- tz - lists.iana.org

Fiji DST 2015/16
by Ken Rylander Sept. 2, 2015

Sept. 2, 2015

Hi Paul, Some changes to the Fiji DST 2015/16 were communicated to us. See attached order from Fiji Government. There is nothing yet on their website, http://www.fiji.gov.fj, so I can't provide a link yet. The attached order states: "the daylight saving period is 1 hour in advance of the standard time commencing at 2.00 am on Sunday 1st November, 2015 and ending at 3.00 am on Sunday 17th January, 2016." Fiji rules in tzdata2015f: Rule Fiji 2014 max - Nov Sun>=1 2:00 1:00 S Rule Fiji 2015 max - Jan Sun>=18 3:00 0 - I guess that means that the start of DST on Nov 1, will work fine. But the second rule need to be changed, so DST will on Jan 17, 2016. Regards, Ken Rylander | Solution Development and BI Manager - Digicel Pacific | Digicel (Fiji) Ltd | P.O.Box 13811, Suva, Fiji | Ground Floor, Kadavu House, Victoria Parade, Suva, Fiji | Mobile (679) 701 5353 | Fax (679) 334 3733 | ken.rylander(a)digicelgroup.com<mailto:ken.rylander@digicelgroup.com> | www.digicelpacific.com<http://www.digicelpacific.com/> ________________________________ Notice of Confidentiality: The information contained in this communication is intended solely for the use of the individual or entity to whom it is addressed and others authorized to receive it. It may contain confidential or legally privileged information. If you are not the intended recipient you are hereby notified that any disclosure, copying, distribution or taking any action in reliance on the contents of this information is strictly prohibited and may be unlawful. If you have received this communication in error, please notify us immediately by responding to this email and then delete it from your system.

2 2

[PROPOSED PATCH] Fix localtime bug with Anchorage after 2037
by Paul Eggert Aug. 28, 2015

Aug. 28, 2015

* NEWS: Document this. * localtime.c (tzloadbody): Handle America/Anchorage after 2037, even though the sum of the abbreviations' sizes (42) plus the sum of the extended abbreviations' sizes (10) exceeds TZ_MAX_CHARS (50). Do this by reusing existing abbreviations. Thanks to Bradley White for reporting the bug. Perhaps we should also consider increasing TZ_MAX_CHARS from its currently-low value. --- NEWS | 8 +++++++ localtime.c | 71 +++++++++++++++++++++++++++++++++++++++---------------------- 2 files changed, 54 insertions(+), 25 deletions(-) diff --git a/NEWS b/NEWS index b477250..ffe3fb2 100644 --- a/NEWS +++ b/NEWS @@ -1,5 +1,13 @@ News for the tz database +Unreleased, experimental changes + + Changes affecting code + + localtime no longer mishandles America/Anchorage after 2037. + (Thanks to Bradley White for reporting the bug.) + + Release 2015f - 2015-08-10 18:06:56 -0700 Changes affecting future time stamps diff --git a/localtime.c b/localtime.c index ded8f7b..ac34e5d 100644 --- a/localtime.c +++ b/localtime.c @@ -569,31 +569,52 @@ tzloadbody(char const *name, struct state *sp, bool doextend, up->buf[nread - 1] = '\0'; if (tzparse(&up->buf[1], ts, false) - && ts->typecnt == 2 - && sp->charcnt + ts->charcnt <= TZ_MAX_CHARS) { - for (i = 0; i < 2; ++i) - ts->ttis[i].tt_abbrind += - sp->charcnt; - for (i = 0; i < ts->charcnt; ++i) - sp->chars[sp->charcnt++] = - ts->chars[i]; - i = 0; - while (i < ts->timecnt && - ts->ats[i] <= - sp->ats[sp->timecnt - 1]) - ++i; - while (i < ts->timecnt && - sp->timecnt < TZ_MAX_TIMES) { - sp->ats[sp->timecnt] = - ts->ats[i]; - sp->types[sp->timecnt] = - sp->typecnt + - ts->types[i]; - ++sp->timecnt; - ++i; - } - sp->ttis[sp->typecnt++] = ts->ttis[0]; - sp->ttis[sp->typecnt++] = ts->ttis[1]; + && ts->typecnt == 2) { + + /* Attempt to reuse existing abbreviations. + Without this, America/Anchorage would stop + working after 2037 when TZ_MAX_CHARS is 50, as + sp->charcnt equals 42 (for LMT CAT CAWT CAPT AHST + AHDT YST AKDT AKST) and ts->charcnt equals 10 + (for AKST AKDT). Reusing means sp->charcnt can + stay 42 in this example. */ + int gotabbr = 0; + int charcnt = sp->charcnt; + for (i = 0; i < 2; i++) { + char *tsabbr = ts->chars + ts->ttis[i].tt_abbrind; + int j; + for (j = 0; j < charcnt; j++) + if (strcmp(sp->chars + j, tsabbr) == 0) { + ts->ttis[i].tt_abbrind = j; + gotabbr++; + break; + } + if (! (j < charcnt)) { + int tsabbrlen = strlen(tsabbr); + if (j + tsabbrlen < TZ_MAX_CHARS) { + strcpy(sp->chars + j, tsabbr); + charcnt = j + tsabbrlen + 1; + ts->ttis[i].tt_abbrind = j; + gotabbr++; + } + } + } + if (gotabbr == 2) { + sp->charcnt = charcnt; + for (i = 0; i < ts->timecnt; i++) + if (sp->ats[sp->timecnt - 1] < ts->ats[i]) + break; + while (i < ts->timecnt + && sp->timecnt < TZ_MAX_TIMES) { + sp->ats[sp->timecnt] = ts->ats[i]; + sp->types[sp->timecnt] = (sp->typecnt + + ts->types[i]); + sp->timecnt++; + i++; + } + sp->ttis[sp->typecnt++] = ts->ttis[0]; + sp->ttis[sp->typecnt++] = ts->ttis[1]; + } } } if (sp->timecnt > 1) { -- 2.1.4

