) type = addtype(offset, ab, rp->r_stdoff != 0, rp->r_todisstd, rp->r_todisgmt); if (rp->r_hiyear == ZIC_MAX - && ! (lastatmax && ktime < lastatmax->at)) - lastatmax = &attypes[timecnt]; + && ! (0 <= lastatmax + && ktime < attypes[lastatmax].at)) + lastatmax = timecnt; addtt(ktime, type); } } @@ -2557,8 +2558,8 @@ error(_("can't determine time zone abbreviation to use just after until time")); starttime = tadd(starttime, -gmtoff); } } - if (lastatmax) - lastatmax->dontmerge = true; + if (0 <= lastatmax) + attypes[lastatmax].dontmerge = true; if (do_extend) { /* ** If we're extending the explicitly listed observations -- 2.7.4

1 0

Bloomberg on possible Massachusetts time zone shift
by Arthur David Olson Aug. 30, 2016

Aug. 30, 2016

"Massachusetts Could Swap Time Zones for Later Winter Sunsets" (To Atlantic Time.) http://www.bloomberg.com/news/articles/2016-08-29/winter-s-hated-4-p-m-suns… --ado

2 1

[PROPOSED PATCH 1/4] Prefer +-NN notation for UT-offset time zones in comments.
by Paul Eggert Aug. 26, 2016

Aug. 26, 2016

This does not change any code or data; just comments. --- NEWS | 38 +++++++++++++++++++------------------- Theory | 2 +- africa | 2 +- asia | 51 +++++++++++++++++++++++++-------------------------- australasia | 26 +++++++++++++------------- backzone | 14 +++++++------- europe | 8 +------- localtime.c | 2 +- northamerica | 2 +- private.h | 1 + southamerica | 4 ++-- 11 files changed, 72 insertions(+), 78 deletions(-) diff --git a/NEWS b/NEWS index 4c18e55..b0da284 100644 --- a/NEWS +++ b/NEWS @@ -465,7 +465,7 @@ Release 2015d - 2015-04-24 08:09:46 -0700 Changes affecting past time stamps - America/Whitehorse switched from UTC-9 to UTC-8 on 1967-05-28, not + America/Whitehorse switched from -09 to -08 on 1967-05-28, not 1966-07-01. Also, Yukon's time zone history is documented better. (Thanks to Brian Inglis and Dennis Ferguson.) @@ -499,12 +499,12 @@ Release 2015c - 2015-04-11 08:55:55 -0700 The 1918 transition was September 10, not September 1. - The UTC-4 time observed from 1932 to 1942 is now considered to be + The -04 time observed from 1932 to 1942 is now considered to be standard time, not year-round DST. - Santiago observed DST (UTC-3) from 1946-07-15 through 1946-08-31, + Santiago observed DST (-03) from 1946-07-15 through 1946-08-31, then reverted to standard time, then switched its time zone to - UTC-5 on 1947-04-01. + -05 on 1947-04-01. Assume transitions before 1968 were at 00:00, since we have no data saying otherwise. @@ -633,7 +633,7 @@ Release 2014j - 2014-11-10 17:37:11 -0800 Changes affecting current and future time stamps - Turks & Caicos' switch from US eastern time to UTC-4 year-round + Turks & Caicos' switch from US eastern time to -04 year-round did not occur on 2014-11-02 at 02:00. It's currently scheduled for 2015-11-01 at 02:00. (Thanks to Chris Walton.) @@ -669,7 +669,7 @@ Release 2014i - 2014-10-21 22:04:57 -0700 years will use a similar pattern. A new Zone Pacific/Bougainville, for the part of Papua New Guinea - that plans to switch from UTC+10 to UTC+11 on 2014-12-28 at 02:00. + that plans to switch from +10 to +11 on 2014-12-28 at 02:00. (Thanks to Kiley Walbom for the heads-up.) Changes affecting time zone abbreviations @@ -679,8 +679,8 @@ Release 2014i - 2014-10-21 22:04:57 -0700 to its more-traditional value MSK on 2014-10-26 at 01:00. (Thanks to Alexander Bokovoy for the heads-up about Belarus.) - The new abbreviation IDT stands for the pre-1976 use of