libc/time/theory.html

<!DOCTYPE html>
<html lang="en">
<head>
  <title>Theory and pragmatics of the tz code and data</title>
  <meta charset="UTF-8">
  <style>
    pre {margin-left: 2em; white-space: pre-wrap;}
  </style>
</head>

<body>
<h1>Theory and pragmatics of the <code><abbr>tz</abbr></code> code and data</h1>
  <h3>Outline</h3>
  <nav>
    <ul>
      <li><a href="#scope">Scope of the <code><abbr>tz</abbr></code>
	  database</a></li>
      <li><a href="#naming">Timezone identifiers</a></li>
      <li><a href="#abbreviations">Time zone abbreviations</a></li>
      <li><a href="#accuracy">Accuracy of the <code><abbr>tz</abbr></code>
	  database</a></li>
      <li><a href="#functions">Time and date functions</a></li>
      <li><a href="#stability">Interface stability</a></li>
      <li><a href="#leapsec">Leap seconds</a></li>
      <li><a href="#calendar">Calendrical issues</a></li>
      <li><a href="#planets">Time and time zones off earth</a></li>
    </ul>
  </nav>

<section>
  <h2 id="scope">Scope of the <code><abbr>tz</abbr></code> database</h2>
<p>
The <a
href="https://www.iana.org/time-zones"><code><abbr>tz</abbr></code>
database</a> attempts to record the history and predicted future of
civil time scales.
It organizes <a href="tz-link.html">time zone and daylight saving time
data</a> by partitioning the world into <a
href="https://en.wikipedia.org/wiki/List_of_tz_database_time_zones"><dfn>timezones</dfn></a>
whose clocks all agree about timestamps that occur after the <a
href="https://en.wikipedia.org/wiki/Unix_time">POSIX Epoch</a>
(1970-01-01 00:00:00 <a
href="https://en.wikipedia.org/wiki/Coordinated_Universal_Time"><abbr
title="Coordinated Universal Time">UTC</abbr></a>).
Although 1970 is a somewhat-arbitrary cutoff, there are significant
challenges to moving the cutoff earlier even by a decade or two, due
to the wide variety of local practices before computer timekeeping
became prevalent.
Most timezones correspond to a notable location and the database
records all known clock transitions for that location;
some timezones correspond instead to a fixed <abbr>UTC</abbr> offset.
</p>

<p>
Each timezone typically corresponds to a geographical region that is
smaller than a traditional time zone, because clocks in a timezone
all agree after 1970 whereas a traditional time zone merely
specifies current standard time. For example, applications that deal
with current and future timestamps in the traditional North
American mountain time zone can choose from the timezones
<code>America/Denver</code> which observes US-style daylight saving
time (<abbr>DST</abbr>),
and <code>America/Phoenix</code> which does not observe <abbr>DST</abbr>.
Applications that also deal with past timestamps in the mountain time
zone can choose from over a dozen timezones, such as
<code>America/Boise</code>, <code>America/Edmonton</code>, and
<code>America/Hermosillo</code>, each of which currently uses mountain
time but differs from other timezones for some timestamps after 1970.
</p>

<p>
Clock transitions before 1970 are recorded for location-based timezones,
because most systems support timestamps before 1970 and could
misbehave if data entries were omitted for pre-1970 transitions.
However, the database is not designed for and does not suffice for
applications requiring accurate handling of all past times everywhere,
as it would take far too much effort and guesswork to record all
details of pre-1970 civil timekeeping.
Although some information outside the scope of the database is
collected in a file <code>backzone</code> that is distributed along
with the database proper, this file is less reliable and does not
necessarily follow database guidelines.
</p>

<p>
As described below, reference source code for using the
<code><abbr>tz</abbr></code> database is also available.
The <code><abbr>tz</abbr></code> code is upwards compatible with <a
href="https://en.wikipedia.org/wiki/POSIX">POSIX</a>, an international
standard for <a
href="https://en.wikipedia.org/wiki/Unix">UNIX</a>-like systems.
As of this writing, the current edition of POSIX is: <a
href="https://pubs.opengroup.org/onlinepubs/9699919799/"> The Open
Group Base Specifications Issue 7</a>, IEEE Std 1003.1-2017, 2018
Edition.
Because the database's scope encompasses real-world changes to civil
timekeeping, its model for describing time is more complex than the
standard and daylight saving times supported by POSIX.1-2017.
A <code><abbr>tz</abbr></code> timezone corresponds to a ruleset that can
have more than two changes per year, these changes need not merely
flip back and forth between two alternatives, and the rules themselves
can change at times.
Whether and when a timezone changes its clock,
and even the timezone's notional base offset from <abbr>UTC</abbr>,
are variable.
It does not always make sense to talk about a timezone's
"base offset", which is not necessarily a single number.
</p>

</section>

<section>
  <h2 id="naming">Timezone identifiers</h2>
<p>
Each timezone has a name that uniquely identifies the timezone.
Inexperienced users are not expected to select these names unaided.
Distributors should provide documentation and/or a simple selection
interface that explains each name via a map or via descriptive text like
"Czech Republic" instead of the timezone name "<code>Europe/Prague</code>".
If geolocation information is available, a selection interface can
locate the user on a timezone map or prioritize names that are
geographically close. For an example selection interface, see the
<code>tzselect</code> program in the <code><abbr>tz</abbr></code> code.
The <a href="https://cldr.unicode.org">Unicode Common Locale Data
Repository</a> contains data that may be useful for other selection
interfaces; it maps timezone names like <code>Europe/Prague</code> to
locale-dependent strings like "Prague", "Praha", "", and "".
</p>

<p>
The naming conventions attempt to strike a balance
among the following goals:
</p>

<ul>
  <li>
    Uniquely identify every timezone where clocks have agreed since 1970.
    This is essential for the intended use: static clocks keeping local
    civil time.
  </li>
  <li>
    Indicate to experts where the timezone's clocks typically are.
  </li>
  <li>
    Be robust in the presence of political changes.
    For example, names are typically not tied to countries, to avoid
    incompatibilities when countries change their name (e.g.,
    Swaziland&rarr;Eswatini) or when locations change countries (e.g., Hong
    Kong from UK colony to China).
    There is no requirement that every country or national
    capital must have a timezone name.
  </li>
  <li>
    Be portable to a wide variety of implementations.
  </li>
  <li>
    Use a consistent naming conventions over the entire world.
  </li>
</ul>

<p>
Names normally have the form
<var>AREA</var><code>/</code><var>LOCATION</var>, where
<var>AREA</var> is a continent or ocean, and
<var>LOCATION</var> is a specific location within the area.
North and South America share the same area, '<code>America</code>'.
Typical names are '<code>Africa/Cairo</code>',
'<code>America/New_York</code>', and '<code>Pacific/Honolulu</code>'.
Some names are further qualified to help avoid confusion; for example,
'<code>America/Indiana/Petersburg</code>' distinguishes Petersburg,
Indiana from other Petersburgs in America.
</p>

<p>
Here are the general guidelines used for
choosing timezone names,
in decreasing order of importance:
</p>

