- tz - lists.iana.org

Hiccup in Easter Island 1982
by Jesper Norgaard Welen Feb. 22, 2006

Feb. 22, 2006

It seems unlikely that Easter Island changed from GMT-6 to GMT-7 for 4 hours (from 20 to 24) on 14 th. of March 1982, and then switched back to GMT-6. Rather it is more likely they stayed on the same time, by changing from one rule to another on 13.th of March at 21:00, which results in no time change at all on 13.th. or 14.th. of March in Easter Island. I don't know if this is something zic mends, but at least the Jodatime java package which uses the tz data directly, reflects this double DST change in their extracted data files, while Jodatime handles gracefully (zic-like) the hiccups of this type that only lasts for an hour (for instance the Knox Indiana change that was discussed earlier). Old rule: -7:00 Chile EAS%sT 1982 Mar 14 # Easter I Time -6:00 Chile EAS%sT New rule: -7:00 Chile EAS%sT 1982 Mar 13 21:00 # Easter I Time -6:00 Chile EAS%sT Regards, - Jesper Jesper Nørgaard Welen Email: jnorgard(a)Prodigy.Net.mx Project Leader (Líder de Proyecto) Software CIMMYT - Centro Internacional de Mejoramiento de Maíz y Trigo Dirección: CIMMYT Int. c/o Jesper Nørgaard Km. 45, Carretera México-Veracruz El Batán Texcoco, Edo. de México CP 56130 MEXICO Tel.: +52 (55) 58-04-20-04 ext. 1374 Fax: +52 (55) 58-04-75-58 Tel. Casa: 53-10-05-95 ó 53-10-97-78 Download the free program World Time Explorer, I made: http://www.cimmyt.org/timezone/index.htm

2 1

First Sunday
by Jesper Norgaard Welen Feb. 22, 2006

Feb. 22, 2006

There are a few instances in the tz database where a rule is using the string "Sun<=7" rather than "Sun>=1" to mean the first Sunday in the month. I would recommend using "Sun>=1", because I think it is clearer. At least I confused "Sun<=7" with "Sun>=7" once when reading these rules, which is quite a different thing. In the africa file: Rule Algeria 1916 1919 - Oct Sun<=7 23:00s 0 - In the antarctica file: Rule ArgAQ 1967 1968 - Oct Sun<=7 0:00 1:00 S Rule ArgAQ 1968 1969 - Apr Sun<=7 0:00 0 - ... Rule ArgAQ 1974 1976 - Oct Sun<=7 0:00 1:00 S Rule ArgAQ 1975 1977 - Apr Sun<=7 0:00 0 - Regards, - Jesper Jesper Nørgaard Welen Email: jnorgard(a)Prodigy.Net.mx Project Leader (Líder de Proyecto) Software CIMMYT - Centro Internacional de Mejoramiento de Maíz y Trigo Dirección: CIMMYT Int. c/o Jesper Nørgaard Km. 45, Carretera México-Veracruz El Batán Texcoco, Edo. de México CP 56130 MEXICO Tel.: +52 (55) 58-04-20-04 ext. 1374 Fax: +52 (55) 58-04-75-58 Tel. Casa: 53-10-05-95 ó 53-10-97-78 Download the free program World Time Explorer, I made: http://www.cimmyt.org/timezone/index.htm