UT+8 in - Indochina, to distinguish it better from ICT (UT+7). + The new abbreviation IDT stands for the pre-1976 use of +08 in + Indochina, to distinguish it better from ICT (+07). Changes affecting past time stamps @@ -791,7 +791,7 @@ Release 2014g - 2014-08-28 12:31:23 -0700 Changes affecting future time stamps - Turks & Caicos is switching from US eastern time to UTC-4 year-round, + Turks & Caicos is switching from US eastern time to -04 year-round, modeled as a switch from EST/EDT to AST on 2014-11-02 at 02:00. [As noted in 2014j, this switch was later delayed.] @@ -975,7 +975,7 @@ Release 2014f - 2014-08-05 17:42:36 -0700 This change does not affect UTC offsets, only time zone abbreviations. (Thanks to Rich Tibbett and many others.) - Asia/Novokuznetsk shifts from NOVT to KRAT (remaining on UTC+7) + Asia/Novokuznetsk shifts from NOVT to KRAT (remaining on +07) effective 2014-10-26 at 02:00 local time. The time zone abbreviation for Xinjiang Time (observed in Ürümqi) @@ -1004,8 +1004,8 @@ Release 2014f - 2014-08-05 17:42:36 -0700 zones Asia/Harbin, Asia/Chongqing, and Asia/Kashgar have been removed; backwards-compatibility links still work, albeit with different behaviors for time stamps before May 1980. Asia/Urumqi's - 1980 transition to UTC+8 has been removed, so that it is now at - UTC+6 and not UTC+8. (Thanks to Luther Ma and to Alois Treindl; + 1980 transition to +08 has been removed, so that it is now at + +06 and not +08. (Thanks to Luther Ma and to Alois Treindl; Treindl sent helpful translations of two papers by Guo Qingsheng.) Some zones have been turned into links, when they differed from existing @@ -1035,7 +1035,7 @@ Release 2014f - 2014-08-05 17:42:36 -0700 Europe/Helsinki's 1942 fall-back transition was 10-04 at 01:00, not 10-03 at 00:00. (Thanks to Konstantin Hyppönen.) - Pacific/Pago_Pago has been changed from UTC-11:30 to UTC-11 for the period + Pacific/Pago_Pago has been changed from -11:30 to -11 for the period from 1911 to 1950. Pacific/Chatham has been changed to New Zealand standard time plus @@ -1363,15 +1363,15 @@ Release 2013h - 2013-10-25 15:32:32 -0700 Changes affecting current and future time stamps: - Libya has switched its time zone back to UTC+2 without DST, - instead of UTC+1 with DST. (Thanks to Even Scharning.) + Libya has switched its time zone back to +02 without DST, + instead of +01 with DST. (Thanks to Even Scharning.) Western Sahara (Africa/El_Aaiun) uses Morocco's DST rules. (Thanks to Gwillim Law.) Changes affecting future time stamps: - Acre and (we guess) western Amazonas will switch from UTC-4 to UTC-5 + Acre and (we guess) western Amazonas will switch from -04 to -05 on 2013-11-10. This affects America/Rio_Branco and America/Eirunepe. (Thanks to Steffen Thorsen.) @@ -1458,7 +1458,7 @@ Release 2013f - 2013-09-24 23:37:36 -0700 Tocantins will very likely not observe DST starting this spring. (Thanks to Steffen Thorsen.) - Jordan will likely stay at UTC+3 indefinitely, and will not fall + Jordan will likely stay at +03 indefinitely, and will not fall back this fall. Palestine will fall back at 00:00, not 01:00. (Thanks to Steffen Thorsen.) @@ -1515,7 +1515,7 @@ Release 2013e - 2013-09-19 23:50:04 -0700 Benny Lin.) This affects Asia/Dili, Asia/Jakarta, Asia/Jayapura, Asia/Makassar, and Asia/Pontianak. - Use ART (UTC-3, standard time), rather than WARST (also UTC-3, but + Use ART (-03, standard time), rather than WARST (also UTC-3, but daylight saving time) for