<ul>
  <li>
    Use only valid POSIX file name components (i.e., the parts of
    names other than '<code>/</code>').
    Do not use the file name components '<code>.</code>' and
    '<code>..</code>'.
    Within a file name component, use only <a
    href="https://en.wikipedia.org/wiki/ASCII">ASCII</a> letters,
    '<code>.</code>', '<code>-</code>' and '<code>_</code>'.
    Do not use digits, as that might create an ambiguity with <a
    href="https://pubs.opengroup.org/onlinepubs/9699919799/basedefs/V1_chap08.html#tag_08_03">POSIX.1-2017
    <code>TZ</code> strings</a>.
    A file name component must not exceed 14 characters or start with
    '<code>-</code>'.
    E.g., prefer <code>America/Noronha</code> to
    <code>America/Fernando_de_Noronha</code>.
    Exceptions: see the discussion of legacy names below.
  </li>
  <li>
    A name must not be empty, or contain '<code>//</code>', or
    start or end with '<code>/</code>'.
  </li>
  <li>
    Do not use names that differ only in case.
    Although the reference implementation is case-sensitive, some
    other implementations are not, and they would mishandle names
    differing only in case.
  </li>
  <li>
    If one name <var>A</var> is an initial prefix of another
    name <var>AB</var> (ignoring case), then <var>B</var> must not
    start with '<code>/</code>', as a regular file cannot have the
    same name as a directory in POSIX.
    For example, <code>America/New_York</code> precludes
    <code>America/New_York/Bronx</code>.
  </li>
  <li>
    Uninhabited regions like the North Pole and Bouvet Island
    do not need locations, since local time is not defined there.
  </li>
  <li>
    If all the clocks in a timezone have agreed since 1970,
    do not bother to include more than one timezone
    even if some of the clocks disagreed before 1970.
    Otherwise these tables would become annoyingly large.
  </li>
  <li>
    If boundaries between regions are fluid, such as during a war or
    insurrection, do not bother to create a new timezone merely
    because of yet another boundary change. This helps prevent table
    bloat and simplifies maintenance.
  </li>
  <li>
    If a name is ambiguous, use a less ambiguous alternative;
    e.g., many cities are named San Jos and Georgetown, so
    prefer <code>America/Costa_Rica</code> to
    <code>America/San_Jose</code> and <code>America/Guyana</code>
    to <code>America/Georgetown</code>.
  </li>
  <li>
    Keep locations compact.
    Use cities or small islands, not countries or regions, so that any
    future changes do not split individual locations into different
    timezones.
    E.g., prefer <code>Europe/Paris</code> to <code>Europe/France</code>,
    since
    <a href="https://en.wikipedia.org/wiki/Time_in_France#History">France
    has had multiple time zones</a>.
  </li>
  <li>
    Use mainstream English spelling, e.g., prefer
    <code>Europe/Rome</code> to <code>Europa/Roma</code>, and
    prefer <code>Europe/Athens</code> to the Greek
    <code>/</code> or the Romanized
    <code>Evrpi/Athna</code>.
    The POSIX file name restrictions encourage this guideline.
  </li>
  <li>
    Use the most populous among locations in a region,
    e.g., prefer <code>Asia/Shanghai</code> to
    <code>Asia/Beijing</code>.
    Among locations with similar populations, pick the best-known
    location, e.g., prefer <code>Europe/Rome</code> to
    <code>Europe/Milan</code>.
  </li>
  <li>
    Use the singular form, e.g., prefer <code>Atlantic/Canary</code> to
    <code>Atlantic/Canaries</code>.
  </li>
  <li>
    Omit common suffixes like '<code>_Islands</code>' and
    '<code>_City</code>', unless that would lead to ambiguity.
    E.g., prefer <code>America/Cayman</code> to
    <code>America/Cayman_Islands</code> and
    <code>America/Guatemala</code> to
    <code>America/Guatemala_City</code>, but prefer
    <code>America/Mexico_City</code> to
    <code>America/Mexico</code>
    because <a href="https://en.wikipedia.org/wiki/Time_in_Mexico">the
    country of Mexico has several time zones</a>.
  </li>
  <li>
    Use '<code>_</code>' to represent a space.
  </li>
  <li>
    Omit '<code>.</code>' from abbreviations in names.
    E.g., prefer <code>Atlantic/St_Helena</code> to
    <code>Atlantic/St._Helena</code>.
  </li>
  <li>
    Do not change established names if they only marginally violate
    the above guidelines.
    For example, do not change the existing name <code>Europe/Rome</code> to
    <code>Europe/Milan</code> merely because Milan's population has grown
    to be somewhat greater than Rome's.
  </li>
  <li>
    If a name is changed, put its old spelling in the
    '<code>backward</code>' file as a link to the new spelling.
    This means old spellings will continue to work.
    Ordinarily a name change should occur only in the rare case when
    a location's consensus English-language spelling changes; for example,
    in 2008 <code>Asia/Calcutta</code> was renamed to <code>Asia/Kolkata</code>
    due to long-time widespread use of the new city name instead of the old.
  </li>
</ul>

<p>
Guidelines have evolved with time, and names following old versions of
these guidelines might not follow the current version. When guidelines
have changed, old names continue to be supported. Guideline changes
have included the following:
</p>

<ul>
<li>
Older versions of this package used a different naming scheme.
See the file '<code>backward</code>' for most of these older names
(e.g., '<code>US/Eastern</code>' instead of '<code>America/New_York</code>').
The other old-fashioned names still supported are
'<code>WET</code>', '<code>CET</code>', '<code>MET</code>', and
'<code>EET</code>' (see the file '<code>europe</code>').
</li>

<li>
Older versions of this package defined legacy names that are
incompatible with the first guideline of location names, but which are
still supported.
These legacy names are mostly defined in the file
'<code>etcetera</code>'.
Also, the file '<code>backward</code>' defines the legacy names
'<code>Etc/GMT0</code>', '<code>Etc/GMT-0</code>', '<code>Etc/GMT+0</code>',
'<code>GMT0</code>', '<code>GMT-0</code>' and '<code>GMT+0</code>',
and the file '<code>northamerica</code>' defines the legacy names
'<code>EST5EDT</code>', '<code>CST6CDT</code>',
'<code>MST7MDT</code>', and '<code>PST8PDT</code>'.
</li>

<li>
Older versions of these guidelines said that
there should typically be at least one name for each <a
href="https://en.wikipedia.org/wiki/ISO_3166-1"><abbr
title="International Organization for Standardization">ISO</abbr>
3166-1</a> officially assigned two-letter code for an inhabited
country or territory.
This old guideline has been dropped, as it was not needed to handle
timestamps correctly and it increased maintenance burden.
</li>
</ul>

<p>
The file <code>zone1970.tab</code> lists geographical locations used
to name timezones.
It is intended to be an exhaustive list of names for geographic
regions as described above; this is a subset of the timezones in the data.
Although a <code>zone1970.tab</code> location's
<a href="https://en.wikipedia.org/wiki/Longitude">longitude</a>
corresponds to
its <a href="https://en.wikipedia.org/wiki/Local_mean_time">local mean
time (<abbr>LMT</abbr>)</a> offset with one hour for every 15&deg;
east longitude, this relationship is not exact.
The backward-compatibility file <code>zone.tab</code> is similar
but conforms to the older-version guidelines related to <abbr>ISO</abbr> 3166-1;
it lists only one country code per entry and unlike <code>zone1970.tab</code>
it can list names defined in <code>backward</code>.
Applications that process only timestamps from now on can instead use the file
<code>zonenow.tab</code>, which partitions the world more coarsely,
into regions where clocks agree now and in the predicted future;
this file is smaller and simpler than <code>zone1970.tab</code>
and <code>zone.tab</code>.
</p>

<p>
The database defines each timezone name to be a zone, or a link to a zone.
The source file <code>backward</code> defines links for backward
compatibility; it does not define zones.
Although <code>backward</code> was originally designed to be optional,
nowadays distributions typically use it
and no great weight should be attached to whether a link
is defined in <code>backward</code> or in some other file.
The source file <code>etcetera</code> defines names that may be useful
on platforms that do not support POSIX.1-2017-style <code>TZ</code> strings;
no other source file other than <code>backward</code>
contains links to its zones.
One of <code>etcetera</code>'s names is <code>Etc/UTC</code>,
used by functions like <code>gmtime</code> to obtain leap
second information on platforms that support leap seconds.
Another <code>etcetera</code> name, <code>GMT</code>,
is used by older code releases.
</p>
</section>