2 1

tz changes to pacify GCC 4.0.2 on Solaris 8
by Paul Eggert Feb. 22, 2006

Feb. 22, 2006

When compiling the latest tz code with GCC 4.0.2 on Solaris 8 (64-bit), I got some warnings that looked like this: zic.c:979: warning: comparison between signed and unsigned zic.c:1086: warning: comparison between signed and unsigned asctime.c:106: warning: '0' flag ignored with precision and '%d' printf format Here are proposed patches to remove all the places where I observed problems in this area. Not all the size_t-related problems were caught by GCC, but I expect other compilers might have similar complaints, so I figured I might as well fix them all while I was thinking about it. Since the affected quantities are all too small to be worried about overflowing a signed 16-bit quantity on practical machines, I don't think this patch fixes any real-world bugs or changes any real-world behavior, except perhaps that it makes the code a bit smaller and faster on 64-bit hosts because we don't keep converting 64-bit quantities to 32 bits and back. =================================================================== RCS file: RCS/private.h,v retrieving revision 2006.2.0.1 retrieving revision 2006.2.0.2 diff -pu -r2006.2.0.1 -r2006.2.0.2 --- private.h 2006/02/21 21:07:30 2006.2.0.1 +++ private.h 2006/02/22 00:24:05 2006.2.0.2 @@ -236,8 +236,8 @@ extern char * asctime_r(); char * icalloc P((int nelem, int elsize)); char * icatalloc P((char * old, const char * new)); char * icpyalloc P((const char * string)); -char * imalloc P((int n)); -void * irealloc P((void * pointer, int size)); +char * imalloc P((size_t n)); +void * irealloc P((void * pointer, size_t size)); void icfree P((char * pointer)); void ifree P((char * pointer)); const char * scheck P((const char * string, const char * format)); =================================================================== RCS file: RCS/ialloc.c,v retrieving revision 1997.3 retrieving revision 1997.3.0.1 diff -pu -r1997.3 -r1997.3.0.1 --- ialloc.c 1997/03/07 21:06:34 1997.3 +++ ialloc.c 2006/02/22 00:27:00 1997.3.0.1 @@ -12,9 +12,9 @@ static char elsieid[] = "@(#)ialloc.c 8. char * imalloc(n) -const int n; +const size_t n; { - return malloc((size_t) nonzero(n)); + return malloc(nonzero(n)); } char * @@ -30,11 +30,11 @@ int elsize; void * irealloc(pointer, size) void * const pointer; -const int size; +const size_t size; { if (pointer == NULL) return imalloc(size); - return realloc((void *) pointer, (size_t) nonzero(size)); + return realloc((void *) pointer, nonzero(size)); } char * @@ -43,7 +43,7 @@ char * const old; const char * const new; { register char * result; - register int oldsize, newsize; + register size_t oldsize, newsize; newsize = (new == NULL) ? 0 : strlen(new); if (old == NULL) =================================================================== RCS file: RCS/localtime.c,v retrieving revision 2006.2.0.1 retrieving revision 2006.2.0.2 diff -pu -r2006.2.0.1 -r2006.2.0.2 --- localtime.c 2006/02/21 21:07:30 2006.2.0.1 +++ localtime.c 2006/02/22 00:24:05 2006.2.0.2 @@ -340,8 +340,9 @@ register const int doextend; register const char * p; register int i; register int fid; - register int stored; - register int nread; + register int read_return; + register size_t stored; + register size_t nread; union { struct tzhead tzhead; char buf[2 * sizeof(struct tzhead) + @@ -385,12 +386,14 @@ register const int doextend; if ((fid = open(name, OPEN_MODE)) == -1) return -1; } - nread = read(fid, u.buf, sizeof u.buf); - if (close(fid) < 0 || nread <= 0) + read_return = read(fid, u.buf, sizeof u.buf); + if (close(fid) < 0 || read_return <= 0) return -1; + nread = read_return; for (stored = 4; stored <= 8; stored *= 2) { int ttisstdcnt; int ttisgmtcnt; + size_t j; ttisstdcnt = (int) detzcode(u.tzhead.tzh_ttisstdcnt); ttisgmtcnt = (int) detzcode(u.tzhead.tzh_ttisgmtcnt); @@ -511,12 +514,12 @@ register const int doextend; if (u.tzhead.tzh_version[0] == '\0') break; nread -= p - u.buf; - for (i = 0; i < nread; ++i) - u.buf[i] = p[i]; + for (j = 0; j < nread; ++j) + u.buf[j] = p[j]; /* ** If this is a narrow integer time_t system, we're done. */ - if (stored >= (int) sizeof(time_t) && TYPE_INTEGRAL(time_t)) + if (stored >= sizeof(time_t) && TYPE_INTEGRAL(time_t)) break; } if (doextend && nread > 2 && =================================================================== RCS file: RCS/scheck.c,v retrieving revision 2006.2 retrieving revision 2006.2.0.1 diff -pu -r2006.2 -r2006.2.0.1 --- scheck.c 2006/02/20 15:08:17 2006.2 +++ scheck.c 2006/02/22 00:24:05 2006.2.0.1 @@ -23,7 +23,7 @@ const char * const format; result = ""; if (string == NULL || format == NULL) return result; - fbuf = imalloc((int) (2 * strlen(format) + 4)); + fbuf = imalloc(2 * strlen(format) + 4); if (fbuf == NULL) return result; fp = format; =================================================================== RCS file: RCS/zic.c,v retrieving revision 2006.2 retrieving revision 2006.2.0.1 diff -pu -r2006.2 -r2006.2.0.1 --- zic.c 2006/02/20 15:08:17 2006.2 +++ zic.c 2006/02/22 00:24:05 2006.2.0.1 @@ -167,8 +167,8 @@ static int leapseen; static int leapminyear; static int leapmaxyear; static int linenum; -static int max_abbrvar_len; -static int max_format_len; +static size_t max_abbrvar_len; +static size_t max_format_len; static zic_t max_time; static int max_year; static zic_t min_time; @@ -979,7 +979,7 @@ const int nfields; if (max_abbrvar_len < strlen(r.r_abbrvar)) max_abbrvar_len = strlen(r.r_abbrvar); rules = (struct rule *) (void *) erealloc((char *) rules, - (int) ((nrules + 1) * sizeof *rules)); + (nrules + 1) * sizeof *rules); rules[nrules++] = r; } @@ -996,7 +996,7 @@ const int nfields; return FALSE; } if (strcmp(fields[ZF_NAME], TZDEFAULT) == 0 && lcltime != NULL) { - buf = erealloc(buf, (int) (132 + strlen(TZDEFAULT))); + buf = erealloc(buf, 132 + strlen(TZDEFAULT)); (void) sprintf(buf, _("\"Zone %s\" line and -l option are mutually exclusive"), TZDEFAULT); @@ -1004,7 +1004,7 @@ _("\"Zone %s\" line and -l option are mu return FALSE; } if (strcmp(fields[ZF_NAME], TZDEFRULES) == 0 && psxrules != NULL) { - buf = erealloc(buf, (int) (132 + strlen(TZDEFRULES))); + buf = erealloc(buf, 132 + strlen(TZDEFRULES)); (void) sprintf(buf, _("\"Zone %s\" line and -p option are mutually exclusive"), TZDEFRULES); @@ -1014,7 +1014,7 @@ _("\"Zone %s\" line and -p option are mu for (i = 0; i < nzones; ++i) if (zones[i].z_name != NULL && strcmp(zones[i].z_name, fields[ZF_NAME]) == 0) { - buf = erealloc(buf, (int) (132 + + buf = erealloc(buf, (132 + strlen(fields[ZF_NAME]) + strlen(zones[i].z_filename))); (void) sprintf(buf, @@ -1112,7 +1112,7 @@ const int iscont; } } zones = (struct zone *) (void *) erealloc((char *) zones, - (int) ((nzones + 1) * sizeof *zones)); + (nzones + 1) * sizeof *zones); zones[nzones++] = z; /* ** If there was an UNTIL field on this line, @@ -1249,7 +1249,7 @@ const int nfields; l.l_from = ecpyalloc(fields[LF_FROM]); l.l_to = ecpyalloc(fields[LF_TO]); links = (struct link *) (void *) erealloc((char *) links, - (int) ((nlinks + 1) * sizeof *links)); + (nlinks + 1) * sizeof *links); links[nlinks++] = l; } @@ -1557,7 +1557,7 @@ const char * const string; ++leapi32; } fullname = erealloc(fullname, - (int) (strlen(directory) + 1 + strlen(name) + 1)); + strlen(directory) + 1 + strlen(name) + 1); (void) sprintf(fullname, "%s/%s", directory, name); /* ** Remove old file, if any, to snap links. @@ -1942,8 +1942,8 @@ const int zonecount; register char * startbuf; register char * ab; register char * envvar; - register int max_abbr_len; - register int max_envvar_len; + register size_t max_abbr_len; + register size_t max_envvar_len; max_abbr_len = 2 + max_format_len + max_abbrvar_len; max_envvar_len = 2 * max_abbr_len + 5 * 9; @@ -2310,7 +2310,7 @@ const char * const type; if (type == NULL || *type == '\0') return TRUE; - buf = erealloc(buf, (int) (132 + strlen(yitcommand) + strlen(type))); + buf = erealloc(buf, 132 + strlen(yitcommand) + strlen(type)); (void) sprintf(buf, "%s %d %s", yitcommand, year, type); result = system(buf); if (WIFEXITED(result)) switch (WEXITSTATUS(result)) { @@ -2401,7 +2401,7 @@ register char * cp; if (cp == NULL) return NULL; array = (char **) (void *) - emalloc((int) ((strlen(cp) + 1) * sizeof *array)); + emalloc((strlen(cp) + 1) * sizeof *array); nsubs = 0; for ( ; ; ) { while (isascii((unsigned char) *cp) && =================================================================== RCS file: RCS/asctime.c,v retrieving revision 2006.2 retrieving revision 2006.2.0.1 diff -pu -r2006.2 -r2006.2.0.1 --- asctime.c 2006/02/20 15:08:17 2006.2 +++ asctime.c 2006/02/22 00:24:05 2006.2.0.1 @@ -37,14 +37,24 @@ static char elsieid[] = "@(#)asctime.c 8 ** The ISO C 1999 and POSIX 1003.1-2004 standards prohibit padding the year, ** but many implementations pad anyway; most likely the standards are buggy. */ +#ifdef __GNUC__ +/* Avoid "warning: '0' flag ignored with precision and '%d' printf format". */ +#define ASCTIME_FMT "%.3s %.3s%3d %2.2d:%2.2d:%2.2d %-4s\n" +#else #define ASCTIME_FMT "%.3s %.3s%3d %02.2d:%02.2d:%02.2d %-4s\n" +#endif /* ** For years that are more than four digits we put extra spaces before the year ** so that code trying to overwrite the newline won't end up overwriting ** a digit within a year and truncating the year (operating on the assumption ** that no output is better than wrong output). */ +#ifdef __GNUC__ +/* Avoid "warning: '0' flag ignored with precision and '%d' printf format". */ +#define ASCTIME_FMT_B "%.3s %.3s%3d %2.2d:%2.2d:%2.2d %s\n" +#else #define ASCTIME_FMT_B "%.3s %.3s%3d %02.2d:%02.2d:%02.2d %s\n" +#endif #define STD_ASCTIME_BUF_SIZE 26 /*

1 0

Re: forego
by Mccabe Alisa Feb. 21, 2006

Feb. 21, 2006

testament Mr. Dick represented the Government or the Opposition as the case daughters station in life, the projects I may contemplate for her replicate such another position something else, entirely different; the mouth a little at the same time - and produced my last letter to

1 0

RE: Tzdata2006b.tar.gz, tz32code2006b.tar.gz, and tz64code2006b.tar.gz
by Olson, Arthur David (NIH/NCI) [E] Feb. 21, 2006

Feb. 21, 2006

If we don't see reports of problems with the new stuff, I'll be happy to go with tzcode2006c.tar.gz rather than tz64code2006c.tar.gz on the next iteration; I wasn't inclined to be so prejudicial this time around. The question then becomes whether to change the name tz32code206b.tar.gz. Perhaps old-tzcode-32-bit-outputx.tar.gz would do the trick? --ado -----Original Message----- From: Paul Eggert [mailto:eggert@cs.ucla.edu] Sent: Tuesday, February 21, 2006 4:06 PM To: tz(a)lecserver.nci.nih.gov Subject: Re: Tzdata2006b.tar.gz, tz32code2006b.tar.gz, and tz64code2006b.tar.gz "Olson, Arthur David $NIH/NCI$ [E]" <olsona(a)dc37a.nci.nih.gov> writes: > ftp://elsie.nci.nih.gov/pub/tzdata2006b.tar.gz > ftp://elsie.nci.nih.gov/pub/tz32code2006b.tar.gz > ftp://elsie.nci.nih.gov/pub/tz64code2006b.tar.gz > > The 32-bit code is there for folks who may need it; with providence, > there will be little or no need to update it. This means we'll need to update the instructions in tz-link.htm, which currently say this: wget 'ftp://elsie.nci.nih.gov/pub/tz*.tar.gz' gzip -dc tzcode*.tar.gz | tar -xf - gzip -dc tzdata*.tar.gz | tar -xf - and this won't work nicely, due to the two code tarballs overridding each other. The names tz32code2006b.tar.gz and tz64code2006b.tar.gz are a bit confusing, since people might naturally get confused into thinking that the tarballs are for 32-bit and 64-bit development platforms, or for 32-bit and 64-bit runtime platforms, whereas neither is the case: they are really for 32-bit and 64-bit tz files, which is a different matter. Here's an idea. How about if we rename tz32code2006b.tar.gz to pre2038-tzcode2006b.tar.gz, and rename tz64code2006b.tar.gz back to tzcode2006b.tar.gz? That way, I won't have to change tz-link.htm (so it's less work for me :-), and also, users will be able to see more clearly why one might prefer the current to the previous version. Also, this naming convention fits in better with the recently-removed 'classictzcode' and 'classictzdata' convention. If you still prefer the current naming style, please let me know, so that I can propose a patch to tz-link.htm. Thanks.

