Re: tz leap second bug encountered by sendmail 8.8.8 in Australia
The basic problem is that tz code looks at the system file .../zoneinfo/GMT to decide what leap seconds to use when computing GMT; this leads to bogus results if the user wants leap seconds and the system file doesn't (or vice versa).
Might I proffer that this is not a problem with the tz code, but a truthful representation of the situation presented to it. If "the user wants leap seconds and the system file doesn't (or vice versa)," the user is just kidding themselves. Either the system clock ticks leap seconds (a stationary epoch), or it doesn't. All tz files used on the system should reflect that reality. The suggested gmtoff() algorithm will always produce such varying results if localtime() and gmtime() cannot agree on minute boundaries, whether the discrepancy is a leapsecond artifact or not. Bradley
Date: Sat, 23 May 1998 00:25:34 -0400 (EDT) From: Bradley White <bww@fore.com> If "the user wants leap seconds and the system file doesn't (or vice versa)," the user is just kidding themselves. Either the system clock ticks leap seconds (a stationary epoch), or it doesn't. All tz files used on the system should reflect that reality. That's good advice, but unfortunately the current tz package does not follow it, and this apparently causes problems in practice. By default the current tz distribution generates both "posix" and "right" subdirectories, with the default being "posix". It's natural to assume that if all processes set the TZ environment variable to "right/Australia/Adelaide", then the system will have leap second support with Adelaide localtime. Unfortunately this doesn't work as expected, as it botches the localtime-gmtime offset. One way to fix the problem is to remove the `posix' and `right' subdirectories; this will make the configuration error less likely. Also, the "GMT" link should be moved out of the "backward" file and into the "etcetera" file, as the "GMT" link is currently essential for proper leap second support. (I didn't know this when I originally moved it to the "backward" file -- sorry about that.) I enclose a proposed patch below. The Makefile change might be a bit controversial, as it makes it harder to enable leap second support, but the other two changes should be noncontroversial given the current state of the code. If these changes are acceptable, then it should be a win for gmtime to look at the local time file's leap second info instead of looking at the "GMT" file's leap second info, since the leap second info of the two files ought to agree, and looking at just one file means one less file to read. But this is only a performance issue; it shouldn't affect correctness. =================================================================== RCS file: RCS/etcetera,v retrieving revision 1997.11 retrieving revision 1997.11.1.1 diff -u -r1997.11 -r1997.11.1.1 --- etcetera 1997/12/29 14:31:50 1997.11 +++ etcetera 1998/05/24 07:28:25 1997.11.1.1 @@ -10,6 +10,10 @@ Zone Etc/UTC 0 - UTC Zone Etc/UCT 0 - UCT +# The following link is an important special case, +# as functions like gmtime load the "GMT" file to handle leap seconds properly. +Link Etc/GMT GMT + Link Etc/UTC Etc/Universal Link Etc/UTC Etc/Zulu =================================================================== RCS file: RCS/backward,v retrieving revision 1997.10 retrieving revision 1997.10.1.1 diff -u -r1997.10 -r1997.10.1.1 --- backward 1997/12/11 17:44:35 1997.10 +++ backward 1998/05/24 07:28:25 1997.10.1.1 @@ -38,7 +38,6 @@ Link Africa/Cairo Egypt Link Europe/Dublin Eire Link Europe/London GB -Link Etc/GMT GMT Link Etc/GMT+0 GMT+0 Link Etc/GMT-0 GMT-0 Link Etc/GMT0 GMT0 =================================================================== RCS file: RCS/Makefile,v retrieving revision 1998.3 retrieving revision 1998.3.1.1 diff -u -r1998.3 -r1998.3.1.1 --- Makefile 1998/02/28 17:12:55 1998.3 +++ Makefile 1998/05/24 07:28:25 1998.3.1.1 @@ -58,21 +58,16 @@ # If you always want time values interpreted as "seconds since the epoch # (not counting leap seconds)", use -# REDO= posix_only +# REDO= posix # below. If you always want right time values interpreted as "seconds since # the epoch" (counting leap seconds)", use -# REDO= right_only -# below. If you want both sets of data available, with leap seconds not -# counted normally, use -# REDO= posix_right -# below. If you want both sets of data available, with leap seconds counted -# normally, use -# REDO= right_posix +# REDO= right # below. # POSIX mandates that leap seconds not be counted; for compatibility with it, -# use either "posix_only" or "posix_right". +# use "posix". Do not use both "posix" and "right" files simultaneously, +# as this causes either localtime or gmtime to yield incorrect results. -REDO= posix_right +REDO= posix # Since "." may not be in PATH... @@ -287,21 +282,12 @@ cp yearistype.sh yearistype chmod +x yearistype -posix_only: zic $(TDATA) +posix: zic $(TDATA) $(ZIC) -y $(YEARISTYPE) -d $(TZDIR) -L /dev/null $(TDATA) -right_only: zic leapseconds $(TDATA) +right: zic leapseconds $(TDATA) $(ZIC) -y $(YEARISTYPE) -d $(TZDIR) -L leapseconds $(TDATA) -other_two: zic leapseconds $(TDATA) - $(ZIC) -y $(YEARISTYPE) -d $(TZDIR)/posix -L /dev/null $(TDATA) - $(ZIC) -y $(YEARISTYPE) \ - -d $(TZDIR)/right -L leapseconds $(TDATA) - -posix_right: posix_only other_two - -right_posix: right_only other_two - zones: $(REDO) $(TZLIB): $(LIBOBJS)
Since we're probably going to change the tz code anyway to address the posix/right problem, I cobbled together all the other proposed tz changes that I've been saving up. Here they are: * We now have info for Davis Station, Antarctica. * Ephraim Silverberg forwarded a brief note about time in Palestine. And I added the country code PS for Palestine, jumping the gun a bit; this should help simplify time zone setup in Israel and in areas administered by the Palestinian Authority. I added commentary to help explain this. * I updated an obsolete link in WWW.htm. * Peter Ilieve and Joseph Myers updated the official info about time in Britain. * At the suggestion of someone who's coordinating this data with ISO country codes, the Clipperton and Midway entries belong in the australasia file, as this munges political boundaries less. Clipperton is administered from French Polynesia, and Midway is one of the US Minor Outlying Islands. * Similarly, I amalgamated the commentary for the Scattered Islands (Iles Eparses). =================================================================== RCS file: RCS/australasia,v retrieving revision 1998.4 retrieving revision 1998.4.1.1 diff -u -r1998.4 -r1998.4.1.1 --- australasia 1998/04/06 17:47:07 1998.4 +++ australasia 1998/05/24 07:28:25 1998.4.1.1 @@ -207,6 +207,8 @@ -9:30 - MART # Marquesas Time Zone Pacific/Tahiti -9:58:16 - LMT 1912 Oct # Papeete -10:00 - TAHT # Tahiti Time +# Clipperton (near North America) is administered from French Polynesia; +# it is uninhabited. # Guam # Zone NAME GMTOFF RULES FORMAT [UNTIL] @@ -402,6 +404,12 @@ # Kingman # uninhabited +# Midway +Zone Pacific/Midway -11:49:28 - LMT 1901 + -11:00 - NST 1967 Apr # N=Nome + -11:00 - BST 1983 Nov 30 # B=Bering + -11:00 - SST # S=Samoa + # Palmyra # uninhabited since World War II; was probably like Pacific/Kiritimati =================================================================== RCS file: RCS/europe,v retrieving revision 1998.4 retrieving revision 1998.4.1.1 diff -u -r1998.4 -r1998.4.1.1 --- europe 1998/04/23 15:37:16 1998.4 +++ europe 1998/05/24 07:28:25 1998.4.1.1 @@ -147,7 +147,7 @@ # time of sunrise and sunset in The Times, when BDST was in effect, and # if you find a zone reference it will say, "All times B.D.S.T." -# From Peter Ilieve <peter@aldie.co.uk> (1998-01-12): +# From Peter Ilieve <peter@aldie.co.uk> (1998-04-19): # The following list attempts to show the complete history of Summer Time # legislation in the United Kingdom, and has quite a bit to say about # the Isle of Man and the Channel Islands as well. @@ -200,7 +200,7 @@ # seen it I don't know if it used Greenwich mean time, some other definition, # or just said that Isle of Man time would be the same as in Great Britain. # -# - The Isle of Man (War Legislation) Act, 1914 (4 & 5 Geo. 5. c. 62) (???) +# - The Isle of Man (War Legislation) Act, 1914 (4 & 5 Geo. 5. c. 62) # # Gives the power, by Order in Council, to extend wartime legislation # to the Isle of Man. @@ -258,6 +258,14 @@ # # The matching Isle of Man Order for 1918 with the same dates and times. # +# - The Termination of the Present War (Definition) Act, 1918 +# (8 & 9 Geo. 5. c. 59) +# +# This gave power to specify a legal end date for the war just ended, +# which would affect things like the Summer Time Act, 1916, which applied +# only in wartime. This date was to be close to the date of formal +# ratification of the treaty or treaties of peace. +# # - S.R.&O. 1919, No. 297 # # An Order made under the Summer Time Act, 1916 giving dates @@ -285,12 +293,12 @@ # Monday 25 October (the time remaining 02:00 GMT). The 1989 Green # Paper (Cm 722) says this was done because of a coal strike. # -# - The War Emergency Laws (Continuance) Act, 1920 (10 & 11 Geo. 5. c. 5) (???) +# - The War Emergency Laws (Continuance) Act, 1920 (10 Geo. 5. c. 5) # # This extends the power to make Orders under the Summer Time Act, 1916 # for a period of 12 months after the termination of the war. -# I haven't seen this one so I don't know when it came into force, or -# when the law deemed the termination of the war to have been. +# Came into force on 31 March 1920. Although the war had been over for more +# than 12 months by then the legal end date had not yet been set. # # - S.R.&O. 1921, No. 363 # @@ -302,11 +310,16 @@ # # The matching Isle of Man Order for 1921 with the same dates and times. # -# - S.R.&O. 1922, No. 264 (???) +# - S.R.&O. 1922, No. 264 # -# This probably defines Summer Time for 1922 as the Summer Time Act, 1922 -# was passed after the start date. Dates from Cm 722: -# 1922: 26 March to 8 October +# An Order made under the Summer Time Act, 1916 and the War +# Emergency Laws (Continuance) Act, 1920 giving dates for Summer Time +# in 1921 of Sunday 26 March to Sunday 8 October, both at 02:00 GMT. +# It also mentions the arrangements for defining the legal end date +# for the late war. An Order was made on 10 August 1921, under the +# Termination of the Present War (Definition) Act, 1918, setting +# a date of 31 August 1921. This means the powers of the Summer Time +# Act, 1916 would finally expire on 31 August 1922. # # - S.R.