- tz - lists.iana.org

[PROPOSED 1/5] New Makefile macro BACKWARD
by Paul Eggert May 23, 2017

May 23, 2017

* NEWS: Document this. * Makefile (BACKWARD): New macro. (YDATA, INSTALLARGS): Use it. --- Makefile | 10 +++++++++- NEWS | 4 ++++ 2 files changed, 13 insertions(+), 1 deletion(-) diff --git a/Makefile b/Makefile index c62d46f..6373920 100644 --- a/Makefile +++ b/Makefile @@ -92,6 +92,13 @@ LIBDIR= $(TOPDIR)/lib REDO= posix_right +# For backward-compatibility links for old zone names, use +# BACKWARD= backward pacificnew +# To omit these links, use +# BACKWARD= + +BACKWARD= backward pacificnew + # If you want out-of-scope and often-wrong data from the file 'backzone', use # PACKRATDATA= backzone # To omit this data, use @@ -384,7 +391,7 @@ WEB_PAGES= tz-art.htm tz-how-to.html tz-link.htm DOCS= $(MANS) date.1 $(MANTXTS) $(WEB_PAGES) PRIMARY_YDATA= africa antarctica asia australasia \ europe northamerica southamerica -YDATA= $(PRIMARY_YDATA) pacificnew etcetera backward +YDATA= $(PRIMARY_YDATA) etcetera $(BACKWARD) NDATA= systemv factory TDATA= $(YDATA) $(NDATA) ZONETABLES= zone1970.tab zone.tab @@ -482,6 +489,7 @@ leapseconds: $(LEAP_DEPS) # Arguments to pass to submakes of install_data. # They can be overridden by later submake arguments. INSTALLARGS = \ + BACKWARD=$(BACKWARD) \ DESTDIR=$(DESTDIR) \ LEAPSECONDS='$(LEAPSECONDS)' \ PACKRATDATA='$(PACKRATDATA)' \ diff --git a/NEWS b/NEWS index 4e8091b..dcd49eb 100644 --- a/NEWS +++ b/NEWS @@ -20,6 +20,10 @@ Unreleased, experimental changes in civil timekeeping. (Thanks to Robert Elz and Bradley White for noticing glitches in the code that uncovered this problem.) + 'make BACKWARD=' now suppresses backward-compatibility names + like 'US/Pacific' that are defined in the 'backward' and + 'pacificnew' files. + Several minor changes have been made to the code to make it a bit easier to port to MS-Windows. (Thanks to Kees Dekker for reporting the problems.) -- 2.9.4

1 4

Fwd: input needed on creation of a new sub-package for raw zone data
by Patsy Franklin May 23, 2017

May 23, 2017

Sorry, should have replied to all on this one. ---------- Forwarded message ---------- From: Patsy Franklin <pfrankli(a)redhat.com> Date: Tue, May 23, 2017 at 9:26 AM Subject: Re: [tz] input needed on creation of a new sub-package for raw zone data To: Brian.Inglis(a)systematicsw.ab.ca Hi Brian, Thanks for the feedback! On Mon, May 22, 2017 at 1:54 PM, Brian Inglis <Brian.Inglis(a)systematicsw.ab. ca> wrote: > On 2017-05-22 07:03, Patsy Franklin wrote: > > I'd appreciate community input on this plan. Ideally if others decide > > to also ship a similar subpackage we can use a common naming > > convention and install directory. > > > > We are planning to ship a new subpackage for users who want to have > > access to the raw zone data files e.g. leapseconds, and in a > > pre-determined install location e.g. /usr/share/zoneinfo/zonedata/. > > How about using /usr/share/tzdata{,-src}/ maybe with subdirectories > tzdata-2017b, etc. > > We really want to avoid using src to prevent any confusion between the full src rpm that we ship and this subset of the zone data. > > The raw zone data is useful for users designing their own interfaces > > on top of this data. The broadest flexibility is offered by the raw > > zone data, and while the compiled binary data is versioned and more > > stable, some users have expressed a request for the raw zone data. > > The understanding is that the raw zone data format may change. > > > > In summary: > > Our current plan is to use -zonedata in naming the subpackage, For > > example, tzdata-zonedata-2017b-1.el7. > > How about tzdata-src-2017b...? > > We already provide tzdata-2017b-1.el7.src.rpm which contains the full set of sources. The location of these sources is defined by the installer. > We plan to install the files in this subpackage under > > /usr/share/zoneinfo/zonedata/. > > Adding zonedata does not add any useful information about the contents > and storing under /usr/share/zoneinfo/ could be confusing as there are > already subdirectories posix and right holding copies of standard and > leapsecond compensated binary data files, so TZ=posix/... and > TZ=right/... are valid zones, and users may expect something similar to > happen using TZ=zonedata/... which would be unfortunate. > > That's a good point. I hadn't considered that case. We chose zonedata to clarify that it was only the zone data - not zic, zdump, binaries, etc. It was requested that we provide this subset of files in a consistent location at install time which is why we grouped it under the zoneinfo directory. Is there a more appropriate place to install it? Do you have any other suggestions for a name since I think src is ambiguous for our purposes? > Just as an example we would ship the following files: > > LICENSE > > version > > africa > > antarctica > > asia > > australasia > > europe > > northamerica > > southamerica > > pacificnew > > etcetera > > backward > > systemv > > factory > > backzone > > iso3166.tab > > leapseconds > > leap-seconds.list > > zone1970.tab > > zone.tab > > README, CONTRIBUTING, NEWS, and Theory should be included, also > tz-how-to.html which documents how to define and use the sources. > leap-seconds.list should be a (symbolic?) link to the canonical file > version leap-seconds.<timestamp> e.g. leap-seconds.3692908800. > We ship README and Theory with the base package. CONTRIBUTING, NEWS and tz-how-to.html is shipped with our src package. Do we still need to duplicate these in the optional subpackages? BTW, I forgot to mention that our spec file will not allow the new subpackage to be installed without the corresponding version of the base package. The base package does not require the optional new subpackage but will check to insure that it is in sync if it is installed. Do we still need to version the leap-seconds file if it is tied to a specific tzdata version? Thanks! Patsy > > -- > Take care. Thanks, Brian Inglis, Calgary, Alberta, Canada >