1 0

North Korea ultimatum deadline time confusion
by Paul Eggert Aug. 22, 2015

Aug. 22, 2015

North Korea has delivered a 48-hour ultimatum to the South Korean Defense Ministry, saying that it will respond militarily unless South Korea stops broadcasting propaganda via loudspeakers over the border. There seems to be some confusion about the ultimatum's exact deadline. Ju-Min Park and James Person of Reuters report that although the ultimatum is "uncharacteristically specific", the deadline is "around 5 p.m. (0800 GMT) on Saturday in Seoul". Could the time confusion be due to the recent time zone change in North Korea? Park JM, Pearson J. North Korea goes on war footing against South Korea as deadline looms. Reuters 2015-08-21 12:53 -0400. http://www.reuters.com/article/2015/08/21/us-northkorea-southkorea-border-i… $ date -u; TZ=Asia/Pyongyang date; TZ=Asia/Seoul date Fri Aug 21 17:12:37 UTC 2015 Sat Aug 22 01:42:37 KST 2015 Sat Aug 22 02:12:37 KST 2015

2 3

Time zone humor
by Matt Johnson Aug. 18, 2015

Aug. 18, 2015

For the art file http://www.maniacworld.com/time-zones-explained.html Enjoy!

3 4

tzfiles contain Unix epoch for the first transition time
by Eric Erhardt Aug. 15, 2015

Aug. 15, 2015

I am working on enabling the .NET TimeZoneInfo class to read time zone information from tzfiles. I've hit a snag with the latest tzdata 2015f. (I'm not sure when this change started, but the problem doesn't occur with the tzfiles that are shipped with an Ubuntu 14.04 distribution.) The problem is that the 2015f version of the tzdata contains an initial "Transition Time" that is out of order. The beginning of the America/Chicago tzfile looks like the following: Transition Time Transition Offset 01/01/1970 00:00:00 -05:50:36 11/18/1883 18:00:00 -06:00:00 03/31/1918 08:00:00 -05:00:00 DST 10/27/1918 07:00:00 -06:00:00 Notice the first entry is for 1970, and then the next entry is for 1883. This breaks the documentation in 'man tzfile': The above header is followed by tzh_timecnt four-byte values of type long, sorted in ascending order. These values are written in "standard" byte order. Each is used as a transition time (as returned by time(2)) at which the rules for computing local time change. This causes the TimeZoneInfo parsing code to throw an exception because it is assuming these transitions are sorted in ascending order. Is this an intentional change in the tzfiles? If so, will the tzfile man page be updated for this change? Eric Erhardt

8 18

[PROPOSED PATCH] Fix comments re NK transition time, abbreviation
by Paul Eggert Aug. 15, 2015