San Luis, Argentina since 2009. Changes affecting Godthåb time stamps after 2037 if version mismatch @@ -1972,7 +1972,7 @@ Release 2012f - 2012-09-12 23:17:03 -0700 Release 2012e - 2012-08-02 20:44:55 -0700 - * australasia (Pacific/Fakaofo): Tokelau is UTC+13, not UTC+14. + * australasia (Pacific/Fakaofo): Tokelau is +13, not +14. (Thanks to Steffen Thorsen.) * Use a single version number for both code and data. diff --git a/Theory b/Theory index 60678fe..d8d3882 100644 --- a/Theory +++ b/Theory @@ -323,7 +323,7 @@ Errors in the tz database arise from many sources: * Sometimes historical timekeeping was specified more precisely than what the tz database can handle. For example, from 1909 to - 1937 Netherlands clocks were legally UT+00:19:32.13, but the tz + 1937 Netherlands clocks were legally +00:19:32.13, but the tz database cannot represent the fractional second. * Even when all the timestamp transitions recorded by the tz database diff --git a/africa b/africa index 50f29d5..43c8e23 100644 --- a/africa +++ b/africa @@ -464,7 +464,7 @@ Zone Africa/Monrovia -0:43:08 - LMT 1882 # http://www.libyaherald.com/2013/10/24/correction-no-time-change-tomorrow/ # # From Paul Eggert (2013-10-25): -# For now, assume they're reverting to the pre-2012 rules of permanent UTC+2. +# For now, assume they're reverting to the pre-2012 rules of permanent +02. # Rule NAME FROM TO TYPE IN ON AT SAVE LETTER/S Rule Libya 1951 only - Oct 14 2:00 1:00 S diff --git a/asia b/asia index ee2c436..a935476 100644 --- a/asia +++ b/asia @@ -380,7 +380,7 @@ Rule PRC 1987 1991 - Apr Sun>=10 0:00 1:00 D # Lewiston (ME) Daily Sun (1939-05-29), p 17, said "Even the time is # different - the occupied districts going by Tokyo time, an hour # ahead of that prevailing in the rest of Shanghai." Guess that the -# Xujiahui Observatory was under French control and stuck with UT+8. +# Xujiahui Observatory was under French control and stuck with +08. # # In earlier versions of this file, China had many separate Zone entries, but # this was based on what were apparently incorrect data in Shanks & Pottenger. @@ -389,26 +389,26 @@ Rule PRC 1987 1991 - Apr Sun>=10 0:00 1:00 D # Proposed in 1918 and theoretically in effect until 1949 (although in practice # mainly observed in coastal areas), the five zones were: # -# Changbai Time ("Long-white Time", Long-white = Heilongjiang area) UT+8.5 +# Changbai Time ("Long-white Time", Long-white = Heilongjiang area) +08:30 # Asia/Harbin (currently a link to Asia/Shanghai) # Heilongjiang (except Mohe county), Jilin # -# Zhongyuan Time ("Central plain Time") UT+8 +# Zhongyuan Time ("Central plain Time") -08 # Asia/Shanghai # most of China # This currently represents most other zones as well, # as apparently these regions have been the same since 1970. # Milne gives 8:05:43.2 for Xujiahui Observatory time; round to nearest. -# Guo says Shanghai switched to UT+8 "from the end of the 19th century". +# Guo says Shanghai switched to +08 "from the end of the 19th century". # -# Long-shu Time (probably due to Long and Shu being two names of that area) UT+7 +# Long-shu Time (probably due to Long and Shu being two names of that area) +07 # Asia/Chongqing (currently a link to Asia/Shanghai) # Guangxi, Guizhou, Hainan, Ningxia, Sichuan, Shaanxi, and Yunnan; # most of Gansu; west Inner Mongolia; west Qinghai; and the Guangdong # counties Deqing, Enping, Kaiping, Luoding, Taishan, Xinxing, # Yangchun, Yangjiang, Yu'nan, and