<section>
  <h2 id="abbreviations">Time zone abbreviations</h2>
<p>
When this package is installed, it generates time zone abbreviations
like '<code>EST</code>' to be compatible with human tradition and POSIX.
Here are the general guidelines used for choosing time zone abbreviations,
in decreasing order of importance:
</p>

<ul>
  <li>
    Use three to six characters that are ASCII alphanumerics or
    '<code>+</code>' or '<code>-</code>'.
    Previous editions of this database also used characters like
    space and '<code>?</code>', but these characters have a
    special meaning to the
    <a href="https://en.wikipedia.org/wiki/Unix_shell">UNIX shell</a>
    and cause commands like
    '<code><a href="https://pubs.opengroup.org/onlinepubs/9699919799/utilities/V3_chap02.html#set">set</a>
    `<a href="https://pubs.opengroup.org/onlinepubs/9699919799/utilities/date.html">date</a>`</code>'
    to have unexpected effects.
    Previous editions of this guideline required upper-case letters, but the
    Congressman who introduced
    <a href="https://en.wikipedia.org/wiki/Chamorro_Time_Zone">Chamorro
    Standard Time</a> preferred "ChST", so lower-case letters are now
    allowed.
    Also, POSIX from 2001 on relaxed the rule to allow '<code>-</code>',
    '<code>+</code>', and alphanumeric characters from the portable
    character set in the current locale.
    In practice ASCII alphanumerics and '<code>+</code>' and
    '<code>-</code>' are safe in all locales.

    <p>
    In other words, in the C locale the POSIX extended regular
    expression <code>[-+[:alnum:]]{3,6}</code> should match the
    abbreviation.
    This guarantees that all abbreviations could have been specified by a
    POSIX.1-2017 <code>TZ</code> string.
    </p>
  </li>
  <li>
    Use abbreviations that are in common use among English-speakers,
    e.g., 'EST' for Eastern Standard Time in North America.
    We assume that applications translate them to other languages
    as part of the normal localization process; for example,
    a French application might translate 'EST' to 'HNE'.

    <p>
    <small>These abbreviations (for standard/daylight/etc. time) are:
      ACST/ACDT Australian Central,
      AST/ADT/APT/AWT/ADDT Atlantic,
      AEST/AEDT Australian Eastern,
      AHST/AHDT Alaska-Hawaii,
      AKST/AKDT Alaska,
      AWST/AWDT Australian Western,
      BST/BDT Bering,
      CAT/CAST Central Africa,
      CET/CEST/CEMT Central European,
      ChST Chamorro,
      CST/CDT/CWT/CPT Central [North America],
      CST/CDT China,
      GMT/BST/IST/BDST Greenwich,
      EAT East Africa,
      EST/EDT/EWT/EPT Eastern [North America],
      EET/EEST Eastern European,
      GST/GDT Guam,
      HST/HDT/HWT/HPT Hawaii,
      HKT/HKST/HKWT Hong Kong,
      IST India,
      IST/GMT Irish,
      IST/IDT/IDDT Israel,
      JST/JDT Japan,
      KST/KDT Korea,
      MET/MEST Middle European (a backward-compatibility alias for
	Central European),
      MSK/MSD Moscow,
      MST/MDT/MWT/MPT Mountain,
      NST/NDT/NWT/NPT/NDDT Newfoundland,
      NST/NDT/NWT/NPT Nome,
      NZMT/NZST New Zealand through 1945,
      NZST/NZDT New Zealand 1946&ndash;present,
      PKT/PKST Pakistan,
      PST/PDT/PWT/PPT Pacific,
      PST/PDT Philippine,
      SAST South Africa,
      SST Samoa,
      UTC Universal,
      WAT/WAST West Africa,
      WET/WEST/WEMT Western European,
      WIB Waktu Indonesia Barat,
      WIT Waktu Indonesia Timur,
      WITA Waktu Indonesia Tengah,
      YST/YDT/YWT/YPT/YDDT Yukon</small>.
    </p>
  </li>
  <li>
    <p>
    For times taken from a city's longitude, use the
    traditional <var>x</var>MT notation.
    The only abbreviation like this in current use is '<abbr>GMT</abbr>'.
    The others are for timestamps before 1960,
    except that Monrovia Mean Time persisted until 1972.
    Typically, numeric abbreviations (e.g., '<code>-</code>004430' for
    MMT) would cause trouble here, as the numeric strings would exceed
    the POSIX length limit.
    </p>

    <p>
    <small>These abbreviations are:
      AMT Asuncin, Athens;
      BMT Baghdad, Bangkok, Batavia, Bermuda, Bern, Bogot,
        Brussels, Bucharest;
      CMT Calamarca, Caracas, Chisinau, Coln, Crdoba;
      DMT Dublin/Dunsink;
      EMT Easter;
      FFMT Fort-de-France;
      FMT Funchal;
      GMT Greenwich;
      HMT Havana, Helsinki, Horta, Howrah;
      IMT Irkutsk, Istanbul;
      JMT Jerusalem;
      KMT Kaunas, Kyiv, Kingston;
      LMT Lima, Lisbon, local;
      MMT Macassar, Madras, Mal, Managua, Minsk, Monrovia, Montevideo,
	Moratuwa, Moscow;
      PLMT Ph Lin;
      PMT Paramaribo, Paris, Perm, Pontianak, Prague;
      PMMT Port Moresby;
      PPMT Port-au-Prince;
      QMT Quito;
      RMT Rangoon, Riga, Rome;
      SDMT Santo Domingo;
      SJMT San Jos;
      SMT Santiago, Simferopol, Singapore, Stanley;
      TBMT Tbilisi;
      TMT Tallinn, Tehran;
      WMT Warsaw.</small>
    </p>

    <p>
    <small>A few abbreviations also follow the pattern that
    <abbr>GMT</abbr>/<abbr>BST</abbr> established for time in the UK.
    They are:
      BMT/BST for Bermuda 1890&ndash;1930,
      CMT/BST for Calamarca Mean Time and Bolivian Summer Time
	1890&ndash;1932,
      DMT/IST for Dublin/Dunsink Mean Time and Irish Summer Time
	1880&ndash;1916,
      MMT/MST/MDST for Moscow 1880&ndash;1919, and
      RMT/LST for Riga Mean Time and Latvian Summer time 1880&ndash;1926.
    </small>
    </p>
  </li>
  <li>
    Use '<abbr>LMT</abbr>' for local mean time of locations before the
    introduction of standard time; see "<a href="#scope">Scope of the
    <code><abbr>tz</abbr></code> database</a>".
  </li>
  <li>
    If there is no common English abbreviation, use numeric offsets like
    <code>-</code>05 and <code>+</code>0530 that are generated
    by <code>zic</code>'s <code>%z</code> notation.
  </li>
  <li>
    Use current abbreviations for older timestamps to avoid confusion.
    For example, in 1910 a common English abbreviation for time
    in central Europe was 'MEZ' (short for both "Middle European
    Zone" and for "Mitteleuropische Zeit" in German).
    Nowadays 'CET' ("Central European Time") is more common in
    English, and the database uses 'CET' even for circa-1910
    timestamps as this is less confusing for modern users and avoids
    the need for determining when 'CET' supplanted 'MEZ' in common
    usage.
  </li>
  <li>
    Use a consistent style in a timezone's history.
    For example, if a history tends to use numeric
    abbreviations and a particular entry could go either way, use a
    numeric abbreviation.
  </li>
  <li>
    Use
    <a href="https://en.wikipedia.org/wiki/Universal_Time">Universal Time</a>
    (<abbr>UT</abbr>) (with time zone abbreviation '<code>-</code>00') for
    locations while uninhabited.
    The leading '<code>-</code>' is a flag that the <abbr>UT</abbr> offset is in
    some sense undefined; this notation is derived
    from <a href="https://datatracker.ietf.org/doc/html/rfc3339">Internet
    <abbr title="Request For Comments">RFC</abbr> 3339</a>.
  </li>
</ul>