1 0

Re: Message to `tz' mailing list
by Robert Elz Feb. 19, 2006

Feb. 19, 2006

Date: Fri, 17 Feb 2006 15:14:09 +0530 From: "N. Raghavendra" <raghu(a)mri.ernet.in> Message-ID: <867j7ugw0m.fsf(a)riemann.mri.ernet.in> | Perhaps I did not express my question clearly. I am trying to | understand how, given | | 1. an integer TIME (viewed as a number of seconds since the Epoch); | and | | 2. a TZFILE, | | one computes the wall clock time corresponding to TIME, in the time | zone defined by TZFILE. The wall clock time should be returned as a | broken-down time value, i.e., as a list of the form | | (SEC, MIN, HOUR, MDAY, MON, YEAR, WDAY, IS_DST, TZ_ABBREV, GMT_OFF) That's exactly what localtime() does isn't it? All the code is available (and free to use) - just use it (either the learn from, or simply to compile into your application). The raw tzfile isn't used of course, it is converted (using zic) into a binary format - but if you needed to you could combine the roles of zic and localtime into one function. kre

2 3

RE: Message to `tz' mailing list
by Olson, Arthur David (NIH/NCI) [E] Feb. 17, 2006

Feb. 17, 2006

> I am confused about the use of the UTC/local indicators and the standard/wall indicators in a tzfile. According to tzfile(5), they "are used when > a time zone file is used in handling POSIX-style time zone environment variables". Does this mean that these indicators are irrelevant when computing > local time using only a tzfile (without looking at the environment). That's correct: the ttisgmt and ttisstd indicators can be ignored if you're simply trying to figure out what wall clock time is for the zone that's associated with the file. (The indicators are there for the case when the time zone file is used as a "template" for the DST rules to be used when handling POSIX-style TZ environment variables; you need to know whether the transition time is defined in terms of UTC, local wall clock time, or local standard time to be able to know when translating the transition time for use in another zone.) --ado

3 2

Proposed tz code cleanups
by Arthur David Olson Feb. 16, 2006

Feb. 16, 2006

Along with the 64-bit changes previously circulated on the time zone list and due to show up in the ftp version on Monday, here are proposed STDC, long, and overflow cleanups from Paul Eggert and Ken Pizzini. diff -r -c old/date.c new/date.c *** old/date.c Thu Jan 26 10:19:21 2006 --- new/date.c Thu Feb 16 11:05:15 2006 *************** *** 1,6 **** #ifndef lint #ifndef NOID ! static char elsieid[] = "@(#)date.c 7.42"; /* ** Modified from the UCB version with the SCCS ID appearing below. */ --- 1,6 ---- #ifndef lint #ifndef NOID ! static char elsieid[] = "@(#)date.c 7.44"; /* ** Modified from the UCB version with the SCCS ID appearing below. */ *************** *** 254,260 **** struct timeval tv; tv.tv_sec = (int) adjust; ! tv.tv_usec = (int) ((adjust - tv.tv_sec) * 1000000); if (adjtime(&tv, (struct timeval *) NULL) != 0) oops("adjtime"); #endif /* HAVE_ADJTIME */ --- 254,260 ---- struct timeval tv; tv.tv_sec = (int) adjust; ! tv.tv_usec = (int) ((adjust - tv.tv_sec) * 1000000L); if (adjtime(&tv, (struct timeval *) NULL) != 0) oops("adjtime"); #endif /* HAVE_ADJTIME */ *************** *** 352,360 **** --- 352,364 ---- /*ARGSUSED*/ static void + #ifdef __STDC__ + reset(const time_t newt, const int nflag) + #else /* !defined __STDC__ */ reset(newt, nflag) const time_t newt; const int nflag; + #endif /* !defined __STDC__ */ { register int fid; time_t oldt; *************** *** 619,628 **** --- 623,636 ---- #define ATOI2(ar) (ar[0] - '0') * 10 + (ar[1] - '0'); ar += 2; static time_t + #ifdef __STDC__ + convert(register const char * const value, const int dousg, const time_t t) + #else /* !defined __STDC__ */ convert(value, dousg, t) register const char * const value; const int dousg; const time_t t; + #endif /* !defined __STDC__ */ { register const char * cp; register const char * dotp; *************** *** 726,736 **** --- 734,751 ---- */ static void + #ifdef __STDC__ + checkfinal(const char * const value, + const int didusg, + const time_t t, + const time_t oldnow) + #else /* !defined __STDC__ */ checkfinal(value, didusg, t, oldnow) const char * const value; const int didusg; const time_t t; const time_t oldnow; + #endif /* !defined __STDC__ */ { time_t othert; struct tm tm; *************** *** 789,799 **** --- 804,819 ---- } static void + #ifdef __STDC__ + iffy(const time_t thist, const time_t thatt, + const char * const value, const char * const reason) + #else /* !defined __STDC__ */ iffy(thist, thatt, value, reason) const time_t thist; const time_t thatt; const char * const value; const char * const reason; + #endif /* !defined __STDC__ */ { struct tm tm; diff -r -c old/localtime.c new/localtime.c *** old/localtime.c Thu Jan 26 10:19:21 2006 --- new/localtime.c Thu Feb 16 11:07:47 2006 *************** *** 5,11 **** #ifndef lint #ifndef NOID ! static char elsieid[] = "@(#)localtime.c 7.97"; #endif /* !defined NOID */ #endif /* !defined lint */ --- 5,11 ---- #ifndef lint #ifndef NOID ! static char elsieid[] = "@(#)localtime.c 7.98"; #endif /* !defined NOID */ #endif /* !defined lint */ *************** *** 510,518 **** --- 510,522 ---- */ static const char * + #ifdef __STDC__ + getqzname(register const char *strp, const char delim) + #else /* !defined __STDC__ */ getqzname(strp, delim) register const char * strp; const char delim; + #endif /* !defined __STDC__ */ { register char c; diff -r -c old/zdump.c new/zdump.c *** old/zdump.c Thu Jan 26 10:19:21 2006 --- new/zdump.c Thu Feb 16 11:16:58 2006 *************** *** 1,4 **** ! static char elsieid[] = "@(#)zdump.c 7.74"; /* ** This code has been made independent of the rest of the time --- 1,4 ---- ! static char elsieid[] = "@(#)zdump.c 7.76"; /* ** This code has been made independent of the rest of the time *************** *** 15,21 **** #include "ctype.h" /* for isalpha et al. */ #ifndef isascii #define isascii(x) 1 ! #endif #ifndef ZDUMP_LO_YEAR #define ZDUMP_LO_YEAR (-500) --- 15,21 ---- #include "ctype.h" /* for isalpha et al. */ #ifndef isascii #define isascii(x) 1 ! #endif /* !defined isascii */ #ifndef ZDUMP_LO_YEAR #define ZDUMP_LO_YEAR (-500) *************** *** 418,431 **** } } else if (0 > (time_t) -1) { /* ! ** time_t is signed. */ ! register time_t hibit; ! for (hibit = 1; (hibit * 2) != 0; hibit *= 2) ! continue; ! absolute_min_time = hibit; ! absolute_max_time = -(hibit + 1); } else { /* ** time_t is unsigned. --- 418,438 ---- } } else if (0 > (time_t) -1) { /* ! ** time_t is signed. Assume overflow wraps around. */ ! time_t t = 0; ! time_t t1 = 1; ! while (t < t1) { ! t = t1; ! t1 = 2 * t1 + 1; ! } ! ! absolute_max_time = t; ! t = -t; ! absolute_min_time = t - 1; ! if (t < absolute_min_time) ! absolute_min_time = t; } else { /* ** time_t is unsigned. *************** *** 468,477 **** --- 475,488 ---- } static time_t + #ifdef __STDC__ + hunt(char *name, time_t lot, time_t hit) + #else /* !defined __STDC__ */ hunt(name, lot, hit) char * name; time_t lot; time_t hit; + #endif /* !defined __STDC__ */ { time_t t; long diff; *************** *** 541,550 **** --- 552,565 ---- } static void + #ifdef __STDC__ + show(char *zone, time_t t, int v) + #else /* !defined __STDC__ */ show(zone, t, v) char * zone; time_t t; int v; + #endif /* !defined __STDC__ */ { register struct tm * tmp;