&O. 1922, No. 290 (???) # @@ -327,7 +340,7 @@ # dated 13 October 1922. It revokes (among other things) the Order extending # the Summer Time Act, 1916 to the Isle of Man. # -# - The Expiring Laws Continuance Act, 1923 (13 & 14 Geo. 5. c. 37) (jsm) +# - The Expiring Laws Continuance Act, 1923 (13 & 14 Geo. 5. c. 37) # # This extended the Summer Time Act, 1922 (among other things) until # 31 December 1924. @@ -359,15 +372,16 @@ # These were made under the Emergency Powers (Defence) Act, 1939. # They change the end date to be the day after the third Saturday in November. # -# - S.R.&O. 1940, No. 1883 +# - S.R.&O. 1940, No. 172 # # An Order in Council amending the Defence (Summer Time) Regulations, 1939. +# It changed the start date to the day after the fourth Saturday in February +# (ie. 25 Feb 1940). +# +# - S.R.&O. 1940, No. 1883 +# +# Another Order amending the Defence (Summer Time) Regulations, 1939. # This continues summer time throughout the year after it starts in 1940. -# There was another Order (S.R.&O. 1940, No. 172) that I assume had merely -# changed the dates, to start on 4 February, and was then superseded by this -# one. I haven't seen No. 172 so I don't know what end date it would have -# specified. The dates from Cm 722: -# 1940: Summer Time starts on 4 February # # - S.R.&O. 1941, No. 476 # @@ -385,15 +399,20 @@ # # - S.R.&O. 1944, No. 932 # -# This changed the end date of Double Summer Time to 17 September 1944. -# (I don't have the text of this, just a note of what it did, the text almost -# certainly had the `day after the nth Saturday' form.) -# -# - S.R.&O. 1945, No. 312 (???) -# -# Probably defines the dates of Double Summer Time in 1945. -# I do have this quote from Hansard (the official record of the United Kingdom -# Parliament), Oral Answers, 1 March 1945, cols 1559--60: +# Another Order amending the Defence (Summer Time) Regulations, 1939. +# This changed the end date of Double Summer Time to the day after the +# third Saturday in September (ie. 17 September 1944). +# +# - S.R.&O. 1945, No. 312 +# +# Another Order amending the Defence (Summer Time) Regulations, 1939. +# This changes the start and end dates of Double Summer Time to the +# day after the first Sunday in April and the day after the second Saturday +# in July (ie. Mon 2 April to Sun 15 July). +# +# I have this quote from Hansard (the official record of the United Kingdom +# Parliament), Oral Answers, 1 March 1945, cols 1559--60, explaining the +# unusual start on a Monday: # # `58. Major Sir Goronwy Owen asked the Secretary of State for the Home # Department if he is now able to state the Government's proposals @@ -410,15 +429,13 @@ # Sunday, but in the night of Sunday-Monday so that it will operate from # Monday, 2nd April.' # -# Cm 722 gives dates of: -# 1945: Double Summer time: 2 April to 15 July -# 1945: Summer Time ends on 7 October -# # - S.R.