Yunfu. # -# Xin-zang Time ("Xinjiang-Tibet Time") UT+6 +# Xin-zang Time ("Xinjiang-Tibet Time") +06 # Asia/Urumqi # This currently represents Kunlun Time as well, # as apparently the two regions have been the same since 1970. @@ -421,7 +421,7 @@ Rule PRC 1987 1991 - Apr Sun>=10 0:00 1:00 D # Shihezi, Changji, Yanqi, Heshuo, Tuokexun, Tulufan, Shanshan, Hami, # Fukang, Kuitun, Kumukuli, Miquan, Qitai, and Turfan. # -# Kunlun Time UT+5.5 +# Kunlun Time +05:30 # Asia/Kashgar (currently a link to Asia/Urumqi) # West Tibet, including Pulan, Aheqi, Shufu, Shule; # West Xinjiang, including Aksu, Atushi, Yining, Hetian, Cele, Luopu, Nileke, @@ -493,7 +493,7 @@ Rule PRC 1987 1991 - Apr Sun>=10 0:00 1:00 D # having the same time as Beijing. # From Paul Eggert (2014-06-30): -# In the early days of the PRC, Tibet was given its own time zone (UT+6) but +# In the early days of the PRC, Tibet was given its own time zone (+06) but # this was withdrawn in 1959 and never reinstated; see Tubten Khétsun, # Memories of life in Lhasa under Chinese Rule, Columbia U Press, ISBN # 978-0231142861 (2008), translator's introduction by Matthew Akester, p x. @@ -508,12 +508,12 @@ Rule PRC 1987 1991 - Apr Sun>=10 0:00 1:00 D # Republics, the Soviet Union, the Kuomintang, and the People's Republic of # China, and tracking down all these organizations' timekeeping rules would be # quite a trick. Approximate this lost history by a transition from LMT to -# XJT at the start of 1928, the year of accession of the warlord Jin Shuren, +# +06 at the start of 1928, the year of accession of the warlord Jin Shuren, # which happens to be the date given by Shanks & Pottenger (no doubt as a -# guess) as the transition from LMT. Ignore the usage of UT+8 before -# 1986-02-01 under the theory that the transition date to UT+8 is unknown and +# guess) as the transition from LMT. Ignore the usage of +08 before +# 1986-02-01 under the theory that the transition date to +08 is unknown and # that the sort of users who prefer Asia/Urumqi now typically ignored the -# UT+8 mandate back then. +# +08 mandate back then. # Zone NAME GMTOFF RULES FORMAT [UNTIL] # Beijing time, used throughout China; represented by Shanghai. @@ -718,7 +718,7 @@ Zone Asia/Hong_Kong 7:36:42 - LMT 1904 Oct 30 # be found from historical government announcement database. # From Paul Eggert (2014-07-03): -# As per Yu-Cheng Chuang, say that Taiwan was at UT+9 from 1937-10-01 +# As per Yu-Cheng Chuang, say that Taiwan was at +09 from 1937-10-01 # until 1945-09-21 at 01:00, overriding Shanks & Pottenger. # Likewise, use Yu-Cheng Chuang's data for DST in Taiwan. @@ -917,7 +917,7 @@ Zone Asia/Kolkata 5:53:28 - LMT 1880 # Kolkata # These would be the earliest possible times for a change. # Régimes horaires pour le monde entier, by Henri Le Corre, (Éditions # Traditionnelles, 1987, Paris) says that Java and Madura switched -# from JST to UTC+07:30 on 1945-09-23, and gives 1944-09-01 for Jayapura +# from +09 to +07:30 on 1945-09-23, and gives 1944-09-01 for Jayapura # (Hollandia). For now, assume all Indonesian locations other than Jayapura # switched on 1945-09-23. # @@ -930,9 +930,9 @@ Zone Asia/Kolkata 5:53:28 - LMT 1880 # Kolkata # Indonesia, <http://time.kim.lipi.go.id/time-eng.php> (2006-09-29). # The abbreviations are: # -# WIB - UTC+7 - Waktu Indonesia Barat (Indonesia western time) -# WITA - UTC+8 - Waktu Indonesia Tengah (Indonesia central time) -# WIT - UTC+9 - Waktu Indonesia Timur (Indonesia