1 0

FW: Message to `tz' mailing list
by Olson, Arthur David (NIH/NCI) [E] Feb. 16, 2006

Feb. 16, 2006

N. Raghavendra has been added to the time zone mailing list. --ado -----Original Message----- From: N. Raghavendra [mailto:raghu@mri.ernet.in] Sent: Wednesday, February 15, 2006 11:44 AM To: Olson, Arthur David (NIH/NCI) [E] Subject: Message to `tz' mailing list ... UTC and std indicators: I am trying to write a Common Lisp program to compute local time using the `tz' database. I am confused about the use of the UTC/local indicators and the standard/wall indicators in a tzfile. According to tzfile(5), they "are used when a time zone file is used in handling POSIX-style time zone environment variables". Does this mean that these indicators are irrelevant when computing local time using only a tzfile (without looking at the environment). I have tried reading `localtime.c' using my negligible knowledge of C, but haven't understood the role of the indicators. Any pointers on this would be most welcome. TIA, Raghavendra. -- N. Raghavendra <raghu(a)mri.ernet.in> | See message headers for contact Harish-Chandra Research Institute | and OpenPGP details.

1 0

RE: America/Edmonton 2007
by Paul Schauble Feb. 11, 2006

Feb. 11, 2006

In cases like this I really have to wonder about the custom of using the largest city. To me, it makes much more sense to name the entity to which the change applies. Not just whatever city happens to be there. Naming the province, county, state, or whatever terms the changes applies to conveys more information. In this case, I think that should be America/Alberta or America/Canada/Alberta. In some the recent Indiana things, I think the tz name should have been America/Indiana/county. And I still think it should be America/Navajo, not America/Shiprock. Thanks for your attention, ++PLS -----Original Message----- From: tz-request(a)elsie.nci.nih.gov [mailto:tz-request@elsie.nci.nih.gov] On Behalf Of Chris Walton-TM Sent: Tuesday, February 07, 2006 2:53 PM To: tz(a)lecserver.nci.nih.gov Subject: FW: America/Edmonton 2007 -----Original Message----- From: Chris Walton-TM Sent: February 7, 2006 1:42 AM To: tz(a)lecserver.nci.nih.gov Subject: America/Edmonton 2007 The Alberta government announced on Feb 2 that it will extend daylight savings by four weeks starting in 2007. This will match the changes that have been made in the United States. I am hoping that "America/Edmonton" can be updated in the next tzdata release. Here is a link from the CBC website. http://www.cbc.ca/calgary/story/ca_daylight20060202.html Thanks. -chris

7 10

RE: Pulaski County rebels against the Department of Transportation
by Olson, Arthur David (NIH/NCI) [E] Feb. 9, 2006

Feb. 9, 2006

We're faced with the ugly possibility that clocks will be set two different ways in Pulaski County; offices under federal control (such as post offices) may follow U. S. Department of Transportation regulations. Particularly touchy: any federal court offices located in Pulaski County. --ado -----Original Message----- From: Paul Eggert [mailto:eggert@cs.ucla.edu] Sent: Wednesday, February 08, 2006 4:15 AM To: tz(a)lecserver.nci.nih.gov Subject: Pulaski County rebels against the Department of Transportation According to an AP article <http://www.indystar.com/apps/pbcs.dll/article?AID=/20060207/NEWS01/6020 70408/1006>, officials in Pulaski County, Indiana, have changed their minds, and want to undo the time-zone change that they asked the feds to grant a few months ago. "While it's official, it's unofficial: We will not set our clocks with Chicago; we'll set our clocks with Indianapolis," said Pulaski County Council President Sam Frain. Even if they follow through with this threat, we won't have to create a new zone; Pulaski County residents can continue to use the America/Indiana/Indianapolis zone. However, we should adjust zone.tab and the comments to make it clear that Pulaski County is still using Indianapolis time.

1 0

RE: America/Edmonton 2007
by Olson, Arthur David (NIH/NCI) [E] Feb. 9, 2006

Feb. 9, 2006

One piece of evidence from the current Indiana brouhaha: > According to an AP article > <http://www.indystar.com/apps/pbcs.dll/article?AID=/20060207/NEWS01/6020 70408/1006>, > officials in Pulaski County, Indiana, have changed their minds, and want to undo the time-zone change that they asked the feds to grant a few months > ago. "While it's official, it's unofficial: We will not set our clocks with Chicago; we'll set our clocks with Indianapolis," > said Pulaski County Council President Sam Frain. Citizen Frain uses cities (rather than counties or states) to describe what Pulaski is to do. --ado

1 0

RE: America/Edmonton and other zone naming issues
by Paul Schauble Feb. 8, 2006

Feb. 8, 2006

We actually might be able to do that in the US. As far as I know, all of these zone changes are done by country. The only exceptions I know of in the states where incorporated cities are more or less equivalent to counties. So we might have America/Indiana/Pulaski Pulaski County America/Arizona America/Navajo Navajo reservation, or America/Dineh If we made links for America/<State>/<Jurisdiction> to whatever was right, we would cover 99.9% of these changes without people needing to change the name they use. <Jurisdiction> is usually county, but could be an incorporated city or an indian nation, or a few other things. I think the real need, though, is to tie the zone name to the jurisdiction making the change, and not to pick an arbitrary city as a proxy for the deciding jurisdiction. Where an entire state follows the same rules, don't create the county level. Just for reference, the US has 3,007 entities named "County" 16 Boroughs in Alaska 11 Census Areas in Alaska (for areas not organized into Boroughs by the State) 64 Parishes in Louisiana 42 Independent Cities (1 in Maryland, 1 in Missouri, 1 in Nevada, and the remainder in Virginia) 1 District - the Federal District or District of Columbia. -----Original Message----- From: tz-request(a)elsie.nci.nih.gov [mailto:tz-request@elsie.nci.nih.gov] On Behalf Of Mark Davis Sent: Wednesday, February 08, 2006 8:23 AM To: tz(a)lecserver.nci.nih.gov Cc: tz(a)lecserver.nci.nih.gov Subject: Re: America/Edmonton and other zone naming issues What would be even better would be physical boundaries for each given zone. Right now, within multi-zone countries, the timezone database alone is insufficient to determine the behavior of any given city (or other location). Mark

2 1

Pulaski County rebels against the Department of Transportation
by Paul Eggert Feb. 8, 2006

Feb. 8, 2006

According to an AP article <http://www.indystar.com/apps/pbcs.dll/article?AID=/20060207/NEWS01/60207040…>, officials in Pulaski County, Indiana, have changed their minds, and want to undo the time-zone change that they asked the feds to grant a few months ago. "While it's official, it's unofficial: We will not set our clocks with Chicago; we'll set our clocks with Indianapolis," said Pulaski County Council President Sam Frain. Even if they follow through with this threat, we won't have to create a new zone; Pulaski County residents can continue to use the America/Indiana/Indianapolis zone. However, we should adjust zone.tab and the comments to make it clear that Pulaski County is still using Indianapolis time.