<p>
Application writers should note that these abbreviations are ambiguous
in practice: e.g., 'CST' means one thing in China and something else
in North America, and 'IST' can refer to time in India, Ireland or
Israel.
To avoid ambiguity, use numeric <abbr>UT</abbr> offsets like
'<code>-</code>0600' instead of time zone abbreviations like 'CST'.
</p>
</section>

<section>
  <h2 id="accuracy">Accuracy of the <code><abbr>tz</abbr></code> database</h2>
<p>
The <code><abbr>tz</abbr></code> database is not authoritative, and it
surely has errors.
Corrections are welcome and encouraged; see the file <code>CONTRIBUTING</code>.
Users requiring authoritative data should consult national standards
bodies and the references cited in the database's comments.
</p>

<p>
Errors in the <code><abbr>tz</abbr></code> database arise from many sources:
</p>

<ul>
  <li>
    The <code><abbr>tz</abbr></code> database predicts future
    timestamps, and current predictions
    will be incorrect after future governments change the rules.
    For example, if today someone schedules a meeting for 13:00 next
    October 1, Casablanca time, and tomorrow Morocco changes its
    daylight saving rules, software can mess up after the rule change
    if it blithely relies on conversions made before the change.
  </li>
  <li>
    The pre-1970 entries in this database cover only a tiny sliver of how
    clocks actually behaved; the vast majority of the necessary
    information was lost or never recorded.
    Thousands more timezones would be needed if
    the <code><abbr>tz</abbr></code> database's scope were extended to
    cover even just the known or guessed history of standard time; for
    example, the current single entry for France would need to split
    into dozens of entries, perhaps hundreds.
    And in most of the world even this approach would be misleading
    due to widespread disagreement or indifference about what times
    should be observed.
    In her 2015 book
    <cite><a
    href="https://www.hup.harvard.edu/catalog.php?isbn=9780674286146">The
    Global Transformation of Time, 1870&ndash;1950</a></cite>,
    Vanessa Ogle writes
    "Outside of Europe and North America there was no system of time
    zones at all, often not even a stable landscape of mean times,
    prior to the middle decades of the twentieth century".
    See: Timothy Shenk, <a
href="https://www.dissentmagazine.org/blog/booked-a-global-history-of-time-vanessa-ogle">Booked:
      A Global History of Time</a>. <cite>Dissent</cite> 2015-12-17.
  </li>
  <li>
    Most of the pre-1970 data entries come from unreliable sources, often
    astrology books that lack citations and whose compilers evidently
    invented entries when the true facts were unknown, without
    reporting which entries were known and which were invented.
    These books often contradict each other or give implausible entries,
    and on the rare occasions when they are checked they are
    typically found to be incorrect.
  </li>
  <li>
    For the UK the <code><abbr>tz</abbr></code> database relies on
    years of first-class work done by
    Joseph Myers and others; see
    "<a href="https://www.polyomino.org.uk/british-time/">History of
    legal time in Britain</a>".
    Other countries are not done nearly as well.
  </li>
  <li>
    Sometimes, different people in the same city maintain clocks
    that differ significantly.
    Historically, railway time was used by railroad companies (which
    did not always
    agree with each other), church-clock time was used for birth
    certificates, etc.
    More recently, competing political groups might disagree about
    clock settings. Often this is merely common practice, but
    sometimes it is set by law.
    For example, from 1891 to 1911 the <abbr>UT</abbr> offset in France
    was legally <abbr>UT</abbr> +00:09:21 outside train stations and
    <abbr>UT</abbr> +00:04:21 inside. Other examples include
    Chillicothe in 1920, Palm Springs in 1946/7, and Jerusalem and
    rmqi to this day.
  </li>
  <li>
    Although a named location in the <code><abbr>tz</abbr></code>
    database stands for the containing region, its pre-1970 data
    entries are often accurate for only a small subset of that region.
    For example, <code>Europe/London</code> stands for the United
    Kingdom, but its pre-1847 times are valid only for locations that
    have London's exact meridian, and its 1847 transition
    to <abbr>GMT</abbr> is known to be valid only for the L&amp;NW and
    the Caledonian railways.
  </li>
  <li>
    The <code><abbr>tz</abbr></code> database does not record the
    earliest time for which a timezone's
    data entries are thereafter valid for every location in the region.
    For example, <code>Europe/London</code> is valid for all locations
    in its region after <abbr>GMT</abbr> was made the standard time,
    but the date of standardization (1880-08-02) is not in the
    <code><abbr>tz</abbr></code> database, other than in commentary.
    For many timezones the earliest time of
    validity is unknown.
  </li>
  <li>
    The <code><abbr>tz</abbr></code> database does not record a
    region's boundaries, and in many cases the boundaries are not known.
    For example, the timezone
    <code>America/Kentucky/Louisville</code> represents a region
    around the city of Louisville, the boundaries of which are
    unclear.
  </li>
  <li>
    Changes that are modeled as instantaneous transitions in the
    <code><abbr>tz</abbr></code>
    database were often spread out over hours, days, or even decades.
  </li>
  <li>
    Even if the time is specified by law, locations sometimes
    deliberately flout the law.
  </li>
  <li>
    Early timekeeping practices, even assuming perfect clocks, were
    often not specified to the accuracy that the
    <code><abbr>tz</abbr></code> database requires.
  </li>
  <li>
    The <code><abbr>tz</abbr></code> database cannot represent stopped clocks.
    However, on 1911-03-11 at 00:00, some public-facing French clocks
    were changed by stopping them for a few minutes to effect a transition.
    The <code><abbr>tz</abbr></code> database models this via a
    backward transition; the relevant French legislation does not
    specify exactly how the transition was to occur.
  </li>
  <li>
    Sometimes historical timekeeping was specified more precisely
    than what the <code><abbr>tz</abbr></code> code can handle.
    For example, from 1880 to 1916 clocks in Ireland observed Dublin Mean
    Time (estimated to be <abbr>UT</abbr>
    &minus;00:25:21.1); although the <code><abbr>tz</abbr></code>
    source data can represent the .1 second, TZif files and the code cannot.
    In practice these old specifications were rarely if ever
    implemented to subsecond precision.
  </li>
  <li>
    Even when all the timestamp transitions recorded by the
    <code><abbr>tz</abbr></code> database are correct, the
    <code><abbr>tz</abbr></code> rules that generate them may not
    faithfully reflect the historical rules.
    For example, from 1922 until World War II the UK moved clocks
    forward the day following the third Saturday in April unless that
    was Easter, in which case it moved clocks forward the previous
    Sunday.
    Because the <code><abbr>tz</abbr></code> database has no
    way to specify Easter, these exceptional years are entered as
    separate <code><abbr>tz</abbr> Rule</code> lines, even though the
    legal rules did not change.
    When transitions are known but the historical rules behind them are not,
    the database contains <code>Zone</code> and <code>Rule</code>
    entries that are intended to represent only the generated
    transitions, not any underlying historical rules; however, this
    intent is recorded at best only in commentary.
  </li>
  <li>
    The <code><abbr>tz</abbr></code> database models time
    using the <a
    href="https://en.wikipedia.org/wiki/Proleptic_Gregorian_calendar">proleptic
    Gregorian calendar</a> with days containing 24 equal-length hours
    numbered 00 through 23, except when clock transitions occur.
    Pre-standard time is modeled as local mean time.
    However, historically many people used other calendars and other timescales.
    For example, the Roman Empire used
    the <a href="https://en.wikipedia.org/wiki/Julian_calendar">Julian
    calendar</a>,
    and <a href="https://en.wikipedia.org/wiki/Roman_timekeeping">Roman
    timekeeping</a> had twelve varying-length daytime hours with a
    non-hour-based system at night.
    And even today, some local practices diverge from the Gregorian
    calendar with 24-hour days. These divergences range from
    relatively minor, such as Japanese bars giving times like "24:30" for the
    wee hours of the morning, to more-significant differences such as <a
    href="https://theworld.org/stories/2015-01-30/if-you-have-meeting-ethiopia-you-better-double-check-time">the
    east African practice of starting the day at dawn</a>, renumbering
    the Western 06:00 to be 12:00. These practices are largely outside
    the scope of the <code><abbr>tz</abbr></code> code and data, which
    provide only limited support for date and time localization
    such as that required by POSIX.1-2017.
    If <abbr>DST</abbr> is not used a different time zone
    can often do the trick; for example, in Kenya a <code>TZ</code> setting
    like <code>&lt;-03&gt;3</code> or <code>America/Cayenne</code> starts
    the day six hours later than <code>Africa/Nairobi</code> does.
  </li>
  <li>
    Early clocks were less reliable, and data entries do not represent
    clock error.
  </li>
  <li>
    The <code><abbr>tz</abbr></code> database assumes Universal Time
    (<abbr>UT</abbr>) as an origin, even though <abbr>UT</abbr> is not
    standardized for older timestamps.
    In the <code><abbr>tz</abbr></code> database commentary,
    <abbr>UT</abbr> denotes a family of time standards that includes
    Coordinated Universal Time (<abbr>UTC</abbr>) along with other
    variants such as <abbr>UT1</abbr> and <abbr>GMT</abbr>,
    with days starting at midnight.
    Although <abbr>UT</abbr> equals <abbr>UTC</abbr> for modern
    timestamps, <abbr>UTC</abbr> was not defined until 1960, so
    commentary uses the more general abbreviation <abbr>UT</abbr> for
    timestamps that might predate 1960.
    Since <abbr>UT</abbr>, <abbr>UT1</abbr>, etc. disagree slightly,
    and since pre-1972 <abbr>UTC</abbr> seconds varied in length,
    interpretation of older timestamps can be problematic when
    subsecond accuracy is needed.
  </li>
  <li>
    Civil time was not based on atomic time before 1972, and we do not
    know the history of
    <a href="https://en.wikipedia.org/wiki/Earth's_rotation">earth's
    rotation</a> accurately enough to map <a
    href="https://en.wikipedia.org/wiki/International_System_of_Units"><abbr
    title="International System of Units">SI</abbr></a> seconds to
    historical <a href="https://en.wikipedia.org/wiki/Solar_time">solar time</a>
    to more than about one-hour accuracy.
    See: Stephenson FR, Morrison LV, Hohenkerk CY.
    <a href="https://dx.doi.org/10.1098/rspa.2016.0404">Measurement of
    the Earth's rotation: 720 BC to AD 2015</a>.
    <cite>Proc Royal Soc A</cite>. 2016;472:20160404.
    Also see: Espenak F. <a
    href="https://eclipse.gsfc.nasa.gov/SEhelp/uncertainty2004.html">Uncertainty
    in Delta T (T)</a>.
  </li>
  <li>
    The relationship between POSIX time (that is, <abbr>UTC</abbr> but
    ignoring <a href="https://en.wikipedia.org/wiki/Leap_second">leap
    seconds</a>) and <abbr>UTC</abbr> is not agreed upon.
    This affects time stamps during the leap second era (1972&ndash;2035).
    Although the POSIX
    clock officially stops during an inserted leap second, at least one
    proposed standard has it jumping back a second instead; and in
    practice POSIX clocks more typically either progress glacially during
    a leap second, or are slightly slowed while near a leap second.
  </li>
  <li>
    The <code><abbr>tz</abbr></code> database does not represent how
    uncertain its information is.
    Ideally it would contain information about when data entries are
    incomplete or dicey.
    Partial temporal knowledge is a field of active research, though,
    and it is not clear how to apply it here.
  </li>
</ul>

