- tz - lists.iana.org

Clerical patches
by Tim Parenti Aug. 16, 2014

Aug. 16, 2014

A few more clerical patches, relative to 7217d26487456e93094be4c3befb6cdccfdc7bcc, dated Wed Aug 13 11:46:39 2014 -0700. 0001 fixes up some more minor spacing issues missed in my last round of patches on 2014-08-09. 0002 fixes another typo in recent additions to NEWS, and adds a space after the redirection operator in the new CONTRIBUTING file for added clarity. 0003 renames the fields of Link lines to be more descriptive and more like the parameters of 'ln'. (Personally, I have found LINK-FROM and LINK-TO to be the opposite of my expectations, and these are not particularly descriptive names anyway.) 0004 standardizes Link lines to use single tabs. These can be compared at https://github.com/timparenti/tz-experimental/compare/eggert:master...timpa… -- Tim Parenti

3 8

[PROPOSED PATCH] * private.h (timegm, timelocal) [time_tz]: Define replacements.
by Paul Eggert Aug. 16, 2014

Aug. 16, 2014

Needed on Ubuntu 14.04 when compiling with -D_GNU_SOURCE -Dtime_tz=.... --- private.h | 10 ++++++++++ 1 file changed, 10 insertions(+) diff --git a/private.h b/private.h index 6d3e6aa..38f3f43 100644 --- a/private.h +++ b/private.h @@ -278,6 +278,10 @@ static time_t sys_time(time_t *x) { return time(x); } # define time2posix tz_time2posix # undef time_t # define time_t tz_time_t +# undef timegm +# define timegm tz_timegm +# undef timelocal +# define timelocal tz_timelocal # undef timeoff # define timeoff tz_timeoff @@ -316,6 +320,12 @@ void tzsetwall(void); # if !defined offtime || defined time_tz struct tm *offtime(time_t const *, long); # endif +# if !defined timegm || defined time_tz +time_t timegm(struct tm *); +# endif +# if !defined timelocal || defined time_tz +time_t timelocal(struct tm *); +# endif # if !defined timeoff || defined time_tz time_t timeoff(struct tm *, long); # endif -- 1.9.1

1 0

Re: [tz] The case against time zones
by Patrice Scattolin Aug. 16, 2014

Aug. 16, 2014

It would be a small sacrifice for the Physics community for the betterment of mankind. Alas it was never meant to be. Because even if you change the second, you don't eliminate the need for the leap second. Luckikly with the current lenght we always need to add seconds. With a better second the error would be randomly distributed on both sides. I am not aware of a case where we have to remove an extraneous second. Imagine a day that is missing a second and ends with 11:59:58! ----- Original Message ----- From: lester(a)lsces.co.uk To: tz(a)iana.org Sent: Saturday, August 16, 2014 2:15:22 AM GMT -05:00 US/Canada Eastern Subject: Re: [tz] The case against time zones On 15/08/14 21:02, Clive D.W. Feather wrote: >>> No, it's the length of the earth day that is changing over time as >>> > > earth's rotation slows due to natural forces. >>> > > http://www.washingtonpost.com/national/health-science/why-in-the-jurassic-e… >> > >> > Long term yes. Shorter term, say next 100 years, a small increase in >> > period of a second would be a suitable alternative to the 'problem' of >> > leap seconds ... > At the cost of changing every physical constant, starting with the force of > gravity and the speed of light. Helps to type first :( Finally somebody spotted the problem ... But if the original calculation of number of cycles for a second was wrong everything is wrong now anyway? :( -- Lester Caine - G8HFL ----------------------------- Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk

1 0

[PATCH] Remove lint when debugging by compiling with -Dtime_tz='whatever'.
by Paul Eggert Aug. 16, 2014