1 0

FW: America/Edmonton 2007
by Chris Walton-TM Feb. 8, 2006

Feb. 8, 2006

-----Original Message----- From: Chris Walton-TM Sent: February 7, 2006 1:42 AM To: tz(a)lecserver.nci.nih.gov Subject: America/Edmonton 2007 The Alberta government announced on Feb 2 that it will extend daylight savings by four weeks starting in 2007. This will match the changes that have been made in the United States. I am hoping that "America/Edmonton" can be updated in the next tzdata release. Here is a link from the CBC website. http://www.cbc.ca/calgary/story/ca_daylight20060202.html Thanks. -chris

2 1

Re: FW: tzcode2006a.tar.gz on Irix 6.5.18m
by Paul Eggert Feb. 8, 2006

Feb. 8, 2006

Andrew Donaldson <A.Donaldson(a)bom.gov.au> writes: > Yep, that fixes it. Thanks for checking it. > Do you plan to include this in the next version? Yes, I'll submit it as part of my next proposed patch; Arthur David Olson has the final say.

1 0

Re: FW: tzcode2006a.tar.gz on Irix 6.5.18m
by Paul Eggert Feb. 8, 2006

Feb. 8, 2006

Thanks for looking into the problem. zdump.c relies on undefined behavior on overflow, which is the problem here, but if we can assume C99+LIA-1 semantics (the common case) we can get a reliable result. Also, I think it's safer to avoid 'register'. Does the following patch to zdump.c fix things for you? --- zdump.c 2006/02/02 20:42:22 2005.18.0.1 +++ zdump.c 2006/02/08 00:33:28 @@ -418,14 +418,21 @@ _("%s: use of -v on system with floating } } else if (0 > (time_t) -1) { /* - ** time_t is signed. + ** time_t is signed. Assume overflow wraps around. */ - register time_t hibit; + time_t t = 0; + time_t t1 = 1; - for (hibit = 1; (hibit * 2) != 0; hibit *= 2) - continue; - absolute_min_time = hibit; - absolute_max_time = -(hibit + 1); + while (t < t1) { + t = t1; + t1 = 2 * t1 + 1; + } + + absolute_max_time = t; + t = -t; + absolute_min_time = t - 1; + if (t < absolute_min_time) + absolute_min_time = t; } else { /* ** time_t is unsigned.

1 0

RE: OT- trouble connecting to elsie-
by Olson, Arthur David (NIH/NCI) [E] Feb. 7, 2006

Feb. 7, 2006

I've just update the TCP wrapper in user here at elsie; give ftp a try and let me know one way or the other. Thanks! --ado -----Original Message----- From: Dave Cantor [mailto:Dave@Cantor.mv.com] Sent: Monday, February 06, 2006 8:26 AM To: Paul Eggert Cc: tz(a)lecserver.nci.nih.gov Subject: Re: OT- trouble connecting to elsie- On 5 Feb 2006 at 22:59, Paul Eggert wrote: > > Date: Fri, 03 Feb 2006 09:51:19 -0500 > > From: "Dave Cantor" <Dave(a)Cantor.mv.com> > > Message-ID: <43E32797.7719.2F8A382(a)Dave.Cantor.mv.com> > > | I have not been able to connect to elsie.nci.nih.gov via ftp, or, > > | rather, whenever I connect, the remote host disconnects. Is > > | anyone else having this trouble? > In the past this has happened when people attempt to connect to elsie > from blacklisted IP addresses. I don't know who's in charge of the > blacklist, unfortunately. It could be something as simple as the > reverse-DNS entry for your IP address; is it set up properly? As far as I can tell, the reverse-DNS entry is correct: host174-106.defiant.tvcconnect.net, IP address 216.255.174.106, Maybe it's blacklisted for some reason. Thanks, Dave C.

2 1

FW: Bug#351049: libc6: America/Havana DST info wrong for 2005/2006
by Olson, Arthur David (NIH/NCI) [E] Feb. 6, 2006

Feb. 6, 2006

Lionel Elie Mamane is not on the time zone mailing list; direct replies appropriately. --ado -----Original Message----- From: Lionel Elie Mamane [mailto:lionel@mamane.lu] Sent: Thursday, February 02, 2006 9:18 AM To: Debian Bug Tracking System Subject: Bug#351049: libc6: America/Havana DST info wrong for 2005/2006 Package: libc6 Version: 2.3.5-12.1 Severity: normal Cuba did not switch back to non-DST time on October 2005, but the timezone info in libc thinks it did: Rule Cuba 2005 max - Oct lastSun 0:00s 0 S That line should probably be changed to Rule Cuba 2006 max - Oct lastSun 0:00s 0 S Source: http://www.timeanddate.com/worldclock/timezone.html?n=99 (and a friend that spent vacation in Cuba) Note that the timezone/northamerica says: # From Evert van der Veer via Steffen Thorsen (2004-10-28): # Cuba is not going back to standard time this year. # From Paul Eggert (2004-10-28): # http://www.granma.cu/ingles/2004/septiembre/juev30/41medid-i.html # says that it's due to a problem at the Antonio Guiteras # thermoelectric plant, and says "This October there will be no return # to normal hours (after daylight saving time)". # For now, let's assume that it's a one-year temporary measure. ^^^^^^^^^^^^^^^^^^^^^^^^^ So it became a two-year temporary measure? We are not sure if it won't become a 3+ year temporary measure. -- System Information: Debian Release: testing/unstable APT prefers unstable APT policy: (500, 'unstable'), (1, 'experimental') Architecture: i386 (x86_64) Shell: /bin/sh linked to /bin/bash Kernel: Linux 2.6.15-deb1-64bit Locale: LANG=fr_LU.UTF-8, LC_CTYPE=fr_LU.UTF-8 (charmap=UTF-8) -- no debconf information

5 11

RE: Misspelling With the Australia/Lord_Howe Time Zone
by Olson, Arthur David (NIH/NCI) [E] Feb. 6, 2006