2 1

Looking for a way to cut down binary zoneinfo size. Does anyone have the way to?
by Viktor Sergiienko May 19, 2017

May 19, 2017

Hello, I'm developing for a pretty small-sized embedded system, and 2.9M of binaries is a bit too much for us. The system never uses a date before 2007. I estimate I can slim it down too 500K by cutting down the old data. 1. I made a script to cut down the source data: all the rules and all the zonelines before 2001. Compiling this data, I get a lot of errors like: > "northamerica", line 2537: %s in ruleless zone > "northamerica", line 2618: unruly zone because some zones are left without rules, but with %s format in them. 2. I tried to modify zic itself, but apparently understanding it will take me days. Just setting min_year to 2001 results in bunch of > can't determine time zone abbreviation to use just after until time Does anyone have a ready solution? Thanks!

7 11

[PROPOSED 1/2] Rename ‘lint’ to ‘GCC_LINT’
by Paul Eggert May 18, 2017

May 18, 2017

* Makefile (GCC_DEBUG_FLAGS): * private.h (INITIALIZE): Rename ‘lint’ to ‘GCC_LINT’, as the macro is intended for GCC, not for traditional ‘lint’. --- Makefile | 2 +- private.h | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/Makefile b/Makefile index e7c6ac0..bcfdcb2 100644 --- a/Makefile +++ b/Makefile @@ -159,7 +159,7 @@ LDLIBS= # (or some other number) to set the maximum time zone abbreviation length # that zic will accept without a warning (the default is 6) # $(GCC_DEBUG_FLAGS) if you are using recent GCC and want lots of checking -GCC_DEBUG_FLAGS = -Dlint -g3 -O3 -fno-common -fstrict-aliasing \ +GCC_DEBUG_FLAGS = -DGCC_LINT -g3 -O3 -fno-common -fstrict-aliasing \ -Wall -Wextra \ -Wbad-function-cast -Wcast-align -Wdate-time \ -Wdeclaration-after-statement \ diff --git a/private.h b/private.h index e9eb211..6886185 100644 --- a/private.h +++ b/private.h @@ -566,7 +566,7 @@ static time_t const time_t_max = MAXVAL(time_t, TYPE_BIT(time_t)); ** INITIALIZE(x) */ -#ifdef lint +#ifdef GCC_LINT # define INITIALIZE(x) ((x) = 0) #else # define INITIALIZE(x) -- 2.9.4

1 1

Morocco changes May 27th 2017 instead of May 21th
by Info - avisched.com May 16, 2017

May 16, 2017

Morocco is about to switch on May 27th back to UTC from +1. The current release doesn't reflect that (May 21st). Is there any update planned before the actual transition? Many thanks! Martin

5 5

Asia/Harbin, Asia/Chongqing and Asia/Kashgar
by Phake Nick May 13, 2017

May 13, 2017

iirc, according to past conversation on the mailing list, the reason why these timezones that are not in use for many years already still exists in tz database is because of legacy reason, however given that Asia/Ishigaki were deleted from this database in 1999 when it's found to be not in use, is there any reason preventing the same to be done to Asia/Harbin, Asia/Chongqing, and Asia/Kashgar?