eastern time) +# WIB - +07 - Waktu Indonesia Barat (Indonesia western time) +# WITA - +08 - Waktu Indonesia Tengah (Indonesia central time) +# WIT - +09 - Waktu Indonesia Timur (Indonesia eastern time) # # Zone NAME GMTOFF RULES FORMAT [UNTIL] # Java, Sumatra @@ -1864,25 +1864,24 @@ Rule ROK 1957 1960 - Sep Sun>=18 0:00 0 S Rule ROK 1987 1988 - May Sun>=8 2:00 1:00 D Rule ROK 1987 1988 - Oct Sun>=8 3:00 0 S -# From Paul Eggert (2014-10-30): +# From Paul Eggert (2016-08-23): # The Korean Wikipedia entry gives the following sources for UT offsets: # -# 1908: Official Journal Article No. 3994 (Edict No. 5) +# 1908: Official Journal Article No. 3994 (decree No. 5) # 1912: Governor-General of Korea Official Gazette Issue No. 367 # (Announcement No. 338) # 1954: Presidential Decree No. 876 (1954-03-17) # 1961: Law No. 676 (1961-08-07) -# 1987: Law No. 3919 (1986-12-31) # -# The Wikipedia entry also has confusing information about a change -# to UT+9 in April 1910, but then what would be the point of the later change -# to UT+9 on 1912-01-01? Omit the 1910 change for now. +# (Another source "1987: Law No. 3919 (1986-12-31)" was in the 2014-10-30 +# edition of the Korean Wikipedia entry.) # # I guessed that time zone abbreviations through 1945 followed the same # rules as discussed under Taiwan, with nominal switches from JST to KST # when the respective cities were taken over by the Allies after WWII. # -# For Pyongyang we have no information; guess no changes since World War II. +# For Pyongyang, guess no changes from World War II until 2015, as we +# have no information otherwise. # From Steffen Thorsen (2015-08-07): # According to many news sources, North Korea is going to change to @@ -2042,7 +2041,7 @@ Zone Indian/Maldives 4:54:00 - LMT 1880 # Male # Bill Bonnet (2005-05-19) reports that the US Embassy in Ulaanbaatar says # there is only one time zone and that DST is observed, citing Microsoft # Windows XP as the source. Risto Nykänen (2005-05-16) reports that -# travelmongolia.org says there are two time zones (UTC+7, UTC+8) with no DST. +# travelmongolia.org says there are two time zones (+07, +08) with no DST. # Oscar van Vlijmen (2005-05-20) reports that the Mongolian Embassy in # Washington, DC says there are two time zones, with DST observed. # He also found @@ -2678,7 +2677,7 @@ Link Asia/Qatar Asia/Bahrain # earlier date. # # Shanks & Pottenger also state that until 1968-05-01 Saudi Arabia had two -# time zones; the other zone, at UTC+4, was in the far eastern part of +# time zones; the other zone, at +04, was in the far eastern part of # the country. Ignore this, as it's before our 1970 cutoff. # # Zone NAME GMTOFF RULES FORMAT [UNTIL] diff --git a/australasia b/australasia index 0b33f67..fe77d07 100644 --- a/australasia +++ b/australasia @@ -545,7 +545,7 @@ Zone Pacific/Port_Moresby 9:48:40 - LMT 1880 # Base the Bougainville entry on the Arawa-Kieta region, which appears to have # the most people even though it was devastated in the Bougainville Civil War. # -# Although Shanks gives 1942-03-15 / 1943-11-01 for JST, these dates +# Although Shanks gives 1942-03-15 / 1943-11-01 for +09, these dates # are apparently rough guesswork from the starts of military campaigns. # The World War II entries below are instead based on Arawa-Kieta. # The Japanese occupied Kieta in July 1942, @@ -553,8 +553,8 @@ Zone Pacific/Port_Moresby 9:48:40 - LMT 1880 # http://pwencycl.kgbudge.com/B/o/Bougainville.htm # and seem