Feb. 6, 2006

I failed to find a "load" problem here at elsie (see below). (I also tried looking in the SCCS record to no avail.) --ado Script started on Mon 06 Feb 2006 10:17:36 AM EST lecserver$ ls -l total 0 -rw-r--r-- 1 olsona root 0 Feb 6 10:17 typescript lecserver$ ftp elsie.nci.nih.gov Connected to lecserver. 220 lecserver FTP server (Version wu-2.6.2(4)+connect-dos.patch+realpath.patch Tue Apr 13 15:43:02 EDT 2004) ready. Name (elsie.nci.nih.gov:olsona): anonymous 331 Guest login ok, send your complete e-mail address as password. Password: 230 Guest login ok, access restrictions apply. ftp> binary 200 Type set to I. ftp> prompt Interactive mode off. ftp> cd pub 250 CWD command successful. ftp> get tzcode2006a.tar.gz 200 PORT command successful. 150 Opening BINARY mode data connection for tzcode2006a.tar.gz (180636 bytes). 226 Transfer complete. local: tzcode2006a.tar.gz remote: tzcode2006a.tar.gz 180636 bytes received in 0.031 seconds (5761.77 Kbytes/s) ftp> get tzdata2006a.tar.gz 200 PORT command successful. 150 Opening BINARY mode data connection for tzdata2006a.tar.gz (149612 bytes). 226 Transfer complete. local: tzdata2006a.tar.gz remote: tzdata2006a.tar.gz 149612 bytes received in 0.058 seconds (2507.86 Kbytes/s) ftp> quit 221-You have transferred 330248 bytes in 2 files. 221-Total traffic for this session was 331005 bytes in 2 transfers. 221-Thank you for using the FTP service on lecserver. 221 Goodbye. lecserver$ gunzip < *code* | tar xfp - lecserver$ gunzip < *data* | tar xfp - lecserver$ ls -l total 3152 -r--r--r-- 1 olsona root 20768 Jan 26 10:19 africa -r--r--r-- 1 olsona root 12028 Jan 26 10:19 antarctica -r--r--r-- 1 olsona root 4094 Jan 26 10:19 asctime.c -r--r--r-- 1 olsona root 65779 Jan 30 08:59 asia -r--r--r-- 1 olsona root 55475 Jan 26 10:19 australasia -r--r--r-- 1 olsona root 3678 Jan 26 10:19 backward -r--r--r-- 1 olsona root 3417 Jan 26 10:19 checktab.awk -r--r--r-- 1 olsona root 5065 Jan 26 10:19 date.1 -rw-r--r-- 1 olsona root 6151 Jan 30 08:59 date.1.txt -r--r--r-- 1 olsona root 22986 Jan 26 10:19 date.c -r--r--r-- 1 olsona root 1708 Jan 26 10:19 difftime.c -r--r--r-- 1 olsona root 2921 Jan 26 10:19 etcetera -r--r--r-- 1 olsona root 104518 Jan 26 10:19 europe -r--r--r-- 1 olsona root 311 Jan 26 10:19 factory -r--r--r-- 1 olsona root 1273 Jan 26 10:19 ialloc.c -r--r--r-- 1 olsona root 4008 Jan 26 10:19 iso3166.tab -r--r--r-- 1 olsona root 55948 Jan 26 10:19 itca.jpg -r--r--r-- 1 olsona root 3080 Jan 26 10:19 leapseconds -r--r--r-- 1 olsona root 45789 Jan 26 10:19 localtime.c -r--r--r-- 1 olsona root 2010 Jan 26 10:19 logwtmp.c -r--r--r-- 1 olsona root 15516 Jan 26 10:19 Makefile -r--r--r-- 1 olsona root 5766 Jan 26 10:19 newctime.3 -rw-r--r-- 1 olsona root 5942 Jan 30 08:59 newctime.3.txt -r--r--r-- 1 olsona root 6732 Jan 26 10:19 newstrftime.3 -rw-r--r-- 1 olsona root 4995 Jan 30 08:59 newstrftime.3.txt -r--r--r-- 1 olsona root 5338 Jan 26 10:19 newtzset.3 -rw-r--r-- 1 olsona root 6850 Jan 30 08:59 newtzset.3.txt -r--r--r-- 1 olsona root 92970 Jan 30 08:59 northamerica -r--r--r-- 1 olsona root 1112 Jan 26 10:19 pacificnew -r--r--r-- 1 olsona root 6839 Jan 26 10:19 private.h -r--r--r-- 1 olsona root 3739 Jan 26 10:19 README -r--r--r-- 1 olsona root 1105 Jan 26 10:19 scheck.c -r--r--r-- 1 olsona root 19224 Jan 26 10:19 solar87 -r--r--r-- 1 olsona root 19242 Jan 26 10:19 solar88 -r--r--r-- 1 olsona root 19518 Jan 26 10:19 solar89 -r--r--r-- 1 olsona root 45285 Jan 26 10:19 southamerica -r--r--r-- 1 olsona root 18681 Jan 26 10:19 strftime.c -r--r--r-- 1 olsona root 1464 Jan 26 10:19 systemv -r--r--r-- 1 olsona root 24038 Jan 26 10:19 Theory -r--r--r-- 1 olsona root 3290 Jan 26 10:19 time2posix.3 -rw-r--r-- 1 olsona root 3402 Jan 30 08:59 time2posix.3.txt -rw-r--r-- 1 olsona root 0 Feb 6 10:17 typescript -r--r--r-- 1 olsona root 11956 Jan 26 10:19 tz-art.htm -r--r--r-- 1 olsona root 28222 Jan 30 08:59 tz-link.htm -rw-r--r-- 1 olsona root 180636 Feb 6 10:18 tzcode2006a.tar.gz -rw-r--r-- 1 olsona root 149612 Feb 6 10:18 tzdata2006a.tar.gz -r--r--r-- 1 olsona root 3669 Jan 26 10:19 tzfile.5 -rw-r--r-- 1 olsona root 3789 Jan 30 08:59 tzfile.5.txt -r--r--r-- 1 olsona root 4792 Jan 26 10:19 tzfile.h -r--r--r-- 1 olsona root 1100 Jan 26 10:19 tzselect.8 -rw-r--r-- 1 olsona root 1089 Jan 30 08:59 tzselect.8.txt -r--r--r-- 1 olsona root 6880 Jan 26 10:19 tzselect.ksh -r--r--r-- 1 olsona root 5839 Jan 26 10:19 usno1988 -r--r--r-- 1 olsona root 21156 Jan 26 10:19 usno1989 -r--r--r-- 1 olsona root 21154 Jan 26 10:19 usno1989a -r--r--r-- 1 olsona root 10354 Jan 26 10:19 usno1995 -r--r--r-- 1 olsona root 10405 Jan 26 10:19 usno1997 -r--r--r-- 1 olsona root 10405 Jan 26 10:19 usno1998 -r--r--r-- 1 olsona root 404 Jan 26 10:19 workman.sh -r--r--r-- 1 olsona root 610 Jan 26 10:19 yearistype.sh -r--r--r-- 1 olsona root 1365 Jan 26 10:19 zdump.8 -rw-r--r-- 1 olsona root 1449 Jan 30 08:59 zdump.8.txt -r--r--r-- 1 olsona root 15656 Jan 26 10:19 zdump.c -r--r--r-- 1 olsona root 10192 Jan 26 10:19 zic.8 -rw-r--r-- 1 olsona root 10463 Jan 30 08:59 zic.8.txt -r--r--r-- 1 olsona root 53629 Jan 26 10:19 zic.c -r--r--r-- 1 olsona root 17770 Jan 30 08:59 zone.tab lecserver$ grep -i load * date.1:UTC leap seconds are loaded from date.1.txt: are loaded from /usr/local/etc/zoneinfo/posixrules. etcetera:# as functions like gmtime load the "GMT" file to handle leap seconds properly. localtime.c:** The DST rules to use if TZ has no rules and we can't load TZDEFRULES. localtime.c:static void gmtload P((struct state * sp)); localtime.c:static int tzload P((const char * name, struct state * sp)); localtime.c:tzload(name, sp) localtime.c: register int load_result; localtime.c: load_result = tzload(TZDEFRULES, sp); localtime.c: if (load_result != 0) localtime.c: if (*name == '\0' && load_result != 0) localtime.c:gmtload(sp) localtime.c: if (tzload(gmt, sp) != 0) localtime.c: if (tzload((char *) NULL, lclptr) != 0) localtime.c: gmtload(lclptr); localtime.c: } else if (tzload(name, lclptr) != 0) localtime.c: (void) gmtload(lclptr); localtime.c: gmtload(gmtptr); newctime.3:UTC leap seconds are loaded from newctime.3.txt: are loaded from /usr/local/etc/zoneinfo/posixrules. newtzset.3:UTC leap seconds are loaded from newtzset.3.txt: are loaded from /usr/local/etc/zoneinfo/posixrules. tz-link.htm:The following shell commands download tz-link.htm:see the downloaded tz-link.htm:href="http://www.manifold.net/download/freemaps.html">Free Maps and lecserver$ exit script done on Mon 06 Feb 2006 10:18:55 AM EST ________________________________ From: Zeb Olaff [mailto:zebonebo@gmail.com] Sent: Friday, February 03, 2006 8:45 PM To: tz(a)lecserver.nci.nih.gov Subject: Misspelling With the Australia/Lord_Howe Time Zone I am using your database and have found that the full name for the Australia/Lord_Howe reads "Load Howe" instead of "Lord Howe". Hopefully this is the correct email address to send this information to. Thank You, Zeb Olaf

1 0

FW: Misspelling With the Australia/Lord_Howe Time Zone
by Olson, Arthur David (NIH/NCI) [E] Feb. 6, 2006

Feb. 6, 2006

Zeb Olaff is not on the time zone mailing list; direct replies appropriately. --ado ________________________________ From: Zeb Olaff [mailto:zebonebo@gmail.com] Sent: Friday, February 03, 2006 8:45 PM To: tz(a)lecserver.nci.nih.gov Subject: Misspelling With the Australia/Lord_Howe Time Zone I am using your database and have found that the full name for the Australia/Lord_Howe reads "Load Howe" instead of "Lord Howe". Hopefully this is the correct email address to send this information to. Thank You, Zeb Olaf

1 0

Misspelling With the Australia/Lord_Howe Time Zone
by Zeb Olaff Feb. 4, 2006

Feb. 4, 2006

I am using your database and have found that the full name for the Australia/Lord_Howe reads "Load Howe" instead of "Lord Howe". Hopefully this is the correct email address to send this information to. Thank You, Zeb Olaf