Aug. 16, 2014

* localtime.c (LOCALTIME_IMPLEMENTATION): New macro. (time) [time_tz]: Move here from private.h, and make external. * private.h (sys_time) [time_tz]: Define only if LOCALTIME_IMPLEMENTATION, so that the code doesn't define an unused static function sys_time that the compiler might complain about. (time) [time_tz]: Move to localtime.c. --- localtime.c | 18 ++++++++++++++++++ private.h | 12 +++--------- 2 files changed, 21 insertions(+), 9 deletions(-) diff --git a/localtime.c b/localtime.c index 840fa21..551b5a2 100644 --- a/localtime.c +++ b/localtime.c @@ -10,7 +10,9 @@ /*LINTLIBRARY*/ +#define LOCALTIME_IMPLEMENTATION #include "private.h" + #include "tzfile.h" #include "fcntl.h" @@ -2022,3 +2024,19 @@ posix2time(time_t t) } #endif /* defined STD_INSPIRED */ + +#ifdef time_tz + +/* Convert from the underlying system's time_t to the ersatz time_tz, + which is called 'time_t' in this file. */ + +time_t +time(time_t *p) +{ + time_t r = sys_time(0); + if (p) + *p = r; + return r; +} + +#endif diff --git a/private.h b/private.h index c778035..6d3e6aa 100644 --- a/private.h +++ b/private.h @@ -248,7 +248,9 @@ typedef unsigned long uintmax_t; ** typical platforms. */ #ifdef time_tz +# ifdef LOCALTIME_IMPLEMENTATION static time_t sys_time(time_t *x) { return time(x); } +# endif # undef ctime # define ctime tz_ctime @@ -289,15 +291,7 @@ struct tm *gmtime_r(time_t const *restrict, struct tm *restrict); struct tm *localtime(time_t const *); struct tm *localtime_r(time_t const *restrict, struct tm *restrict); time_t mktime(struct tm *); - -static time_t -time(time_t *p) -{ - time_t r = sys_time(0); - if (p) - *p = r; - return r; -} +time_t time(time_t *); #endif /* -- 1.9.1

1 0

[PROPOSED PATCH] * date.c, localtime.c, zdump.c: Simplify prototypes for static functions
by Paul Eggert Aug. 15, 2014