&O. 1945, No. 1208 # # An Order under the Emergency Powers (Defence) Acts, 1939 and 1940 revoking # a long list of things, including the Defence (Summer Time) Regulations, 1939. # This meant that Summer Time reverted to being set by the 1922 and 1925 Acts. +# It was made on 28 September, early enough to end Summer Time on the +# date defined by the 1925 Act: 7 October. # # - The Summer Time Act, 1947 (10 & 11 Geo. 6. c. 16) # @@ -431,21 +448,36 @@ # both to vary the dates and to continue Double Summer Time. It applied # to the Isle of Man and the Channel Islands. # -# - Summer Time Order, 1948 (S.I. 1948/495) (???) -# - Summer Time Order, 1949 (S.I. 1949/373) (???) -# - Summer Time Order, 1950 (S.I. 1950/518) (???) -# - Summer Time Order, 1951 (S.I. 1951/430) (???) -# - Summer Time Order, 1952 (S.I. 1952/451) (???) -# -# These presumably give the dates of Summer Time for the relevent years. -# There was no Double Summer Time. The dates given in the 1989 Green Paper -# for these years are: -# 1948: 14 March to 31 October -# 1949: 3 April to 30 October -# 1950: 16 April to 22 October -# 1951: 15 April to 21 October -# 1952: 20 April to 26 October -# After 1952 things revert back to the 1922 and 1925 Acts. +# - Summer Time Order, 1948 (S.I. 1948/495) +# +# An Order made under the Summer Time Act, 1947. Gave dates for 1948 of +# 14 March and 31 October, both at 02:00 GMT. +# +# Although the 1947 Act had legislated for Double Summer Time, this was +# not continued after 1947. +# +# - Summer Time Order, 1949 (S.I. 1949/373) +# +# Another Order made under the Summer Time Act, 1947. Gave dates for 1949 +# of 3 April and 30 October, both at 02:00 GMT. +# +# - Summer Time Order, 1950 (S.I. 1950/518) +# +# Another Order made under the Summer Time Act, 1947. Gave dates for 1950 +# of 16 April and 22 October, both at 02:00 GMT. +# +# - Summer Time Order, 1951 (S.I. 1951/430) +# +# Another Order made under the Summer Time Act, 1947. Gave dates for 1951 +# of 15 April and 21 October, both at 02:00 GMT. +# +# - Summer Time Order, 1952 (S.I. 1952/451) +# +# Another Order made under the Summer Time Act, 1947. Gave dates for 1952 +# of 20 April and 26 October, both at 02:00 GMT. +# +# This is the last of this run of Orders, so for 1953 things reverted +# to the 1922 and 1925 Acts. # # - The Interpretation Act (Northern Ireland), 1954 (1954 c. 33 (N.I.)) (???) # @@ -659,7 +691,9 @@ # 2000: 26 March to 29 October # 2001: 25 March to 28 October # +# <a href="http://www.hmso.gov.uk/si/si1997/97298201.htm"> # - Summer Time Order 1997 (S.I. 1997/2982) +# </a> # # Implements the eighth Directive. Has the same text about the Isle of Man, # Guernsey and Jersey as the 1994 Order. @@ -757,7 +791,7 @@ # S.R.&O. 1939, No. 1379 Rule GB-Eire 1939 only - Nov Sun>=16 2:00s 0 GMT # S.R.&O. 1940, No. 172 and No. 1883 -Rule GB-Eire 1940 only - Feb 25 2:00s 1:00 BST +Rule GB-Eire 1940 only - Feb Sun>=23 2:00s 1:00 BST # S.R.&O. 1941, No. 476 Rule GB-Eire 1941 only - May Sun>=2 1:00s 2:00 BDST Rule GB-Eire 1941 1943 - Aug Sun>=9 1:00s 1:00 BST @@ -766,9 +800,9 @@ # S.R.&O. 1944, No. 932 Rule GB-Eire 1944 only - Sep Sun>=16 1:00s 1:00 BST # S.R.&O. 1945, No. 312 -Rule GB-Eire 1945 only - Apr 2 1:00s 2:00 BDST +Rule GB-Eire 1945 only - Apr Mon>=2 1:00s 2:00 BDST +Rule GB-Eire 1945 only - Jul Sun>=9 1:00s 1:00 BST # S.R.&O. 1945, No. 1208 -Rule GB-Eire 1945 only - Jul 15 1:00s 1:00 BST Rule GB-Eire 1945 1946 - Oct Sun>=2 2:00s 0 GMT Rule GB-Eire 1946 only - Apr Sun>=9 2:00s 1:00 BST # The Summer Time Act, 1947 @@ -778,9 +812,10 @@ Rule GB-Eire 1947 only - Nov 2 2:00s 0 GMT # Summer Time Order, 1948 (S.I. 1948/495) Rule GB-Eire 1948 only - Mar 14 2:00s 1:00 BST -Rule GB-Eire 1948 1949 - Oct lastSun 2:00s 0 GMT +Rule GB-Eire 1948 only - Oct 31 2:00s 0 GMT # Summer Time Order, 1949 (S.I. 1949/373) Rule GB-Eire 1949 only - Apr 3 2:00s 1:00 BST +Rule GB-Eire 1949 only - Oct 30 2:00s 0 GMT # Summer Time Order, 1950 (S.I. 