2 1

gcc 8 demanded patch (from NetBSD)
by Robert Elz May 11, 2017

May 11, 2017

The following patch was recently added to NetBSD, to appease GCC 8 (apparently) ... "hit" is a bool, and it seems gcc 8 does not like ++ operators applied to one of those. (deleted parts of the diff which just showed the version ID string being updated...) kre Index: src/lib/libc/time/localtime.c diff -u src/lib/libc/time/localtime.c:1.106 src/lib/libc/time/localtime.c:1.107 --- src/lib/libc/time/localtime.c:1.106 Sat Mar 11 18:23:14 2017 +++ src/lib/libc/time/localtime.c Tue May 9 02:30:49 2017 @@ -1716,7 +1716,7 @@ timesub(const time_t *timep, int_fast32_ sp->lsis[i - 1].ls_trans + 1 && sp->lsis[i].ls_corr == sp->lsis[i - 1].ls_corr + 1) { - ++hit; + hit = true; --i; } }

8 13

TZDEFRULESTRING is increasingly out-of-date
by Tom Lane May 7, 2017

May 7, 2017

Due to a script's unintended failure to supply a -p option to zic, I recently had my nose rubbed in the fact that localtime.c still has this as the fallback when the "posixrules" link can't be loaded: /* ** The DST rules to use if TZ has no rules and we can't load TZDEFRULES. ** We default to US rules as of 1999-08-17. ** POSIX 1003.1 section 8.1.1 says that the default DST rules are ** implementation dependent; for historical reasons, US rules are a ** common default. */ #ifndef TZDEFRULESTRING #define TZDEFRULESTRING ",M4.1.0,M10.5.0" #endif /* !defined TZDEFDST */ It's now over ten years since the US last followed those rules; isn't it time to update TZDEFRULESTRING to match current law? I can't really think of any situation where the current default would be of any value now. Apologies if this policy is being followed intentionally; but I couldn't find any discussion of the point by googling. regards, tom lane

2 1

Potted history of timekeeping from the BBC
by John Haxby April 21, 2017

April 21, 2017

This is from a series of things that made the modern economy, it says on the web page: Tick tock: The importance of knowing the right time: http://www.bbc.co.uk/news/business-39129620 In one short page it manages to mention everything but DST :) and it includes a picture of Harrison No 1. (If you're ever near Greenwich I recommend visiting the Royal Observatory, taking the obligatory foot-in-both-hemispheres selfie and then going to look at the clocks.) >From the BBC web page though, I particularly noticed this: > Some financiers recently calculated it was worth spending $300m (£247m) drilling through mountains between Chicago and New York to lay fibre-optic cables in a slightly straighter line. That sped up communication between the two cities' exchanges by three milliseconds. jch

10 10

Re: [tz] leapseconds.awk, LEAP directives and references
by Chris Woodbury April 21, 2017

April 21, 2017

In the course of reviewing the AWK script leapseconds.awk, I was disturbed by the fact that it (1) operated on the comments (probably not a good idea) in the NIST produced leap-seconds.list file rather than the NTP timestamp itself and (2) that it assumed that leap seconds would only occur in December and June. The regulation that governs leap seconds, ITU-R TF.460-6 <https://www.itu.int/rec/R-REC-TF.460-6-200202-I/>, actually allows leap seconds to occur at the end of any UTC month. Annex 1, Section 2: Leap Seconds 2.1 A positive or negative leap-second should be the last second of a UTC month, but first preference should be given to the end of December and June, and second preference to the end of March and September. I note, in the event of a (highly unlikely) negative leap second, the LEAP directive generated is for 23:59:59 which, in the case of a negative leap second, doesn't exist. I reviewed the C code and, it seems to me, the transaction added should actually be for the second before, 23:59:58, as the transaction itself causes 23:59:59 to be skippedaltogether. (Kinda hard to skip an event if you're in it.) I bring this up as I've alrady written the changes necessary but I want to verify the correctness of this last beforesubmitting them. Also, I've an AWK script that uses the IERS file Leap_Seconds.dat rather than the NIST leap-seconds.listfile. Leap_Seconds.dat is a straight list of leap-seconds. USNO MAIA also has straight list of leap-seconds in a different format. See URLs: <http://hpiers.obspm.fr/iers/bul/bulc/Leap_Seconds.dat> <http://maia.usno.navy.mil/ser7/leapsec.dat> Either of these would, particularly the IERS file as it has few but excellent comments, be a better source of leap-second information than the once removed NIST file leap-seconds.list; which is purposed for another task. Lastly, thanks to Dr. Eggert for including the reference I gave him to the AAPT Resource Letter that included Terry Quinn's article in the NIST Time and Frequency Division's PDF archive (1168). Anyone wanting to know about Time and Frequency Measurement might want to read the other articles as well as Dr. Quinn's. Another good, comprehensive article is "The leap second: its history and possible future" by R.A. Nelson, et al, from Metrogia Vol 38, pp. 509-529 (2001). PDF available (with permission) at: <https://www.cl.cam.ac.uk/~mgk25/time/metrologia-leapsecond.pdf>

9 16