<p>
In short, many, perhaps most, of the <code><abbr>tz</abbr></code>
database's pre-1970 and future timestamps are either wrong or
misleading.
Any attempt to pass the
<code><abbr>tz</abbr></code> database off as the definition of time
should be unacceptable to anybody who cares about the facts.
In particular, the <code><abbr>tz</abbr></code> database's
<abbr>LMT</abbr> offsets should not be considered meaningful, and
should not prompt creation of timezones
merely because two locations
differ in <abbr>LMT</abbr> or transitioned to standard time at
different dates.
</p>
</section>

<section>
  <h2 id="functions">Time and date functions</h2>
<p>
The <code><abbr>tz</abbr></code> code contains time and date functions
that are upwards compatible with those of POSIX.
Code compatible with this package is already
<a href="tz-link.html#tzdb">part of many platforms</a>, where the
primary use of this package is to update obsolete time-related files.
To do this, you may need to compile the time zone compiler
'<code>zic</code>' supplied with this package instead of using the
system '<code>zic</code>', since the format of <code>zic</code>'s
input is occasionally extended, and a platform may still be shipping
an older <code>zic</code>.
</p>

<h3 id="POSIX">POSIX.1-2017 properties and limitations</h3>
<ul>
  <li>
    <p>
    In POSIX.1-2017, time display in a process is controlled by the
    environment variable <code>TZ</code>.
    Unfortunately, the POSIX.1-2017
    <code>TZ</code> string takes a form that is hard to describe and
    is error-prone in practice.
    Also, POSIX.1-2017 <code>TZ</code> strings cannot deal with daylight
    saving time rules not based on the Gregorian calendar (as in
    Morocco), or with situations where more than two time zone
    abbreviations or <abbr>UT</abbr> offsets are used in an area.
    </p>

    <p>
    The POSIX.1-2017 <code>TZ</code> string takes the following form:
    </p>

    <p>
    <var>stdoffset</var>[<var>dst</var>[<var>offset</var>][<code>,</code><var>date</var>[<code>/</code><var>time</var>]<code>,</code><var>date</var>[<code>/</code><var>time</var>]]]
    </p>

    <p>
    where:
    </p>

    <dl>
      <dt><var>std</var> and <var>dst</var></dt><dd>
	are 3 or more characters specifying the standard
	and daylight saving time (<abbr>DST</abbr>) zone abbreviations.
	Starting with POSIX.1-2001, <var>std</var> and <var>dst</var>
	may also be in a quoted form like '<code>&lt;+09&gt;</code>';
	this allows "<code>+</code>" and "<code>-</code>" in the names.
      </dd>
      <dt><var>offset</var></dt><dd>
	is of the form
	'<code>[&plusmn;]<var>hh</var>:[<var>mm</var>[:<var>ss</var>]]</code>'
	and specifies the offset west of <abbr>UT</abbr>.
	'<var>hh</var>' may be a single digit;
	0&le;<var>hh</var>&le;24.
	The default <abbr>DST</abbr> offset is one hour ahead of
	standard time.
      </dd>
      <dt><var>date</var>[<code>/</code><var>time</var>]<code>,</code><var>date</var>[<code>/</code><var>time</var>]</dt><dd>
	specifies the beginning and end of <abbr>DST</abbr>.
	If this is absent, the system supplies its own ruleset
	for <abbr>DST</abbr>, typically	current <abbr>US</abbr>
	<abbr>DST</abbr> rules.
      </dd>
      <dt><var>time</var></dt><dd>
	takes the form
	'<var>hh</var><code>:</code>[<var>mm</var>[<code>:</code><var>ss</var>]]'
	and defaults to 02:00.
	This is the same format as the offset, except that a
	leading '<code>+</code>' or '<code>-</code>' is not allowed.
      </dd>
      <dt><var>date</var></dt><dd>
	takes one of the following forms:
	<dl>
	  <dt>J<var>n</var> (1&le;<var>n</var>&le;365)</dt><dd>
	    origin-1 day number not counting February 29
	  </dd>
	  <dt><var>n</var> (0&le;<var>n</var>&le;365)</dt><dd>
	    origin-0 day number counting February 29 if present
	  </dd>
	  <dt><code>M</code><var>m</var><code>.</code><var>n</var><code>.</code><var>d</var>
	    (0[Sunday]&le;<var>d</var>&le;6[Saturday], 1&le;<var>n</var>&le;5,
	    1&le;<var>m</var>&le;12)</dt><dd>
	    for the <var>d</var>th day of week <var>n</var> of
	    month <var>m</var> of the year, where week 1 is the first
	    week in which day <var>d</var> appears, and
	    '<code>5</code>' stands for the last week in which
	    day <var>d</var> appears (which may be either the 4th or
	    5th week).
	    Typically, this is the only useful form; the <var>n</var>
	    and <code>J</code><var>n</var> forms are rarely used.
	  </dd>
	</dl>
      </dd>
    </dl>