1950/518) # Summer Time Order, 1951 (S.I. 1951/430) # Summer Time Order, 1952 (S.I. 1952/451) =================================================================== RCS file: RCS/asia,v retrieving revision 1998.4 retrieving revision 1998.4.1.1 diff -u -r1998.4 -r1998.4.1.1 --- asia 1998/05/14 15:58:34 1998.4 +++ asia 1998/05/24 07:28:25 1998.4.1.1 @@ -881,6 +881,14 @@ # to Palestine's rules. If you have more info about this, please # send it to tz@elsie.nci.nih.gov for incorporation into future editions. +# From IINS News Service - Israel - 1998-03-23 10:38:07 Israel time, +# forwarded by Ephraim Silverberg: +# +# Despite the fact that Israel changed over to daylight savings time +# last week, the PLO Authority (PA) has decided not to turn its clocks +# one-hour forward at this time. As a sign of independence from Israeli rule, +# the PA has decided to implement DST in April. + # These rules for Egypt are stolen from the `africa' file. # Rule NAME FROM TO TYPE IN ON AT SAVE LETTER/S Rule EgyptAsia 1957 only - May 10 0:00 1:00 S =================================================================== RCS file: RCS/WWW.htm,v retrieving revision 1998.4 retrieving revision 1998.4.1.1 diff -u -r1998.4 -r1998.4.1.1 --- WWW.htm 1998/05/14 15:58:34 1998.4 +++ WWW.htm 1998/05/24 07:28:25 1998.4.1.1 @@ -57,7 +57,7 @@ <LI><A HREF="http://community.bellcore.com/mbr/gmt-explained.html">A Few Facts Concerning GMT, UT, and the RGO</A> answers questions like ``What is the difference between GMT and UTC?'' -<LI><A HREF="http://www.energy.ca.gov/energy/daylightsaving.html">Daylight +<LI><A HREF="http://www.energy.ca.gov/daylightsaving.html">Daylight Saving Time -- Saving Time, Saving Energy</A> is a history of DST in the US. <LI><A HREF="http://www.yahoo.com/Science/Weights_and_Measures/Measurements/Time/Daylight_Saving_Time/">Yahoo! - Science:Weights and Measures:Measurements:Time:Daylight Saving Time</A> =================================================================== RCS file: RCS/antarctica,v retrieving revision 1998.3 retrieving revision 1998.3.1.1 diff -u -r1998.3 -r1998.3.1.1 --- antarctica 1998/02/28 17:32:25 1998.3 +++ antarctica 1998/05/24 07:28:25 1998.3.1.1 @@ -41,18 +41,24 @@ # # year-round bases # Casey, Bailey Peninsula, since 1969 -# Davis, Vestfold Hills, since 1957-01-13 (except 1965-01 - 1969-02) +# Davis, Vestfold Hills, since 1957-01-13 (except 1964-11 - 1969-02) # Mawson, Holme Bay, since 1954-02-13 # Zone NAME GMTOFF RULES FORMAT [UNTIL] Zone Antarctica/Casey 0 - ___ 1969 8:00 - WST # Western (Aus) Standard Time -#Zone Antartica/Davis unknown +Zone Antarctica/Davis 0 - ___ 1957 Jan 13 + 7:00 - DAVT 1964 Nov # Davis Time + 0 - ___ 1969 Feb + 7:00 - DAVT Zone Antarctica/Mawson 0 - ___ 1954 Feb 13 6:00 - MAWT # Mawson Time # References: # <a href="http://www.antdiv.gov.au/aad/exop/sfo/casey/casey_aws.html"> # Casey Weather (1998-02-26) # </a> +# <a href="http://www.antdiv.gov.au/aad/exop/sfo/davis/video.html"> +# Davis Station, Antarctica (1998-02-26) +# </a> # <a href="http://www.antdiv.gov.au/aad/exop/sfo/mawson/video.html"> # Mawson Station, Antarctica (1998-02-25) # </a> =================================================================== RCS file: RCS/iso3166.tab,v retrieving revision 1.3 retrieving revision 1.3.1.1 diff -u -r1.3 -r1.3.1.1 --- iso3166.tab 1997/08/31 08:13:53 1.3 +++ iso3166.tab 1998/05/24 07:28:25 1.3.1.1 @@ -3,12 +3,16 @@ # From Paul Eggert <eggert@twinsun.com> (1996-09-03): # # This file contains a table