to have controlled it until their 1945-08-21 surrender. # -# The Autonomous Region of Bougainville plans to switch from UTC+10 to UTC+11 -# on 2014-12-28 at 02:00. They call UTC+11 "Bougainville Standard Time"; +# The Autonomous Region of Bougainville switched from +10 to +11 +# on 2014-12-28 at 02:00. They call +11 "Bougainville Standard Time"; # abbreviate this as BST. See: # http://www.bougainville24.com/bougainville-issues/bougainville-gets-own-tim… # @@ -620,7 +620,7 @@ Link Pacific/Pago_Pago Pacific/Midway # in US minor outlying islands # From Paul Eggert (2014-06-27): # The International Date Line Act 2011 # http://www.parliament.gov.ws/images/ACTS/International_Date_Line_Act__2011_… -# changed Samoa from UTC-11 to UTC+13, effective "12 o'clock midnight, on +# changed Samoa from -11 to +13, effective "12 o'clock midnight, on # Thursday 29th December 2011". The International Date Line was adjusted # accordingly. @@ -715,7 +715,7 @@ Zone Pacific/Funafuti 11:56:52 - LMT 1901 # 1886-1891; Baker was similar but exact dates are not known. # Inhabited by civilians 1935-1942; U.S. military bases 1943-1944; # uninhabited thereafter. -# Howland observed Hawaii Standard Time (UT-10:30) in 1937; +# Howland observed Hawaii Standard Time (-10:30) in 1937; # see page 206 of Elgen M. Long and Marie K. Long, # Amelia Earhart: the Mystery Solved, Simon & Schuster (2000). # So most likely Howland and Baker observed Hawaii Time from 1935 @@ -1473,7 +1473,7 @@ Zone Pacific/Wallis 12:15:20 - LMT 1901 # Zealand time. I understand that is the time they keep locally, anyhow." # For now, assume this practice goes back to the introduction of standard time # in New Zealand, as this would make Chatham Islands time almost exactly match -# LMT back when New Zealand was at UTC+11:30; also, assume Chatham Islands did +# LMT back when New Zealand was at +11:30; also, assume Chatham Islands did # not observe New Zealand's prewar DST. ############################################################################### @@ -1529,7 +1529,7 @@ Zone Pacific/Wallis 12:15:20 - LMT 1901 # For now, we assume the Ladrones switched at the same time as the Philippines; # see Asia/Manila. -# US Public Law 106-564 (2000-12-23) made UTC+10 the official standard time, +# US Public Law 106-564 (2000-12-23) made +10 the official standard time, # under the name "Chamorro Standard Time". There is no official abbreviation, # but Congressman Robert A. Underwood, author of the bill that became law, # wrote in a press release (2000-12-27) that he will seek the use of "ChST". @@ -1541,15 +1541,15 @@ Zone Pacific/Wallis 12:15:20 - LMT 1901 # "I am certain, having lived there for the past decade, that 'Truk' # (now properly known as Chuuk) ... is in the time zone GMT+10." # -# Shanks & Pottenger write that Truk switched from UTC+10 to UTC+11 +# Shanks & Pottenger write that Truk switched from +10 to +11 # on 1978-10-01; ignore this for now. # From Paul Eggert (1999-10-29): # The Federated States of Micronesia Visitors Board writes in # The Federated States of Micronesia - Visitor Information (1999-01-26) # http://www.fsmgov.org/info/clocks.html -# that Truk and Yap are UTC+10, and Ponape and Kosrae are UTC+11. -# We don't know when Kosrae switched from UTC+12; assume January 1 for now. +# that Truk and Yap are +10, and Ponape and Kosrae are +11. +# We don't know when Kosrae switched from +12; assume January 1 for now. # Midway @@ -1615,11 +1615,11 @@ Zone Pacific/Wallis 