    <p>
    Here is an example POSIX.1-2017 <code>TZ</code> string for New
    Zealand after 2007.
    It says that standard time (<abbr>NZST</abbr>) is 12 hours ahead
    of <abbr>UT</abbr>, and that daylight saving time
    (<abbr>NZDT</abbr>) is observed from September's last Sunday at
    02:00 until April's first Sunday at 03:00:
    </p>

    <pre><code>TZ='NZST-12NZDT,M9.5.0,M4.1.0/3'</code></pre>

    <p>
    This POSIX.1-2017 <code>TZ</code> string is hard to remember, and
    mishandles some timestamps before 2008.
    With this package you can use this instead:
    </p>

    <pre><code>TZ='Pacific/Auckland'</code></pre>
  </li>
  <li>
    POSIX does not define the <abbr>DST</abbr> transitions
    for <code>TZ</code> values like
    "<code>EST5EDT</code>".
    Traditionally the current <abbr>US</abbr> <abbr>DST</abbr> rules
    were used to interpret such values, but this meant that the
    <abbr>US</abbr> <abbr>DST</abbr> rules were compiled into each
    time conversion package, and when
    <abbr>US</abbr> time conversion rules changed (as in the United
    States in 1987 and again in 2007), all packages that
    interpreted <code>TZ</code> values had to be updated
    to ensure proper results.
  </li>
  <li>
    The <code>TZ</code> environment variable is process-global, which
    makes it hard to write efficient, thread-safe applications that
    need access to multiple timezones.
  </li>
  <li>
    In POSIX, there is no tamper-proof way for a process to learn the
    system's best idea of local (wall clock) time.
    This is important for applications that an administrator wants
    used only at certain times &ndash; without regard to whether the
    user has fiddled the
    <code>TZ</code> environment variable.
    While an administrator can "do everything in <abbr>UT</abbr>" to
    get around the problem, doing so is inconvenient and precludes
    handling daylight saving time shifts &ndash; as might be required to
    limit phone calls to off-peak hours.
  </li>
  <li>
    POSIX.1-2017 provides no convenient and efficient way to determine
    the <abbr>UT</abbr> offset and time zone abbreviation of arbitrary
    timestamps, particularly for timezones
    that do not fit into the POSIX model.
  </li>
  <li>
    POSIX requires that <code>time_t</code> clock counts exclude leap
    seconds.
  </li>
  <li>
    The <code><abbr>tz</abbr></code> code attempts to support all the
    <code>time_t</code> implementations allowed by POSIX.
    The <code>time_t</code> type represents a nonnegative count of seconds
    since 1970-01-01 00:00:00 <abbr>UTC</abbr>, ignoring leap seconds.
    In practice, <code>time_t</code> is usually a signed 64- or 32-bit
    integer; 32-bit signed <code>time_t</code> values stop working after
    2038-01-19 03:14:07 <abbr>UTC</abbr>, so new implementations these
    days typically use a signed 64-bit integer.
    Unsigned 32-bit integers are used on one or two platforms, and 36-bit
    and 40-bit integers are also used occasionally.
    Although earlier POSIX versions allowed <code>time_t</code> to be a
    floating-point type, this was not supported by any practical system,
    and POSIX.1-2013 and the <code><abbr>tz</abbr></code> code both
    require <code>time_t</code> to be an integer type.
  </li>
</ul>

<h3 id="POSIX-extensions">Extensions to POSIX.1-2017 in the
<code><abbr>tz</abbr></code> code</h3>
<ul>
  <li>
    <p>
    The <code>TZ</code> environment variable is used in generating
    the name of a file from which time-related information is read
    (or is interpreted  la POSIX.1-2017); <code>TZ</code> is no longer
    constrained to be a string containing abbreviations
    and numeric data as described <a href="#POSIX">above</a>.
    The file's format is <dfn><abbr>TZif</abbr></dfn>,
    a timezone information format that contains binary data; see
    <a href="https://datatracker.ietf.org/doc/html/8536">Internet
    <abbr>RFC</abbr> 8536</a>.
    The daylight saving time rules to be used for a
    particular timezone are encoded in the
    <abbr>TZif</abbr> file; the format of the file allows <abbr>US</abbr>,
    Australian, and other rules to be encoded, and
    allows for situations where more than two time zone
    abbreviations are used.
    </p>
    <p>
    It was recognized that allowing the <code>TZ</code> environment
    variable to take on values such as '<code>America/New_York</code>'
    might cause "old" programs (that expect <code>TZ</code> to have a
    certain form) to operate incorrectly; consideration was given to using
    some other environment variable (for example, <code>TIMEZONE</code>)
    to hold the string used to generate the <abbr>TZif</abbr> file's name.
    In the end, however, it was decided to continue using
    <code>TZ</code>: it is widely used for time zone purposes;
    separately maintaining both <code>TZ</code>
    and <code>TIMEZONE</code> seemed a nuisance; and systems where
    "new" forms of <code>TZ</code> might cause problems can simply
    use legacy <code>TZ</code> values such as "<code>EST5EDT</code>" which
    can be used by "new" programs as well as by "old" programs that
    assume pre-POSIX <code>TZ</code> values.
    </p>
  </li>
  <li>
    The code supports platforms with a <abbr>UT</abbr> offset member
    in <code>struct tm</code>, e.g., <code>tm_gmtoff</code>,
    or with a time zone abbreviation member in
    <code>struct tm</code>, e.g., <code>tm_zone</code>. As noted
    in <a href="https://austingroupbugs.net/view.php?id=1533">Austin
    Group defect 1533</a>, a future version of POSIX is planned to
    require <code>tm_gmtoff</code> and <code>tm_zone</code>.
  </li>
  <li>
    Functions <code>tzalloc</code>, <code>tzfree</code>,
    <code>localtime_rz</code>, and <code>mktime_z</code> for
    more-efficient thread-safe applications that need to use multiple
    timezones.
    The <code>tzalloc</code> and <code>tzfree</code> functions
    allocate and free objects of type <code>timezone_t</code>,
    and <code>localtime_rz</code> and <code>mktime_z</code> are
    like <code>localtime_r</code> and <code>mktime</code> with an
    extra <code>timezone_t</code> argument.
    The functions were inspired by <a href="https://netbsd.org/">NetBSD</a>.
  </li>
  <li>
    Negative <code>time_t</code> values are supported, on systems
    where <code>time_t</code> is signed.
  </li>
  <li>
    These functions can account for leap seconds;
    see <a href="#leapsec">Leap seconds</a> below.
  </li>
</ul>