with the following columns: -# 1. ISO 3166 2-character country code. +# 1. ISO 3166-1:1997 2-character country code. See: # 2. The usual English name for the country, # chosen so that alphabetic sorting of subsets produces helpful lists. +# This is not the same as the English name in the ISO 3166 tables. # -# For France in Europe, we follow common practice and use FR, -# even though FX might be more technically correct. +# The PS code below has been reserved for Palestine by the ISO; +# however, the code is not yet officially assigned to Palestine. Please see +# <a href="ftp://ftp.informatik.uni-erlangen.de/pub/doc/ISO/ISO-3166-background"> +# Cord Wischhoefer, Country Code Elements for Palestine (1996-07-24) +# </a>. # # Columns are separated by a single tab. # The table is sorted by country code. @@ -191,6 +195,7 @@ PM St Pierre & Miquelon PN Pitcairn PR Puerto Rico +PS Palestine PT Portugal PW Palau PY Paraguay =================================================================== RCS file: RCS/africa,v retrieving revision 1997.11 retrieving revision 1997.11.1.1 diff -u -r1997.11 -r1997.11.1.1 --- africa 1997/12/29 14:31:49 1997.11 +++ africa 1998/05/24 07:28:25 1997.11.1.1 @@ -105,9 +105,6 @@ 0:52:04 - LMT 1911 May 26 # Luanda Mean Time? 1:00 - WAT -# Bassas da India -# uninhabited - # Benin # Whitman says they switched to 1:00 in 1946, not 1934; go with Shanks. # Zone NAME GMTOFF RULES FORMAT [UNTIL] @@ -240,9 +237,6 @@ 2:35:20 - ADMT 1936 May 5 # Adis Dera MT 3:00 - EAT -# Europa Island -# uninhabited - # Gabon # Zone NAME GMTOFF RULES FORMAT [UNTIL] Zone Africa/Libreville 0:37:48 - LMT 1912 @@ -264,9 +258,6 @@ Zone Africa/Accra -0:00:52 - LMT 1918 0:00 Ghana %s -# Glorioso Is -# uninhabited - # Guinea # Zone NAME GMTOFF RULES FORMAT [UNTIL] Zone Africa/Conakry -0:54:52 - LMT 1912 @@ -280,9 +271,6 @@ -1:00 - WAT 1975 0:00 - GMT -# Juan de Nova -# uninhabited - # Kenya # Zone NAME GMTOFF RULES FORMAT [UNTIL] Zone Africa/Nairobi 2:27:16 - LMT 1928 Jul @@ -450,6 +438,19 @@ # Zone NAME GMTOFF RULES FORMAT [UNTIL] Zone Indian/Reunion 3:41:52 - LMT 1911 Jun # Saint-Denis 4:00 - RET # Reunion Time +# +# Scattered Islands (Iles Eparses) administered from Reunion are as follows. +# The following information about them is taken from +# <a href="http://www.outre-mer.gouv.fr/domtom/ile.htm"> +# Iles Eparses +# </a> +# (1997-07-22, in French). We have no info about their time zone histories. +# +# Bassas da India - uninhabited +# Europa Island - inhabited from 1905 to 1910 by two families +# Glorioso Is - inhabited until at least 1958 +# Juan de Nova - uninhabited +# Tromelin - inhabited until at least 1958 # Rwanda # Zone NAME GMTOFF RULES FORMAT [UNTIL] @@ -553,9 +554,6 @@ Zone Africa/Lome 0:04:52 - LMT 1893 0:00 - GMT -# Tromelin -# uninhabited - # Tunisia # Rule NAME FROM TO TYPE IN ON AT SAVE LETTER/S Rule Tunisia 1939 only - Apr 15 23:00s 1:00 S =================================================================== RCS file: RCS/northamerica,v retrieving revision 1998.2 retrieving revision 1998.2.1.1 diff -u -r1998.2 -r1998.2.1.1 --- northamerica 1998/01/17 19:27:25 1998.2 +++ northamerica 1998/05/24 07:28:25 1998.2.1.1 @@ -262,11 +262,6 @@ -10:30 US H%sT 1947 Jun 8 2:00 -10:00 - HST -Zone Pacific/Midway -11:49:28 - LMT 1901 - -11:00 - NST 1967 Apr # N=Nome - -11:00 - BST 1983 Nov 30 # B=Bering - -11:00 - SST # S=Samoa - # Now we turn to US areas that have diverged from the consensus since 1970. # Arizona mostly uses MST. @@ -1102,9 +1097,6 @@ -5:07 - KMT 1912 Feb # Kingston Mean Time -5:00 - EST -# Clipperton -# uninhabited - # Costa Rica # Shanks gives some very odd dates for 1991, and stops there. # For now, we won't guess further. =================================================================== RCS file: RCS/zone.tab,v retrieving revision 1.2.1.5 retrieving revision 1.2.1.6 diff -u -r1.2.1.5 -r1.2.1.6 --- zone.tab 1997/08/31 08:13:53 1.2.1.5 +++ zone.tab 1998/05/24 07:36:46 1.2.1.6 @@ -178,8 +178,7 @@ ID -0507+11924 Asia/Ujung_Pandang Borneo & Celebes ID -0232+14042 Asia/Jayapura Irian Jaya & the Moluccas IE +5320-00615 Europe/Dublin -IL +3146+03514 Asia/Jerusalem most locations -IL +3130+03428 Asia/Gaza Gaza Strip +IL +3146+03514 Asia/Jerusalem IN +2232+08822 Asia/Calcutta IO -0720+07225 Indian/Chagos IQ +3321+04425 Asia/Baghdad @@ -269,6 +268,7 @@ PM +4703-05620 America/Miquelon PN -2504-13005 Pacific/Pitcairn PR +182806-0660622 America/Puerto_Rico +PS +3130+03428 Asia/Gaza PT +3843-00908 Europe/Lisbon mainland PT +3238-01654 Atlantic/Madeira Madeira Islands PT +3744-02540 Atlantic/Azores Azores
Since we're probably going to change the tz code anyway to address the posix/right problem, I cobbled together all the other proposed tz changes that I've been saving up. Here they are:
[snip]
* At the suggestion of someone who's coordinating this data with ISO country codes, the Clipperton and Midway entries belong in the australasia file, as this munges political boundaries less. Clipperton is administered from French Polynesia, and Midway is one of the US Minor Outlying Islands.
[snip] In the australsia file, how about including Cocos (nearer to Christmas Island than Christmas Island is to Australia) and Norfolk (admittedly further from Australia than, say, New Caledonia, but still closer to Australia's east coast than Perth is) in the "Australian miscellany" section, since they belong to Australia? _______________ Alex LIVINGSTON Macintosh Support IT, Australian Graduate School of Management (AGSM), Uni. of NSW (UNSW) Fax: +61 2 9931-9349 / Phone: +61 2 9931-9264 / Time: UTC + 10 or 11 h.
Date: Mon, 25 May 1998 13:26:14 +1000 From: Alex LIVINGSTON <alex@agsm.unsw.edu.au> In the australsia file, how about including Cocos (nearer to Christmas Island than Christmas Island is to Australia) and Norfolk (admittedly further from Australia than, say, New Caledonia, but still closer to Australia's east coast than Perth is) in the "Australian miscellany" section, since they belong to Australia? Because they have their own ISO country codes. Actually, consistency suggests that Christmas Island should be kept seperate from Australian miscellany, as it has its own ISO country code. E.g. something like the following. This doesn't change the order of the data entries, just the commentary, which is probably why the other coordinator didn't notice this discrepancy. --- australasia 1998/05/24 07:28:25 1998.4.1.1 +++ australasia 1998/05/25 07:19:09 1998.4.1.2 @@ -159,7 +159,4 @@ Zone Australia/Lord_Howe 10:36:20 - LMT # Australian miscellany -# Zone NAME GMTOFF RULES FORMAT [UNTIL] -Zone Indian/Christmas 7:02:52 - LMT 1895 Feb - 7:00 - CXT # Christmas Island Time # # Ashmore Is, Cartier @@ -179,4 +176,8 @@ Zone Indian/Christmas 7:02:52 - LMT 1895 # no information +# Christmas +# Zone NAME GMTOFF RULES FORMAT [UNTIL] +Zone Indian/Christmas 7:02:52 - LMT 1895 Feb + 7:00 - CXT # Christmas Island Time
participants (3)
-
Alex LIVINGSTON -
Bradley White -
Paul Eggert