1 0

Re: FW: tzcode2006a.tar.gz on Irix 6.5.18m
by Paul Eggert Feb. 3, 2006

Feb. 3, 2006

Andrew Donaldson <A.Donaldson(a)bom.gov.au> writes: > Yell out if you would like anything else tested. Thanks for looking into the problem. There were a lot of changes between the non-working and the working zdump.c. I enclose them below. One thing that jumps out is the new setabsoletes() function. I suspect that it is not working properly on your host. Can you verify this, by using a debugger and/or inserting print statements? I would like to know the value of absolute_min_time and absolute_max_time when setabsoletes returns, and I'd also like to know the size and signedness of time_t on your host. (The code is relying on undefined behavior here, so it's not strictly portable.) Thanks. --- tz-2004gg/zdump.c 2004-10-18 08:37:53.000000000 -0700 +++ tz-2004hg/zdump.c 2004-12-07 08:40:04.000000000 -0800 @@ -1,4 +1,4 @@ -static char elsieid[] = "@(#)zdump.c 7.47"; +static char elsieid[] = "@(#)zdump.c 7.57"; /* ** This code has been made independent of the rest of the time @@ -11,6 +11,15 @@ static char elsieid[] = "@(#)zdump.c 7.4 #include "sys/types.h" /* for time_t */ #include "time.h" /* for struct tm */ #include "stdlib.h" /* for exit, malloc, atoi */ +#include "float.h" /* for FLT_MAX and DBL_MAX */ + +#ifndef ZDUMP_LO_YEAR +#define ZDUMP_LO_YEAR (-500) +#endif /* !defined ZDUMP_LO_YEAR */ + +#ifndef ZDUMP_HI_YEAR +#define ZDUMP_HI_YEAR 2500 +#endif /* !defined ZDUMP_HI_YEAR */ #ifndef MAX_STRING_LENGTH #define MAX_STRING_LENGTH 1024 @@ -71,6 +80,10 @@ static char elsieid[] = "@(#)zdump.c 7.4 #define isleap_sum(a, b) isleap((a) % 400 + (b) % 400) #endif /* !defined isleap_sum */ +#define SECSPERDAY ((long) SECSPERHOUR * HOURSPERDAY) +#define SECSPERNYEAR (SECSPERDAY * DAYSPERNYEAR) +#define SECSPERLYEAR (SECSPERNYEAR + SECSPERDAY) + #if HAVE_GETTEXT #include "locale.h" /* for setlocale */ #include "libintl.h" @@ -127,13 +140,18 @@ extern char * optarg; extern int optind; extern char * tzname[2]; +static time_t absolute_min_time; +static time_t absolute_max_time; +static size_t longest; +static char * progname; + static char * abbr P((struct tm * tmp)); static long delta P((struct tm * newp, struct tm * oldp)); +static void dumptime P((const struct tm * tmp)); static time_t hunt P((char * name, time_t lot, time_t hit)); -static size_t longest; -static char * progname; +static void setabsolutes(); static void show P((char * zone, time_t t, int v)); -static void dumptime P((const struct tm * tmp)); +static time_t yeartot P((long y)); int main(argc, argv) @@ -143,18 +161,22 @@ char * argv[]; register int i; register int c; register int vflag; - register char * cutoff; - register int cutyear; - register long cuttime; - char ** fakeenv; + register char * cutarg; + register long cutloyear = ZDUMP_LO_YEAR; + register long cuthiyear = ZDUMP_HI_YEAR; + register time_t cutlotime; + register time_t cuthitime; + register char ** fakeenv; time_t now; time_t t; time_t newt; - time_t hibit; struct tm tm; struct tm newtm; + register struct tm * tmp; + register struct tm * newtmp; - INITIALIZE(cuttime); + INITIALIZE(cutlotime); + INITIALIZE(cuthitime); #if HAVE_GETTEXT (void) setlocale(LC_MESSAGES, ""); #ifdef TZ_DOMAINDIR @@ -169,34 +191,45 @@ char * argv[]; (void) exit(EXIT_SUCCESS); } vflag = 0; - cutoff = NULL; + cutarg = NULL; while ((c = getopt(argc, argv, "c:v")) == 'c' || c == 'v') if (c == 'v') vflag = 1; - else cutoff = optarg; + else cutarg = optarg; if ((c != EOF && c != -1) || (optind == argc - 1 && strcmp(argv[optind], "=") == 0)) { (void) fprintf(stderr, -_("%s: usage is %s [ --version ] [ -v ] [ -c cutoff ] zonename ...\n"), - argv[0], argv[0]); +_("%s: usage is %s [ --version ] [ -v ] [ -c [loyear,]hiyear ] zonename ...\n"), + progname, progname); + (void) exit(EXIT_FAILURE); + } + if (vflag) { + if (cutarg != NULL) { + long lo; + long hi; + char c; + + if (sscanf(cutarg, "%ld%c", &hi, &c) == 1) { + cuthiyear = hi; + } else if (sscanf(cutarg, "%ld,%ld%c", + &lo, &hi, &c) == 2) { + cutloyear = lo; + cuthiyear = hi; + } else { +(void) fprintf(stderr, _("%s: wild -c argument %s\n"), + progname, cutarg); (void) exit(EXIT_FAILURE); } - if (cutoff != NULL) { - int y; - - cutyear = atoi(cutoff); - cuttime = 0; - for (y = EPOCH_YEAR; y < cutyear; ++y) - cuttime += DAYSPERNYEAR + isleap(y); - cuttime *= SECSPERHOUR * HOURSPERDAY; + } + setabsolutes(); + cutlotime = yeartot(cutloyear); + cuthitime = yeartot(cuthiyear); } (void) time(&now); longest = 0; for (i = optind; i < argc; ++i) if (strlen(argv[i]) > longest) longest = strlen(argv[i]); - for (hibit = 1; (hibit << 1) != 0; hibit <<= 1) - continue; { register int from; register int to; @@ -226,58 +259,129 @@ _("%s: usage is %s [ --version ] [ -v ] show(argv[i], now, FALSE); continue; } - /* - ** Get lowest value of t. - */ - t = hibit; - if (t > 0) /* time_t is unsigned */ - t = 0; + t = absolute_min_time; show(argv[i], t, TRUE); t += SECSPERHOUR * HOURSPERDAY; show(argv[i], t, TRUE); - tm = *localtime(&t); + if (t < cutlotime) + t = cutlotime; + tmp = localtime(&t); + if (tmp != NULL) { + tm = *tmp; (void) strncpy(buf, abbr(&tm), (sizeof buf) - 1); + } for ( ; ; ) { - if (cutoff != NULL && t >= cuttime) + if (t >= cuthitime) break; newt = t + SECSPERHOUR * 12; - if (cutoff != NULL && newt >= cuttime) + if (newt >= cuthitime) break; if (newt <= t) break; - newtm = *localtime(&newt); - if (delta(&newtm, &tm) != (newt - t) || + newtmp = localtime(&newt); + if (newtmp != NULL) + newtm = *newtmp; + if ((tmp == NULL || newtmp == NULL) ? (tmp != newtmp) : + (delta(&newtm, &tm) != (newt - t) || newtm.tm_isdst != tm.tm_isdst || - strcmp(abbr(&newtm), buf) != 0) { + strcmp(abbr(&newtm), buf) != 0)) { newt = hunt(argv[i], t, newt); - newtm = *localtime(&newt); - (void) strncpy(buf, abbr(&newtm), + newtmp = localtime(&newt); + if (newtmp != NULL) { + newtm = *newtmp; + (void) strncpy(buf, + abbr(&newtm), (sizeof buf) - 1); } + } t = newt; tm = newtm; + tmp = newtmp; } - /* - ** Get highest value of t. - */ - t = ~((time_t) 0); - if (t < 0) /* time_t is signed */ - t &= ~hibit; + t = absolute_max_time; t -= SECSPERHOUR * HOURSPERDAY; show(argv[i], t, TRUE); t += SECSPERHOUR * HOURSPERDAY; show(argv[i], t, TRUE); } if (fflush(stdout) || ferror(stdout)) { - (void) fprintf(stderr, "%s: ", argv[0]); + (void) fprintf(stderr, "%s: ", progname); (void) perror(_("Error writing standard output")); (void) exit(EXIT_FAILURE); } exit(EXIT_SUCCESS); + /* If exit fails to exit... */ + return EXIT_FAILURE; +} - /* gcc -Wall pacifier */ - for ( ; ; ) +static void +setabsolutes() +{ + if (0.5 == (time_t) 0.5) { + /* + ** time_t is floating. + */ + if (sizeof (time_t) == sizeof (float)) { + absolute_min_time = (time_t) -FLT_MAX; + absolute_max_time = (time_t) FLT_MAX; + } else if (sizeof (time_t) == sizeof (double)) { + absolute_min_time = (time_t) -DBL_MAX; + absolute_max_time = (time_t) DBL_MAX; + } else { + (void) fprintf(stderr, +_("%s: use of -v on system with floating time_t other than float or double\n"), + progname); + (void) exit(EXIT_FAILURE); + } + } else if (0 > (time_t) -1) { + /* + ** time_t is signed. + */ + register time_t hibit; + + for (hibit = 1; (hibit * 2) != 0; hibit *= 2) continue; + absolute_min_time = hibit; + absolute_max_time = -(hibit + 1); + } else { + /* + ** time_t is unsigned. + */ + absolute_min_time = 0; + absolute_max_time = absolute_min_time - 1; + } +} + +static time_t +yeartot(y) +const long y; +{ + register long myy; + register long seconds; + register time_t t; + + myy = EPOCH_YEAR; + t = 0; + while (myy != y) { + if (myy < y) { + seconds = isleap(myy) ? SECSPERLYEAR : SECSPERNYEAR; + ++myy; + if (t > absolute_max_time - seconds) { + t = absolute_max_time; + break; + } + t += seconds; + } else { + --myy; + seconds = isleap(myy) ? SECSPERLYEAR : SECSPERNYEAR; + if (t < absolute_min_time + seconds) { + t = absolute_min_time; + break; + } + t -= seconds; + } + } + return t; } static time_t @@ -287,24 +391,38 @@ time_t lot; time_t hit; { time_t t; + long diff; struct tm lotm; + register struct tm * lotmp; struct tm tm; - static char loab[MAX_STRING_LENGTH]; + register struct tm * tmp; + char loab[MAX_STRING_LENGTH]; - lotm = *localtime(&lot); + lotmp = localtime(&lot); + if (lotmp != NULL) { + lotm = *lotmp; (void) strncpy(loab, abbr(&lotm), (sizeof loab) - 1); - while ((hit - lot) >= 2) { - t = lot / 2 + hit / 2; + } + for ( ; ; ) { + diff = (long) (hit - lot); + if (diff < 2) + break; + t = lot; + t += diff / 2; if (t <= lot) ++t; else if (t >= hit) --t; - tm = *localtime(&t); - if (delta(&tm, &lotm) == (t - lot) && + tmp = localtime(&t); + if (tmp != NULL) + tm = *tmp; + if ((lotmp == NULL || tmp == NULL) ? (lotmp == tmp) : + (delta(&tm, &lotm) == (t - lot) && tm.tm_isdst == lotm.tm_isdst && - strcmp(abbr(&tm), loab) == 0) { + strcmp(abbr(&tm), loab) == 0)) { lot = t; lotm = tm; + lotmp = tmp; } else hit = t; } show(name, lot, TRUE); @@ -321,8 +439,8 @@ delta(newp, oldp) struct tm * newp; struct tm * oldp; { - long result; - int tmy; + register long result; + register int tmy; if (newp->tm_year < oldp->tm_year) return -delta(oldp, newp); @@ -345,15 +463,22 @@ char * zone; time_t t; int v; { - struct tm * tmp; + register struct tm * tmp; (void) printf("%-*s ", (int) longest, zone); if (v) { - dumptime(gmtime(&t)); - (void) printf(" UTC = "); + tmp = gmtime(&t); + if (tmp == NULL) { + (void) printf("%g", (double) t); + } else { + dumptime(tmp); + (void) printf(" UTC"); + } + (void) printf(" = "); } tmp = localtime(&t); dumptime(tmp); + if (tmp != NULL) { if (*abbr(tmp) != '\0') (void) printf(" %s", abbr(tmp)); if (v) { @@ -362,6 +487,7 @@ int v; (void) printf(" gmtoff=%ld", tmp->TM_GMTOFF); #endif /* defined TM_GMTOFF */ } + } (void) printf("\n"); } @@ -394,6 +520,10 @@ register const struct tm * timeptr; register int lead; register int trail; + if (timeptr == NULL) { + (void) printf("NULL"); + return; + } /* ** The packaged versions of localtime and gmtime never put out-of-range ** values in tm_wday or tm_mon, but since this code might be compiled