<h3 id="vestigial">POSIX features no longer needed</h3>
<p>
POSIX and <a href="https://en.wikipedia.org/wiki/ISO_C"><abbr>ISO</abbr> C</a>
define some <a href="https://en.wikipedia.org/wiki/API"><abbr
title="application programming interface">API</abbr>s</a> that are vestigial:
they are not needed, and are relics of a too-simple model that does
not suffice to handle many real-world timestamps.
Although the <code><abbr>tz</abbr></code> code supports these
vestigial <abbr>API</abbr>s for backwards compatibility, they should
be avoided in portable applications.
The vestigial <abbr>API</abbr>s are:
</p>
<ul>
  <li>
    The POSIX <code>tzname</code> variable does not suffice and is no
    longer needed.
    To get a timestamp's time zone abbreviation, consult
    the <code>tm_zone</code> member if available; otherwise,
    use <code>strftime</code>'s <code>"%Z"</code> conversion
    specification.
  </li>
  <li>
    The POSIX <code>daylight</code> and <code>timezone</code>
    variables do not suffice and are no longer needed.
    To get a timestamp's <abbr>UT</abbr> offset, consult
    the <code>tm_gmtoff</code> member if available; otherwise,
    subtract values returned by <code>localtime</code>
    and <code>gmtime</code> using the rules of the Gregorian calendar,
    or use <code>strftime</code>'s <code>"%z"</code> conversion
    specification if a string like <code>"+0900"</code> suffices.
  </li>
  <li>
    The <code>tm_isdst</code> member is almost never needed and most of
    its uses should be discouraged in favor of the abovementioned
    <abbr>API</abbr>s.
    Although it can still be used in arguments to
    <code>mktime</code> to disambiguate timestamps near
    a <abbr>DST</abbr> transition when the clock jumps back on
    platforms lacking <code>tm_gmtoff</code>, this
    disambiguation does not work when standard time itself jumps back,
    which can occur when a location changes to a time zone with a
    lesser <abbr>UT</abbr> offset.
  </li>
</ul>

<h3 id="other-portability">Other portability notes</h3>
<ul>
  <li>
    The <a href="https://en.wikipedia.org/wiki/Version_7_Unix">7th Edition
    UNIX</a> <code>timezone</code> function is not present in this
    package; it is impossible to reliably map <code>timezone</code>'s
    arguments (a "minutes west of <abbr>GMT</abbr>" value and a
    "daylight saving time in effect" flag) to a time zone
    abbreviation, and we refuse to guess.
    Programs that in the past used the <code>timezone</code> function
    may now examine <code>localtime(&amp;clock)-&gt;tm_zone</code>
    (if <code>TM_ZONE</code> is defined) or
    <code>tzname[localtime(&amp;clock)-&gt;tm_isdst]</code>
    (if <code>HAVE_TZNAME</code> is nonzero) to learn the correct time
    zone abbreviation to use.
  </li>
  <li>
    The <a
    href="https://en.wikipedia.org/wiki/History_of_the_Berkeley_Software_Distribution#4.2BSD"><abbr>4.2BSD</abbr></a>
    <code>gettimeofday</code> function is not
    used in this package.
    This formerly let users obtain the current <abbr>UTC</abbr> offset
    and <abbr>DST</abbr> flag, but this functionality was removed in
    later versions of <abbr>BSD</abbr>.
  </li>
  <li>
    In <abbr>SVR2</abbr>, time conversion fails for near-minimum or
    near-maximum <code>time_t</code> values when doing conversions
    for places that do not use <abbr>UT</abbr>.
    This package takes care to do these conversions correctly.
    A comment in the source code tells how to get compatibly wrong
    results.
  </li>
  <li>
    The functions that are conditionally compiled
    if <code>STD_INSPIRED</code> is nonzero should, at this point, be
    looked on primarily as food for thought.
    They are not in any sense "standard compatible" &ndash; some are
    not, in fact, specified in <em>any</em> standard.
    They do, however, represent responses of various authors to
    standardization proposals.
  </li>
  <li>
    Other time conversion proposals, in particular those supported by the
    <a href="https://howardhinnant.github.io/date/tz.html">Time Zone
    Database Parser</a>, offer a wider selection of functions
    that provide capabilities beyond those provided here.
    The absence of such functions from this package is not meant to
    discourage the development, standardization, or use of such
    functions.
    Rather, their absence reflects the decision to make this package
    contain valid extensions to POSIX, to ensure its broad
    acceptability.
    If more powerful time conversion functions can be standardized, so
    much the better.
  </li>
</ul>
</section>

<section>
  <h2 id="stability">Interface stability</h2>
<p>
The <code><abbr>tz</abbr></code> code and data supply the following interfaces:
</p>

<ul>
  <li>
    A set of timezone names as per
      "<a href="#naming">Timezone identifiers</a>" above.
  </li>
  <li>
    Library functions described in "<a href="#functions">Time and date
      functions</a>" above.
  </li>
  <li>
    The programs <code>tzselect</code>, <code>zdump</code>,
    and <code>zic</code>, documented in their man pages.
  </li>
  <li>
    The format of <code>zic</code> input files, documented in
    the <code>zic</code> man page.
  </li>
  <li>
    The format of <code>zic</code> output files, documented in
    the <code>tzfile</code> man page.
  </li>
  <li>
    The format of zone table files, documented in <code>zone1970.tab</code>.
  </li>
  <li>
    The format of the country code file, documented in <code>iso3166.tab</code>.
  </li>
  <li>
    The version number of the code and data, as the first line of
    the text file '<code>version</code>' in each release.
  </li>
</ul>

<p>
Interface changes in a release attempt to preserve compatibility with
recent releases.
For example, <code><abbr>tz</abbr></code> data files typically do not
rely on recently added <code>zic</code> features, so that users can
run older <code>zic</code> versions to process newer data files.
<a href="tz-link.html#download">Downloading
the <code><abbr>tz</abbr></code> database</a> describes how releases
are tagged and distributed.
</p>

<p>
Interfaces not listed above are less stable.
For example, users should not rely on particular <abbr>UT</abbr>
offsets or abbreviations for timestamps, as data entries are often
based on guesswork and these guesses may be corrected or improved.
</p>

<p>
Timezone boundaries are not part of the stable interface.
For example, even though the <samp>Asia/Bangkok</samp> timezone
currently includes Chang Mai, Hanoi, and Phnom Penh, this is not part
of the stable interface and the timezone can split at any time.
If a calendar application records a future event in some location other
than Bangkok by putting "<samp>Asia/Bangkok</samp>" in the event's record,
the application should be robust in the presence of timezone splits
between now and the future time.
</p>
</section>

<section>
  <h2 id="leapsec">Leap seconds</h2>
<p>
Leap seconds were introduced in 1972 to accommodate the
difference between atomic time and the less regular rotation of the earth.
Unfortunately they caused so many problems with civil
timekeeping that they
are <a href="https://www.bipm.org/en/cgpm-2022/resolution-4">planned
to be discontinued by 2035</a>, with some as-yet-undetermined
mechanism replacing them, perhaps after the year 2135.
Despite their impending obsolescence, a record of leap seconds is still
needed to resolve timestamps from 1972 through 2035.
</p>

<p>
The <code><abbr>tz</abbr></code> code and data can account for leap seconds,
thanks to code contributed by Bradley White.
However, the leap second support of this package is rarely used directly
because POSIX requires leap seconds to be excluded and many
software packages would mishandle leap seconds if they were present.
Instead, leap seconds are more commonly handled by occasionally adjusting
the operating system kernel clock as described in
<a href="tz-link.html#precision">Precision timekeeping</a>,
and this package by default installs a <samp>leapseconds</samp> file
commonly used by
<a href="https://www.ntp.org"><abbr title="Network Time Protocol">NTP</abbr></a>
software that adjusts the kernel clock.
However, kernel-clock twiddling approximates UTC only roughly,
and systems needing more precise UTC can use this package's leap
second support directly.
</p>