Aug. 15, 2014

by removing prototypes that aren't needed, and by removing names of argumnts. This will make future changes simpler and easier to follow. --- date.c | 2 -- localtime.c | 69 +++++++++++-------------------------------------------------- zdump.c | 15 +++++++------- 3 files changed, 19 insertions(+), 67 deletions(-) diff --git a/date.c b/date.c index 3d78914..de56652 100644 --- a/date.c +++ b/date.c @@ -60,11 +60,9 @@ static void display(const char *, time_t); static void dogmt(void); static void errensure(void); static void iffy(time_t, time_t, const char *, const char *); -int main(int, char**); static const char * nondigit(const char *); static void oops(const char *); static void reset(time_t, int); -static int sametm(const struct tm *, const struct tm *); static void timeout(FILE *, const char *, const struct tm *); static void usage(void); static void wildinput(const char *, const char *, diff --git a/localtime.c b/localtime.c index e2cf5c2..840fa21 100644 --- a/localtime.c +++ b/localtime.c @@ -126,59 +126,14 @@ struct rule { #define DAY_OF_YEAR 1 /* n = day of year */ #define MONTH_NTH_DAY_OF_WEEK 2 /* Mm.n.d = month, week, day of week */ -/* -** Prototypes for static functions. -*/ - -static int_fast32_t detzcode(const char * codep); -static int_fast64_t detzcode64(const char * codep); -static int differ_by_repeat(time_t t1, time_t t0); -static const char * getzname(const char * strp) ATTRIBUTE_PURE; -static const char * getqzname(const char * strp, const int delim) - ATTRIBUTE_PURE; -static const char * getnum(const char * strp, int * nump, int min, - int max); -static const char * getsecs(const char * strp, int_fast32_t * secsp); -static const char * getoffset(const char * strp, int_fast32_t * offsetp); -static const char * getrule(const char * strp, struct rule * rulep); -static void gmtload(struct state * sp); -static struct tm * gmtsub(const time_t * timep, int_fast32_t offset, - struct tm * tmp); -static struct tm * localsub(const time_t * timep, int_fast32_t offset, - struct tm * tmp); -static int increment_overflow(int * number, int delta); -static int leaps_thru_end_of(int y) ATTRIBUTE_PURE; -static int increment_overflow32(int_fast32_t * number, int delta); -static int increment_overflow_time(time_t *t, int_fast32_t delta); -static int normalize_overflow32(int_fast32_t * tensptr, - int * unitsptr, int base); -static int normalize_overflow(int * tensptr, int * unitsptr, - int base); -static void settzname(void); -static time_t time1(struct tm * tmp, - struct tm * (*funcp)(const time_t *, - int_fast32_t, struct tm *), - int_fast32_t offset); -static time_t time2(struct tm *tmp, - struct tm * (*funcp)(const time_t *, - int_fast32_t, struct tm*), - int_fast32_t offset, int * okayp); -static time_t time2sub(struct tm *tmp, - struct tm * (*funcp)(const time_t *, - int_fast32_t, struct tm*), - int_fast32_t offset, int * okayp, int do_norm_secs); -static struct tm * timesub(const time_t * timep, int_fast32_t offset, - const struct state * sp, struct tm * tmp); -static int tmcomp(const struct tm * atmp, - const struct tm * btmp); -static int_fast32_t transtime(int year, const struct rule * rulep, - int_fast32_t offset) - ATTRIBUTE_PURE; -static int typesequiv(const struct state * sp, int a, int b); -static int tzload(const char * name, struct state * sp, - int doextend); -static int tzparse(const char * name, struct state * sp, - int lastditch); +static struct tm *gmtsub(time_t const *, int_fast32_t, struct tm *); +static int increment_overflow(int *, int); +static int increment_overflow_time(time_t *, int_fast32_t); +static int normalize_overflow32(int_fast32_t *, int *, int); +static struct tm *timesub(time_t const *, int_fast32_t, struct state const *, + struct tm *); +static int typesequiv(struct state const *, int, int); +static int tzparse(char const *, struct state *, int); #ifdef ALL_STATE static struct state * lclptr; @@ -657,7 +612,7 @@ static const int year_lengths[2] = { ** character. */ -static const char * +static const char * ATTRIBUTE_PURE getzname(register const char *strp) { register char c; @@ -677,7 +632,7 @@ getzname(register const char *strp) ** We don't do any checking here; checking is done later in common-case code. */ -static const char * +static const char * ATTRIBUTE_PURE getqzname(register const char *strp, const int delim) { register int c; @@ -839,7 +794,7 @@ getrule(const char *strp, register struct rule *const rulep) ** effect, calculate the year-relative time that rule takes effect. */ -static int_fast32_t +static int_fast32_t ATTRIBUTE_PURE transtime(const int year, register const struct rule *const rulep, const int_fast32_t offset) { @@ -1419,7 +1374,7 @@ offtime(const time_t *const timep, const long offset) ** where, to make the math easy, the answer for year zero is defined as zero. */ -static int +static int ATTRIBUTE_PURE leaps_thru_end_of(register const int y) { return (y >= 0) ? (y / 4 - y / 100 + y / 400) : diff --git a/zdump.c b/zdump.c index 859a169..515a80f 100644 --- a/zdump.c +++ b/zdump.c @@ -207,14 +207,13 @@ static size_t longest; static char * progname; static int warned; -static char * abbr(struct tm * tmp); -static void abbrok(const char * abbrp, const char * zone); -static intmax_t delta(struct tm * newp, struct tm * oldp) ATTRIBUTE_PURE; -static void dumptime(const struct tm * tmp); -static time_t hunt(char * name, time_t lot, time_t hit); -static void show(char * zone, time_t t, int v); -static const char * tformat(void); -static time_t yeartot(intmax_t y) ATTRIBUTE_PURE; +static char *abbr(struct tm *); +static intmax_t delta(struct tm *, struct tm *) ATTRIBUTE_PURE; +static void dumptime(struct tm const *); +static time_t hunt(char *, time_t, time_t); +static void show(char *, time_t, int); +static const char *tformat(void); +static time_t yeartot(intmax_t) ATTRIBUTE_PURE; /* Is A an alphabetic character in the C locale? */ static int -- 1.9.1