12:15:20 - LMT 1901 # ordaining - by a masterpiece of diplomatic flattery - that # the Fourth of July should be celebrated twice in that year." -# Although Shanks & Pottenger says they both switched to UTC-11:30 -# in 1911, and to UTC-11 in 1950. many earlier sources give UTC-11 +# Although Shanks & Pottenger says they both switched to -11:30 +# in 1911, and to -11 in 1950. many earlier sources give -11 # for American Samoa, e.g., the US National Bureau of Standards # circular "Standard Time Throughout the World", 1932. -# Assume American Samoa switched to UTC-11 in 1911, not 1950, +# Assume American Samoa switched to -11 in 1911, not 1950, # and that after 1950 they agreed until (western) Samoa skipped a # day in 2011. Assume also that the Samoas follow the US and New # Zealand's "ST"/"DT" style of daylight-saving abbreviations. diff --git a/backzone b/backzone index 13dc8d4..27a5472 100644 --- a/backzone +++ b/backzone @@ -194,9 +194,9 @@ Zone Africa/Lusaka 1:53:08 - LMT 1903 Mar # Equatorial Guinea # -# Although Shanks says that Malabo switched from UTC to UTC+1 on 1963-12-15, +# Although Shanks says that Malabo switched from +00 to +01 on 1963-12-15, # a Google Books search says that London Calling, Issues 432-465 (1948), p 19, -# says that Spanish Guinea was at GMT+1 back then. The Shanks data entries +# says that Spanish Guinea was at +01 back then. The Shanks data entries # are most likely wrong, but we have nothing better; use them here for now. # Zone Africa/Malabo 0:35:08 - LMT 1912 @@ -479,14 +479,14 @@ Zone Asia/Muscat 3:54:24 - LMT 1920 # From Paul Eggert (2014-08-11), after a heads-up from Stephen Colebourne: # According to a Portuguese decree (1911-05-26) # http://dre.pt/pdf1sdip/1911/05/12500/23132313.pdf -# Portuguese India switched to GMT+5 on 1912-01-01. +# Portuguese India switched to +05 on 1912-01-01. #Zone Asia/Panaji [not enough info to complete] # Cambodia # From Paul Eggert (2014-10-11): # See Asia/Ho_Chi_Minh for the source for most of this data. Also, guess -# (1) Cambodia reverted to UT+7 on 1945-09-02, when Vietnam did, and -# (2) they also reverted to UT+7 on 1953-11-09, the date of independence. +# (1) Cambodia reverted to +07 on 1945-09-02, when Vietnam did, and +# (2) they also reverted to +07 on 1953-11-09, the date of independence. # These guesses are probably wrong but they're better than guessing no # transitions there. Zone Asia/Phnom_Penh 6:59:40 - LMT 1906 Jul 1 @@ -506,8 +506,8 @@ Zone Asia/Tel_Aviv 2:19:04 - LMT 1880 # Laos # From Paul Eggert (2014-10-11): # See Asia/Ho_Chi_Minh for the source for most of this data. -# Trần's book says that Laos reverted to UT+7 on 1955-04-15. -# Also, guess that Laos reverted to UT+7 on 1945-09-02, when Vietnam did; +# Trần's book says that Laos reverted to +07 on 1955-04-15. +# Also, guess that Laos reverted to +07 on 1945-09-02, when Vietnam did; # this is probably wrong but it's better than guessing no transition. Zone Asia/Vientiane 6:50:24 - LMT 1906 Jul 1 7:06:30 - PLMT 1911 May 1 diff --git a/europe b/europe index b02edbd..4219828 100644 --- a/europe +++ b/europe @@ -774,8 +774,6 @@ Zone Europe/Vienna 1:05:21 - LMT 1893 Apr # From Alexander Bokovoy (2014-10-09): # Belarussian government decided against changing to winter time.... # http://eng.belta.by/all_news/society/Belarus-decides-against-adjusting-time… -# From Paul Eggert (2014-10-08): -# Hence Belarus can share time zone abbreviations with Moscow again. # # Zone