<p>
The directly supported mechanism assumes that <code>time_t</code>
counts of seconds since the POSIX epoch normally include leap seconds,
as opposed to POSIX <code>time_t</code> counts which exclude leap seconds.
This modified timescale is converted to <abbr>UTC</abbr>
at the same point that time zone and <abbr>DST</abbr>
adjustments are applied &ndash;
namely, at calls to <code>localtime</code> and analogous functions &ndash;
and the process is driven by leap second information
stored in alternate versions of the <abbr>TZif</abbr> files.
Because a leap second adjustment may be needed even
if no time zone correction is desired,
calls to <code>gmtime</code>-like functions
also need to consult a <abbr>TZif</abbr> file,
conventionally named <samp><abbr>Etc/UTC</abbr></samp>
(<samp><abbr>GMT</abbr></samp> in previous versions),
to see whether leap second corrections are needed.
To convert an application's <code>time_t</code> timestamps to or from
POSIX <code>time_t</code> timestamps (for use when, say,
embedding or interpreting timestamps in portable
<a href="https://en.wikipedia.org/wiki/Tar_(computing)"><code>tar</code></a>
files),
the application can call the utility functions
<code>time2posix</code> and <code>posix2time</code>
included with this package.
</p>

<p>
If the POSIX-compatible <abbr>TZif</abbr> file set is installed
in a directory whose basename is <samp>zoneinfo</samp>, the
leap-second-aware file set is by default installed in a separate
directory <samp>zoneinfo-leaps</samp>.
Although each process can have its own time zone by setting
its <code>TZ</code> environment variable, there is no support for some
processes being leap-second aware while other processes are
POSIX-compatible; the leap-second choice is system-wide.
So if you configure your kernel to count leap seconds, you should also
discard <samp>zoneinfo</samp> and rename <samp>zoneinfo-leaps</samp>
to <samp>zoneinfo</samp>.
Alternatively, you can install just one set of <abbr>TZif</abbr> files
in the first place; see the <code>REDO</code> variable in this package's
<a href="https://en.wikipedia.org/wiki/Makefile">makefile</a>.
</p>
</section>

<section>
  <h2 id="calendar">Calendrical issues</h2>
<p>
Calendrical issues are a bit out of scope for a time zone database,
but they indicate the sort of problems that we would run into if we
extended the time zone database further into the past.
An excellent resource in this area is Edward M. Reingold
and Nachum Dershowitz, <cite><a
href="https://www.cambridge.org/fr/academic/subjects/computer-science/computing-general-interest/calendrical-calculations-ultimate-edition-4th-edition">Calendrical
Calculations: The Ultimate Edition</a></cite>, Cambridge University Press (2018).
Other information and sources are given in the file '<code>calendars</code>'
in the <code><abbr>tz</abbr></code> distribution.
They sometimes disagree.
</p>
</section>

<section>
  <h2 id="planets">Time and time zones off Earth</h2>
<p>
The European Space Agency is <a
href='https://www.esa.int/Applications/Navigation/Telling_time_on_the_Moon'>considering</a>
the establishment of a reference timescale for the Moon, which has
days roughly equivalent to 29.5 Earth days, and where relativistic
effects cause clocks to tick slightly faster than on Earth.
</p>

<p>
Some people's work schedules have used
<a href="https://en.wikipedia.org/wiki/Timekeeping_on_Mars">Mars time</a>.
Jet Propulsion Laboratory (JPL) coordinators kept Mars time on
and off during the
<a href="https://en.wikipedia.org/wiki/Mars_Pathfinder">Mars
Pathfinder</a> mission (1997).
Some of their family members also adapted to Mars time.
Dozens of special Mars watches were built for JPL workers who kept
Mars time during the
<a href="https://en.wikipedia.org/wiki/Mars_Exploration_Rover">Mars
Exploration Rovers (MER)</a> mission (2004&ndash;2018).
These timepieces looked like normal Seikos and Citizens but were adjusted
to use Mars seconds rather than terrestrial seconds, although
unfortunately the adjusted watches were unreliable and appear to have
had only limited use.
</p>

<p>
A Mars solar day is called a "sol" and has a mean period equal to
about 24 hours 39 minutes 35.244 seconds in terrestrial time.
It is divided into a conventional 24-hour clock, so each Mars second
equals about 1.02749125 terrestrial seconds.
(One MER worker noted, "If I am working Mars hours, and Mars hours are
2.5% more than Earth hours, shouldn't I get an extra 2.5% pay raise?")
</p>

<p>
The <a href="https://en.wikipedia.org/wiki/Prime_meridian">prime
meridian</a> of Mars goes through the center of the crater
<a href="https://en.wikipedia.org/wiki/Airy-0">Airy-0</a>, named in
honor of the British astronomer who built the Greenwich telescope that
defines Earth's prime meridian.
Mean solar time on the Mars prime meridian is
called Mars Coordinated Time (<abbr>MTC</abbr>).
</p>

<p>
Each landed mission on Mars has adopted a different reference for
solar timekeeping, so there is no real standard for Mars time zones.
For example, the MER mission defined two time zones "Local
Solar Time A" and "Local Solar Time B" for its two missions, each zone
designed so that its time equals local true solar time at
approximately the middle of the nominal mission.
The A and B zones differ enough so that an MER worker assigned to
the A zone might suffer "Mars lag" when switching to work in the B zone.
Such a "time zone" is not particularly suited for any application
other than the mission itself.
</p>

<p>
Many calendars have been proposed for Mars, but none have achieved
wide acceptance.
Astronomers often use Mars Sol Date (<abbr>MSD</abbr>) which is a
sequential count of Mars solar days elapsed since about 1873-12-29
12:00 <abbr>GMT</abbr>.
</p>

<p>
In our solar system, Mars is the planet with time and calendar most
like Earth's.
On other planets, Sun-based time and calendars would work quite
differently.
For example, although Mercury's
<a href="https://en.wikipedia.org/wiki/Rotation_period">sidereal
rotation period</a> is 58.646 Earth days, Mercury revolves around the
Sun so rapidly that an observer on Mercury's equator would see a
sunrise only every 175.97 Earth days, i.e., a Mercury year is 0.5 of a
Mercury day.
Venus is more complicated, partly because its rotation is slightly
<a href="https://en.wikipedia.org/wiki/Retrograde_motion">retrograde</a>:
its year is 1.92 of its days.
Gas giants like Jupiter are trickier still, as their polar and
equatorial regions rotate at different rates, so that the length of a
day depends on latitude.
This effect is most pronounced on Neptune, where the day is about 12
hours at the poles and 18 hours at the equator.
</p>

<p>
Although the <code><abbr>tz</abbr></code> database does not support
time on other planets, it is documented here in the hopes that support
will be added eventually.
</p>

<p>
Sources for time on other planets:
</p>

<ul>
  <li>
    Michael Allison and Robert Schmunk,
    "<a href="https://www.giss.nasa.gov/tools/mars24/help/notes.html">Technical
      Notes on Mars Solar Time as Adopted by the Mars24 Sunclock</a>"
    (2020-03-08).
  </li>
  <li>
    Zara Mirmalek,
    <em><a href="https://mitpress.mit.edu/books/making-time-mars">Making
	Time on Mars</a></em>, MIT Press (March 2020), ISBN 978-0262043854.
  </li>
  <li>
    Jia-Rui Chong,
    "<a href="https://www.latimes.com/archives/la-xpm-2004-jan-14-sci-marstime14-story.html">Workdays
    Fit for a Martian</a>", <cite>Los Angeles Times</cite>
    (2004-01-14), pp A1, A20&ndash;A21.
  </li>
  <li>
    Tom Chmielewski,
    "<a href="https://www.theatlantic.com/technology/archive/2015/02/jet-lag-is-worse-on-mars/386033/">Jet
    Lag Is Worse on Mars</a>", <cite>The Atlantic</cite> (2015-02-26)
  </li>
  <li>
    Matt Williams,
    "<a href="https://www.universetoday.com/37481/days-of-the-planets/">How
    long is a day on the other planets of the solar system?</a>"
    (2016-01-20).
  </li>
</ul>
</section>

<footer>
  <hr>
  This file is in the public domain, so clarified as of 2009-05-17 by
  Arthur David Olson.
</footer>
</body>
</html>