Aug. 15, 2015

* asia (Asia/Pyongyang): Confirm 00:00 as transition time. Fix commentary about abbreviation. --- asia | 11 ++++++----- 1 file changed, 6 insertions(+), 5 deletions(-) diff --git a/asia b/asia index 4f8756b..78eed56 100644 --- a/asia +++ b/asia @@ -1717,11 +1717,12 @@ Rule ROK 1987 1988 - Oct Sun>=8 3:00 0 S # the 8:30 time zone on August 15, one example: # http://www.bbc.com/news/world-asia-33815049 # -# From Paul Eggert (2015-08-07): -# No transition time is specified; assume 00:00. +# From Paul Eggert (2015-08-15): +# Bells rang out midnight (00:00) Friday as part of the celebrations. See: +# Talmadge E. North Korea celebrates new time zone, 'Pyongyang Time' +# http://news.yahoo.com/north-korea-celebrates-time-zone-pyongyang-time-16403… # There is no common English-language abbreviation for this time zone. -# Use %z rather than invent one. We can't assume %z works everywhere yet, -# so for now substitute its output manually. +# Use KST, as that's what we already use for 1954-1961 in ROK. # Zone NAME GMTOFF RULES FORMAT [UNTIL] Zone Asia/Seoul 8:27:52 - LMT 1908 Apr 1 @@ -1735,7 +1736,7 @@ Zone Asia/Pyongyang 8:23:00 - LMT 1908 Apr 1 8:30 - KST 1912 Jan 1 9:00 - JCST 1937 Oct 1 9:00 - JST 1945 Aug 24 - 9:00 - KST 2015 Aug 15 + 9:00 - KST 2015 Aug 15 00:00 8:30 - KST ############################################################################### -- 2.1.4

1 0

Is there a zip archive withe the TZ zoneinfo data available ?
by Dave Sweeney Aug. 14, 2015

Aug. 14, 2015

Hello, I was wondering if IANA or a trusted partner provides a zip format of the TZ data and source file? Thank You and Regards, Dave Sweeney +1 (978) 451-0895 VMS Software Inc.

2 2

propagation of North Korea and Uruguay changes
by Paul Eggert Aug. 13, 2015

Aug. 13, 2015

As part of routine system updates my Ubuntu 15.04 desktop now has 2015f's changes for North Korea and Uruguay. Ubuntu is downstream from Debian, so I assume other Debian-derived distributions also have the update; <http://metadata.ftp-master.debian.org/changelogs//main/t/tzdata/tzdata_2015…> says the Debian patch is dated 2015-08-11 19:56:35 +0200. <https://access.redhat.com/articles/1187353> says Red Hat takes at least five business days for QA and estimates their patch will be available for installation by tomorrow, which would be right around the time of North Korea's time zone change. Microsoft says an update for Windows in North Korea is planned by early September <http://blogs.technet.com/b/dst2007/archive/2015/08/07/upcoming-windows-dst-…>. I assume North Korea's own Red Star OS will be updated in time, though I don't know how to check this. <https://en.wikipedia.org/wiki/Red_Star_OS>