1 0

FW: Time zone error for US cities.
by Olson, Arthur David (NIH/NCI) [E] Feb. 2, 2006

Feb. 2, 2006

Jian Zhang is not on the time zone mailling list; direct replies appropriately. --ado ________________________________ From: Jian Zhang [mailto:Jian.Zhang@Campbell.com] Sent: Wednesday, February 01, 2006 3:26 PM To: tz(a)lecserver.nci.nih.gov Subject: Time zone error for US cities. Dear Sir/Madam, I am trying to use the Time zone converter and got an error for US cities. The error is: When I try to convert the New_York time from: 30-Apr-1972 02:10:30 To GMT time, it gives me the time as: 31-Dec-1969 23:59:59 That is clearly wrong. Does any one have the same problem? The error is isolated to the 02:00:00 to 03:00:00 on that date. Thanks Email jian.zhang(a)campbell.com Jian ______________________________________________________________________ Campbell & Company, Inc.: The information in this e-mail may contain privileged/confidential information. If you are not the intended recipient, you must not read, use, copy or disseminate the information or take any action in reliance thereupon. If you have received this e-mail in error, please notify Campbell & Company, Inc. immediately by e-mail or telephone and delete the e-mail and any attachments from any computer. The information in this e-mail does not constitute an offer to sell or the solicitation of an offer to buy any securities in any jurisdiction or for the benefit of any person. ______________________________________________________________________ This email has been scanned by the MessageLabs Email Security System. For more information please visit http://www.messagelabs.com/email ______________________________________________________________________

2 1

FW: tzcode2006a.tar.gz on Irix 6.5.18m
by Olson, Arthur David (NIH/NCI) [E] Feb. 2, 2006

Feb. 2, 2006

Andrew Donaldson is not on the time zone mailing list; direct replies appropriately. --ado -----Original Message----- From: Andrew Donaldson [mailto:A.Donaldson@bom.gov.au] Sent: Wednesday, February 01, 2006 10:48 PM To: tz(a)lecserver.nci.nih.gov Cc: Rob Petrini Subject: tzcode2006a.tar.gz on Irix 6.5.18m Dear Sir, We had just had reason to update some of our SGI machines' timezone data to reflect the changes in Melbourne, Australia due to the Commonwealth Games. Please note that we are hardly experts in this area, so we may have missed something obvious. We found that the zdump utility, used with the -v option, would only output the first times. $ ./zdump -v Australia/Melbourne Australia/Melbourne Thu Jan 1 00:00:00 1970 UTC = Thu Jan 1 10:00:00 1970 EST isdst=0 Australia/Melbourne Fri Jan 2 00:00:00 1970 UTC = Fri Jan 2 10:00:00 1970 EST isdst=0 Australia/Melbourne Tue Dec 30 23:59:59 1969 UTC = Wed Dec 31 09:59:59 1969 EST isdst=0 Australia/Melbourne Wed Dec 31 23:59:59 1969 UTC = Thu Jan 1 09:59:59 1970 EST isdst=0 After a bit of head scratching, we proved that the data files were valid, by viewing them on a linux computer, so we tried the source code from the classictzcode.tar.gz file which produced valid results both with the classictzdata.tar.gz and tzdata2006a.tar.gz data, the later including the updates to Melbourne time that we desired. I am unsure why the tzcode2006a.tar.gz version of zdump behaved differently. To compile tzcode2006a.tar.gz I had to make these changes: $ diff localtime.c localtime.c.orig 512d511 < /* 517,518d515 < */ < const char * getqzname P((const char * strp, const char delim)) To compile classictzcode.tar.gz I had to make these changes: $ diff zic.c zic.c.orig 64,65c64 < /*extern int getopt P((int argc, char * argv[], const char * options));*/ < extern int getopt P((int, char *const *, const char *)); --- > extern int getopt P((int argc, char * argv[], const char * options)); and $ diff getopt.c getopt.c.orig 52d51 < /* 57,58d55 < */ < int getopt(int nargc, char *const *nargv, const char *ostr) I hope our findings are useful to you. regards, Andrew Donaldson.

2 1