1 0

Proposed reversions, for moving forward
by Tim Parenti Aug. 14, 2014

Aug. 14, 2014

I propose the attached patches, described as follows, be applied to the master branch of our repository so that we have a stable starting point from which to adopt better Git practices moving forward. To assist in review, I have pushed these commits to https://github.com/timparenti/tz-experimental/commits/revert-eggert-2014-07. The first patch aims to revert, in their entirety, Paul’s three commits related to zone-linking, dated 2014-07-08, -09, and -15. Further work on this effort should be done on a new and separate branch. The second patch aims to revert, in their entirety, Paul’s two commits related to the introduction of time.tab, dated 2014-07-18 and -19. Further work on this effort should be done on a new and separate branch. The third, fourth, and fifth patches reapply changes cherry-picked from the above commits which relate to data and are less controversial. These are relative to the current state of master at 87163d184d7cc0c704d2b505adcfb23528203951 <https://github.com/eggert/tz/commit/87163d184d7cc0c704d2b505adcfb23528203951>, dated Tue Jul 22 20:54:19 2014 -0700. -- Tim Parenti

16 46

Philippines may do off-and-on DST starting November
by Paul Eggert Aug. 14, 2014

Aug. 14, 2014

Last week Philippine House Deputy Minority Leader Arnel Ty urged President Aquino to declare daylight saving time from November to January and again from March to June, in an attempt to save electrical power. We had some obsolete commentary about a failed attempt to do something similar in 2006, and the attached commentary patch (installed in the experimental version) does some housekeeping in this area.

1 0

Re: [tz] [Tzdist] Time Zone Distribution Protocol WG Milestones (tzdist)
by Daniel Migault Aug. 13, 2014

Aug. 13, 2014

Hi Simon, Thanks for the feed back. On Wed, Aug 13, 2014 at 5:26 PM, Simon Perreault <sperreault(a)jive.com> wrote: > Tobias, > > A draft's name is irrelevant. It will be gone once it gets published > as an RFC. Other than in the two drafts' names, "timezone" is absent > from the proposed milestone text so I don't see any terminology issue. > > All, > > My comments on the proposed milestones: > > - I don't see the need for 2-3 months between WGLC and submission to > IESG. If issues are tracked tightly during the whole process > [RFC7282], WGLC becomes a formality and you can send to IESG right > away. In any case, I suggest removing the WGLC milestones: the "send > to IESG" milestones are sufficient. > > I got your point. The time between WGLC and IESG submission has been provisioned in case some issues/comments are lately reported on the mailing list. However we hope this should not happen and encourage people to comment as soon as possible. > - "September 2016 re-charter the WG": This should not be a milestone. > Rechartering is something that is done out of necessity in the present > moment, not something that a WG aims to do in the future. Not having > the milestone does not impact in any way the possibility of a > rechartering. > > I am also fine with that. > Simon > > > On Wed, Aug 13, 2014 at 11:06 AM, Tobias Conradi <tc(a)tobiasconradi.com> > wrote: > > On Wed, Aug 13, 2014 at 3:06 PM, Daniel Migault <mglt.ietf(a)gmail.com> > wrote: > > ... > >> These are the milestones we propose. If you have any comments, please > feel > >> free to post them before Tuesday August 19 so we can consider them. > > > > 2014-08-13 today > > 2014-08-19 the mentioned Tuesday > > = 6 days > > > > This seems rather short, see > > http://www.ietf.org/mail-archive/web/tzdist/current/msg00004.html > > 2014-07-03 proposal to Mr. Daboo to change "timezone" to "time zone" > > > > 2014-08-08 this change has been done for the WG name > > > > = 1 month, 5 days > > > > But simple renaming not yet done for the drafts mentioned in the charter > > -------------- > > The working group will use the following drafts as initial input for its > > work: > > draft-douglass-timezone-service-11 > > draft-daboo-caldav-timezone-ref-01 > > -------------- > > > > So, shall people use drafts with outdated terminology as basis for > discussion? > > > > I prefer to fix the drafts ASAP so that their contents can be > > discussed using consistent terminology. > > > > -- > > Tobias Conradi > > Rheinsberger Str. 18 > > 10115 Berlin > > Germany > > > > http://tobiasconradi.com > > > > _______________________________________________ > > Tzdist mailing list > > Tzdist(a)ietf.org > > https://www.ietf.org/mailman/listinfo/tzdist > -- Daniel Migault Orange Labs -- Security +33 6 70 72 69 58