NAME GMTOFF RULES FORMAT [UNTIL] Zone Europe/Minsk 1:50:16 - LMT 1880 @@ -1294,7 +1292,7 @@ Zone Europe/Paris 0:09:21 - LMT 1891 Mar 15 0:01 # http://www.parlament-berlin.de/pds-fraktion.nsf/727459127c8b66ee85256623004… # says that Bersarin issued an order to use Moscow time on May 20. # However, Moscow did not observe daylight saving in 1945, so -# this was equivalent to CEMT (GMT+3), not GMT+4. +# this was equivalent to +03, not +04. # Rule NAME FROM TO TYPE IN ON AT SAVE LETTER/S @@ -2258,7 +2256,6 @@ Zone Europe/Bucharest 1:44:24 - LMT 1891 Oct # http://www.worldtimezone.com/dst_news/dst_news_russia-map-2014-07.html # From Paul Eggert (2006-03-22): -# Except for Moscow after 1919-07-01, I invented the time zone abbreviations. # Moscow time zone abbreviations after 1919-07-01, and Moscow rules after 1991, # are from Andrey A. Chernov. The rest is from Shanks & Pottenger, # except we follow Chernov's report that 1992 DST transitions were Sat @@ -3080,9 +3077,6 @@ Zone Asia/Magadan 10:03:12 - LMT 1924 May 2 # in Russian.) In addition, Srednekolymsk appears to be a much older # settlement and the population of Zyryanka seems to be declining. # Go with Srednekolymsk. -# -# Since Magadan Oblast moves to UTC+10 on 2014-10-26, we cannot keep using MAGT -# as the abbreviation. Use SRET instead. Zone Asia/Srednekolymsk 10:14:52 - LMT 1924 May 2 10:00 - +10 1930 Jun 21 diff --git a/localtime.c b/localtime.c index 276ce34..6c00c45 100644 --- a/localtime.c +++ b/localtime.c @@ -1510,7 +1510,7 @@ gmtsub(struct state const *sp, time_t const *timep, int_fast32_t offset, #ifdef TM_ZONE /* ** Could get fancy here and deliver something such as - ** "UT+xxxx" or "UT-xxxx" if offset is non-zero, + ** "+xx" or "-xx" if offset is non-zero, ** but this is no time for a treasure hunt. */ tmp->TM_ZONE = ((char *) diff --git a/northamerica b/northamerica index d66135e..0c069bb 100644 --- a/northamerica +++ b/northamerica @@ -3312,7 +3312,7 @@ Zone America/Miquelon -3:44:40 - LMT 1911 May 15 # St Pierre # indicating that the normal ET rules are followed. # # From Paul Eggert (2014-08-19): -# The 2014-08-13 Cabinet meeting decided to stay on UTC-4 year-round. See: +# The 2014-08-13 Cabinet meeting decided to stay on -04 year-round. See: # http://tcweeklynews.com/daylight-savings-time-to-be-maintained-p5353-127.htm # Model this as a switch from EST/EDT to AST ... # From Chris Walton (2014-11-04): diff --git a/private.h b/private.h index 941e91b..3583079 100644 --- a/private.h +++ b/private.h @@ -15,6 +15,7 @@ ** Thank you! */ +/* This string was in the tzdata Factory zone through version 2016f. */ #define GRANDPARENTED "Local time zone must be set--see zic manual page" /* diff --git a/southamerica b/southamerica index 1c38f63..5321451 100644 --- a/southamerica +++ b/southamerica @@ -410,9 +410,9 @@ Rule Arg 2008 only - Oct Sun>=15 0:00 1:00 S # stuck on Summer daylight savings time even though the summer is over. # From Paul Eggert (2013-09-05): -# Perhaps San Luis operates on the legal fiction that it is at UTC-4 +# Perhaps San Luis operates on the legal fiction that it is at -04 # with perpetual summer time, but ordinary usage typically seems to -# just say it's at UTC-3; see, for example, +# just say it's at -03; see, for example, # http://es.wikipedia.org/wiki/Hora_oficial_argentina # We've documented similar situations as being plain changes to # standard time, so let's do that here too. This does not change UTC -- 2.5.5

12 28