Branch: Tag:


2014-10-16 20:27:46 by Henrik Grubbström (Grubba) <>

Updated to tzdata2014h.

1:   News for the tz database    + Release 2014h - 2014-09-25 18:59:03 -0700 +  +  Changes affecting past time stamps +  +  America/Jamaica's 1974 spring-forward transition was Jan. 6, not Apr. 28. +  +  Shanks says Asia/Novokuznetsk switched from LMT (not "NMT") on 1924-05-01, +  not 1920-01-06. The old entry was based on a misinterpretation of Shanks. +  +  Some more zones have been turned into links, when they differed +  from existing zones only for older time stamps. As usual, +  these changes affect UTC offsets in pre-1970 time stamps only. +  Their old contents have been moved to the 'backzone' file. +  The affected zones are: Africa/Blantyre, Africa/Bujumbura, +  Africa/Gaborone, Africa/Harare, Africa/Kigali, Africa/Lubumbashi, +  Africa/Lusaka, Africa/Maseru, and Africa/Mbabane. +  +  Changes affecting code +  +  zdump -V and -v now output gmtoff= values on all platforms, +  not merely on platforms defining TM_GMTOFF. +  +  The tz library's localtime and mktime functions now set tzname to a value +  appropriate for the requested time stamp, and zdump now uses this +  on platforms not defining TM_ZONE, fixing a 2014g regression. +  (Thanks to Tim Parenti for reporting the problem.) +  +  The tz library no longer sets tzname if localtime or mktime fails. +  +  zdump -c no longer mishandles transitions near year boundaries. +  (Thanks to Tim Parenti for reporting the problem.) +  +  An access to uninitalized data has been fixed. +  (Thanks to Jörg Richter for reporting the problem.) +  +  When THREAD_SAFE is defined, the code ports to the C11 memory model. +  A memory leak has been fixed if ALL_STATE and THREAD_SAFE are defined +  and two threads race to initialize data used by gmtime-like functions. +  (Thanks to Andy Heninger for reporting the problems.) +  +  Changes affecting build procedure +  +  'make check' now checks better for properly-sorted data. +  +  Changes affecting documentation and commentary +  +  zdump's gmtoff=N output is now documented, and its isdst=D output +  is now documented to possibly output D values other than 0 or 1. +  +  zdump -c's treatment of years is now documented to use the +  Gregorian calendar and Universal Time without leap seconds, +  and its behavior at cutoff boundaries is now documented better. +  (Thanks to Arthur David Olson and Tim Parenti for reporting the problems.) +  +  Programs are now documented to use the proleptic Gregorian calendar. +  (Thanks to Alan Barrett for the suggestion.) +  +  Fractional-second GMT offsets have been documented for civil time +  in 19th-century Chennai, Jakarta, and New York. +  +    Release 2014g - 2014-08-28 12:31:23 -0700       Changes affecting future time stamps