1 0

Time Zone Distribution Protocol WG Milestones (tzdist)
by Daniel Migault Aug. 13, 2014

Aug. 13, 2014

Hi, We are getting under way with the tzdist WG <http://datatracker.ietf.org/wg/tzdist/charter/> [1]. This IETF working group (WG) goal is to define time zone data distribution protocols. If you want to join please the tzdist WG <https://www.ietf.org/mailman/listinfo/tzdist> [2]. These are the milestones we propose. If you have any comments, please feel free to post them before Tuesday August 19 so we can consider them. Milestones: - September 2014 Adopt time zone data distribution protocol draft as WG document - September 2014 Adopt an CalDAV extension draft as WG document - January 2015 Submit time zone data distribution protocol draft for WGLC - March 2015 Submit time zone data distribution protocol draft as Standards Track RFC - March 2015 Submit CalDAV extension draft for WGLC - June 2015 Submit CalDAV extension draft as Standards Track RFC - September 2016 re-charter the WG [1] http://datatracker.ietf.org/wg/tzdist/charter/ [2] https://www.ietf.org/mailman/listinfo/tzdist -- Daniel Migault Orange Labs -- Security +33 6 70 72 69 58

1 0

[PATCH] * time2posix.3, tzfile.5: Use ASCII double-quote only for C strings.
by Paul Eggert Aug. 12, 2014

Aug. 12, 2014

--- time2posix.3 | 5 ++--- tzfile.5 | 12 ++++++------ 2 files changed, 8 insertions(+), 9 deletions(-) diff --git a/time2posix.3 b/time2posix.3 index d689608..75d8097 100644 --- a/time2posix.3 +++ b/time2posix.3 @@ -28,9 +28,8 @@ time2posix, posix2time \- convert seconds since the Epoch .. IEEE Standard 1003.1 (POSIX) -legislates that a time_t value of -536457599 shall correspond to "Wed Dec 31 23:59:59 UTC 1986." -This effectively implies that POSIX time_t's cannot include leap +requires the time_t value 536457599 to stand for 1986-12-31 23:59:59 UTC. +This effectively implies that POSIX time_t values cannot include leap seconds and, therefore, that the system time must be adjusted as each leap occurs. diff --git a/tzfile.5 b/tzfile.5 index 182cc43..1e67a64 100644 --- a/tzfile.5 +++ b/tzfile.5 @@ -13,7 +13,7 @@ followed by a character identifying the version of the file's format (as of 2013, either an ASCII NUL, or '2', or '3') followed by fifteen bytes containing zeroes reserved for future use, followed by six four-byte integer values -written in a "standard" byte order +written in a standard byte order (the high-order byte of the value is written first). These values are, in order: @@ -28,28 +28,28 @@ The number of standard/wall indicators stored in the file. The number of leap seconds for which data are stored in the file. .TP .I tzh_timecnt -The number of "transition times" for which data are stored +The number of transition times for which data are stored in the file. .TP .I tzh_typecnt -The number of "local time types" for which data are stored +The number of local time types for which data are stored in the file (must not be zero). .TP .I tzh_charcnt -The number of characters of "time zone abbreviation strings" +The number of characters of time zone abbreviation strings stored in the file. .PP The above header is followed by .I tzh_timecnt four-byte signed integer values sorted in ascending order. -These values are written in "standard" byte order. +These values are written in standard byte order. Each is used as a transition time (as returned by .IR time (2)) at which the rules for computing local time change. Next come .I tzh_timecnt one-byte unsigned integer values; -each one tells which of the different types of "local time" types +each one tells which of the different types of local time types described in the file is associated with the same-indexed transition time. These values serve as indices into an array of .I ttinfo -- 1.9.1

1 0