1 0

[PROPOSED PATCH] Support %z in Zone formats
by Paul Eggert Aug. 12, 2015

Aug. 12, 2015

This extends the zic input format to add support for %z, which expands to a UTC offset in as-short-as-possible ISO 8601 format. It's intended to better support zones that do not have an established abbreviation already. This is a change only to zic; it does not affect the data, which can still be compiled with older zic versions. * NEWS, zic.8: Document this. * zic.c (struct zone): New member z_format_specifier. (PERCENT_Z_LEN_BOUND): New constant. (max_abbrvar_len): Initialize to it, rather than to 0. (associate, inzsub, doabbr): Add support for %z. (abbroffset): New function. (doabbr): 2nd arg is now struct zone *, not a char *. All callers changed. --- NEWS | 5 +++++ zic.8 | 13 +++++++++++ zic.c | 82 ++++++++++++++++++++++++++++++++++++++++++++++++++++++------------- 3 files changed, 85 insertions(+), 15 deletions(-) diff --git a/NEWS b/NEWS index b7593e7..9eff28c 100644 --- a/NEWS +++ b/NEWS @@ -25,6 +25,11 @@ Unreleased, experimental changes (Thanks to Jon Skeet and Arthur David Olson.) Constraints on simultaneity are now documented. + The two characters '%z' in a zone format now stand for the UTC + offset, e.g., '-07' for seven hours behind UTC and '+0530' for + five hours and thirty minutes ahead. This better supports time + zone abbreviations conforming to POSIX.1-2001 and later. + Changes affecting installed data files Data entries have been simplified for Atlantic/Canary, Europe/Simferopol, diff --git a/zic.8 b/zic.8 index dfb01db..94b6753 100644 --- a/zic.8 +++ b/zic.8 @@ -350,6 +350,19 @@ The pair of characters is used to show where the .q "variable part" of the time zone abbreviation goes. +Alternately, a format can use the pair of characters +.B %z +to stand for the UTC offset in the form +.RI \(+- hh , +.RI \(+- hhmm , +or +.RI \(+- hhmmss , +using the shortest form that does not lose information, where +.IR hh , +.IR mm , +and +.I ss +are the hours, minutes, and seconds east (+) or west (\(mi) of UTC. Alternately, a slash (/) separates standard and daylight abbreviations. diff --git a/zic.c b/zic.c index 6854033..3c4ae93 100644 --- a/zic.c +++ b/zic.c @@ -76,6 +76,7 @@ struct zone { zic_t z_gmtoff; const char * z_rule; const char * z_format; + char z_format_specifier; zic_t z_stdoff; @@ -130,6 +131,9 @@ static void rulesub(struct rule * rp, static zic_t tadd(zic_t t1, zic_t t2); static bool yearistype(int year, const char * type); +/* Bound on length of what %z can expand to. */ +enum { PERCENT_Z_LEN_BOUND = sizeof "+995959" - 1 }; + static int charcnt; static bool errors; static bool warnings; @@ -139,7 +143,7 @@ static bool leapseen; static zic_t leapminyear; static zic_t leapmaxyear; static int linenum; -static int max_abbrvar_len; +static int max_abbrvar_len = PERCENT_Z_LEN_BOUND; static int max_format_len; static zic_t max_year; static zic_t min_year; @@ -955,7 +959,7 @@ associate(void) ** Note, though, that if there's no rule, ** a '%s' in the format is a bad thing. */ - if (strchr(zp->z_format, '%') != 0) + if (zp->z_format_specifier == 's') error("%s", _("%s in ruleless zone")); } } @@ -1170,6 +1174,7 @@ static bool inzsub(char **fields, int nfields, bool iscont) { register char * cp; + char * cp1; static struct zone z; register int i_gmtoff, i_rule, i_format; register int i_untilyear, i_untilmonth; @@ -1201,13 +1206,21 @@ inzsub(char **fields, int nfields, bool iscont) z.z_linenum = linenum; z.z_gmtoff = gethms(fields[i_gmtoff], _("invalid UT offset"), true); if ((cp = strchr(fields[i_format], '%')) != 0) { - if (*++cp != 's' || strchr(cp, '%') != 0) { + if ((*++cp != 's' && *cp != 'z') || strchr(cp, '%') + || strchr(fields[i_format], '/')) { error(_("invalid abbreviation format")); return false; } } z.z_rule = ecpyalloc(fields[i_rule]); - z.z_format = ecpyalloc(fields[i_format]); + z.z_format = cp1 = ecpyalloc(fields[i_format]); + z.z_format_specifier = cp ? *cp : '\0'; + if (z.z_format_specifier == 'z') { + if (noise) + warning(_("format '%s' not handled by pre-2015 versions of zic"), + z.z_format); + cp1[cp - fields[i_format]] = 's'; + } if (max_format_len < strlen(z.z_format)) max_format_len = strlen(z.z_format); hasuntil = nfields > i_untilyear; @@ -1890,19 +1903,59 @@ writezone(const char *const name, const char *const string, char version) free(fullname); } +static char const * +abbroffset(char *buf, zic_t offset) +{ + char sign = '+'; + int seconds, minutes; + + if (offset < 0) { + offset = -offset; + sign = '-'; + } + + seconds = offset % SECSPERMIN; + offset /= SECSPERMIN; + minutes = offset % MINSPERHOUR; + offset /= MINSPERHOUR; + if (100 <= offset) { + error(_("%%z UTC offset magnitude exceeds 99:59:59")); + return "%z"; + } else { + char *p = buf; + *p++ = sign; + *p++ = '0' + offset / 10; + *p++ = '0' + offset % 10; + if (minutes | seconds) { + *p++ = '0' + minutes / 10; + *p++ = '0' + minutes % 10; + if (seconds) { + *p++ = '0' + seconds / 10; + *p++ = '0' + seconds % 10; + } + } + *p = '\0'; + return buf; + } +} + static size_t -doabbr(char *const abbr, const char *const format, const char *const letters, +doabbr(char *abbr, struct zone const *zp, char const *letters, bool isdst, bool doquotes) { register char * cp; register char * slashp; register size_t len; + char const *format = zp->z_format; slashp = strchr(format, '/'); if (slashp == NULL) { - if (letters == NULL) - strcpy(abbr, format); - else sprintf(abbr, format, letters); + char letterbuf[PERCENT_Z_LEN_BOUND + 1]; + if (zp->z_format_specifier == 'z') + letters = abbroffset(letterbuf, -zp->z_gmtoff); + else if (!letters) + letters = "%s"; + sprintf(abbr, format, letters); } else if (isdst) { strcpy(abbr, slashp + 1); } else { @@ -2127,7 +2180,7 @@ stringzone(char *result, const struct zone *const zpfirst, const int zonecount) if (stdrp == NULL && (zp->z_nrules != 0 || zp->z_stdoff != 0)) return -1; abbrvar = (stdrp == NULL) ? "" : stdrp->r_abbrvar; - len = doabbr(result, zp->z_format, abbrvar, false, true); + len = doabbr(result, zp, abbrvar, false, true); offsetlen = stringoffset(result + len, -zp->z_gmtoff); if (! offsetlen) { result[0] = '\0'; @@ -2136,7 +2189,7 @@ stringzone(char *result, const struct zone *const zpfirst, const int zonecount) len += offsetlen; if (dstrp == NULL) return compat; - len += doabbr(result + len, zp->z_format, dstrp->r_abbrvar, true, true); + len += doabbr(result + len, zp, dstrp->r_abbrvar, true, true); if (dstrp->r_stdoff != SECSPERMIN * MINSPERHOUR) { offsetlen = stringoffset(result + len, -(zp->z_gmtoff + dstrp->r_stdoff)); @@ -2307,8 +2360,7 @@ outzone(const struct zone * const zpfirst, const int zonecount) startoff = zp->z_gmtoff; if (zp->z_nrules == 0) { stdoff = zp->z_stdoff; - doabbr(startbuf, zp->z_format, - NULL, stdoff != 0, false); + doabbr(startbuf, zp, NULL, stdoff != 0, false); type = addtype(oadd(zp->z_gmtoff, stdoff), startbuf, stdoff != 0, startttisstd, startttisgmt); @@ -2400,7 +2452,7 @@ outzone(const struct zone * const zpfirst, const int zonecount) if (ktime < starttime) { startoff = oadd(zp->z_gmtoff, stdoff); - doabbr(startbuf, zp->z_format, + doabbr(startbuf, zp, rp->r_abbrvar, rp->r_stdoff != 0, false); @@ -2410,7 +2462,7 @@ outzone(const struct zone * const zpfirst, const int zonecount) startoff == oadd(zp->z_gmtoff, stdoff)) { doabbr(startbuf, - zp->z_format, + zp, rp->r_abbrvar, rp->r_stdoff != false, @@ -2419,7 +2471,7 @@ outzone(const struct zone * const zpfirst, const int zonecount) } eats(zp->z_filename, zp->z_linenum, rp->r_filename, rp->r_linenum); - doabbr(ab, zp->z_format, rp->r_abbrvar, + doabbr(ab, zp, rp->r_abbrvar, rp->r_stdoff != 0, false); offset = oadd(zp->z_gmtoff, rp->r_stdoff); type = addtype(offset, ab, rp->r_stdoff != 0, -- 2.1.4

3 4

Re: [tz] Time zone humor
by Matt Johnson Aug. 11, 2015

Aug. 11, 2015

Thanks! Sent from my Windows Phone ________________________________ From: Paul Eggert<mailto:eggert@cs.ucla.edu> Sent: ‎8/‎11/‎2015 1:11 PM To: Matt Johnson<mailto:mj1856@hotmail.com> Cc: Time zone mailing list<mailto:tz@iana.org> Subject: Re: Time zone humor Steve Jones did the legwork on the Laugh-In show number; attached is a proposed patch. The git commit log is "* tz-art.htm: Add Goldie Hawn. (Thanks to Matt Johnson & Steve Jones.)"

1 0