marsdate 1.1.7 → 2.0.1

data/README.md

@@ -0,0 +1,650 @@
+# MarsDate
+
+##   A Ruby date/time library for Mars
+
+This library is based on the Martian Common Era calendar created by Hal Fulton.
+Its functionality closely follows that of Ruby's Time class.
+
+**Clock (2.0).** `MarsDateTime` stores one MSD. 00:00 is Airy-0 mean
+midnight (**MTC**). **MXT** is the same instant in SI hours (day to
+~24:39). Civil constructors: `mxt(...)` / `new(y,m,sol,h,min,sec)`
+(MXT), `mtc(...)`, `at(msd)`. Views: `md.mxt` / `md.mtc`. `format(fmt)` is date + MXT;
+`format_mtc(fmt)` / `md.mtc.format` for MTC. `%Z` is the clock
+name (MXT or MTC), not a timezone. No `strftime`.
+1/1/1 is the sol that contains the year-1 northern vernal equinox
+(`EPOCH_MSD = −665773`). Earth time of that midnight is a caption
+(~0001-01-23 13:40, treating TT as UT), not the converter’s input.
+Allison’s Ls series is fitted ~1874–2127; year 1 is an extrapolation.
+A historical ΔT at 1 CE is a few hours and would only slide the
+caption, so we do not model it. `MarsDateTime.today` is the current
+MCE sol at 00:00; `now` is the current instant.
+`MarsDateTime.epoch_caption` returns the year-1 numbers and caveat.
+Design: [`design-issues.md`](design-issues.md).
+
+Tests (Minitest): `ruby test/timescale_test.rb`,
+`ruby test/calendar_test.rb`, `ruby test/test.rb`,
+`ruby test/cli_test.rb`.
+Changelog: [`CHANGELOG.md`](CHANGELOG.md).
+
+## Installation
+
+```sh
+gem install marsdate
+```
+
+## Development and release
+
+Install the development dependencies and run the complete test suite:
+
+```sh
+bundle install
+bundle exec rake
+```
+
+For a release, update `MarsDateTime::VERSION` in
+`lib/marsdate/version.rb` and the changelog, commit those changes, and run
+`./bump-gem` from a clean working tree. The script tests, builds, and verifies
+the packaged CLI without publishing. Review the resulting gem and publish the
+exact path printed by the script.
+
+## API (2.0)
+
+```ruby
+require 'marsdate'
+
+md = MarsDateTime.mxt(1069, 3, 9, 9, 44, 24)
+md.format('%A, %B %e, %Y %H:%M:%S %Z')
+# => "Sunday, M-February  9, 1069 09:44:24 MXT"
+md.format_mtc('%H:%M:%S %Z')
+# => "09:28:45 MTC"
+
+earth = DateTime.new(2010, 1, 1, 0, 0, 0)
+MarsDateTime.new(earth).to_s
+# Sunday, M-February 9, 1069 at MXT 09:44:24… / MTC 09:28:45…
+
+MarsDateTime.today   # current MCE sol at 00:00
+MarsDateTime.now     # current instant
+```
+
+Earth `DateTime` is the instant (`ajd`). `earth_date` comes back UTC.
+
+<pre>
+self.leap?(myear)  # class method for convenience
+self.short?(myear) # short year
+self.long?(myear)  # long year
+self.sols_in_month(m, year)
+self.now              # current instant
+self.today            # current MCE sol at 00:00
+self.epoch_caption    # year-1 Earth captions (TT as UT)
+
+leap?
+short?
+long?
+month_name
+
+initialize(params)
+  3-6: y m sol [h min sec]   # HMS is MXT
+    0: (now)
+    1: MSD or Date or DateTime
+
+self.mxt(y, m, sol, h=0, min=0, sec=0)
+self.mtc(y, m, sol, h=0, min=0, sec=0)
+self.at(msd)            # from_msd is an alias
+md.mxt / md.mtc         # views: date delegates, hour/min/sec, to_s, format
+format(fmt)             # date + MXT (%H). Inspired by Time#strftime.
+format_mxt / format_mtc # no arg: HMS; with fmt: same table, that clock
+md.mxt.format(fmt) / md.mtc.format(fmt)
+
+-(other)    # MarsDateTime, Date, DateTime, Integer, Float
++(sols)
+<=>(other)  # MarsDateTime, Date, or DateTime
+
+earth_date
+
+format(fmt)     # Specifiers (inspired by Time#strftime):
+
+  %a  @day_of_week[0..2]
+  %A  @day_of_week
+  %b  (@month.odd? ? month_name[2..4] : month_name[0..2])
+  %B  month_name
+  %d  zsol
+  %e  ('%2d' % @sol)
+  %F  "#@year-#{zmonth}-#{zsol}"
+  %H  zhh
+  %j  @year_sol.to_s
+  %m  zmonth  # @month.to\_s
+  %M  zmm
+  %s  seconds of this clock (integer)
+  %S  zss   # seconds of this clock
+  %u  (@dow + 1).to_s
+  %U  (@year_sol/7 + 1).to_s
+  %w  @dow.to_s
+  %x  "#@year/#{zmonth}/#{zsol}"
+  %X  "#{zhh}:#{zmm}:#{zss}"
+  %Y  @year.to_s
+  %Z  clock name (MXT or MTC), not a timezone
+  %n  "\n"
+  %t  "\t"
+  %%  "%"
+</pre>
+
+
+### The Martian Common Era calendar
+
+Read more about this below.
+
+### More to come...
+
+What needs doing here:
+  - Restructure the project tree
+  - Tests are Minitest (`ruby test/test.rb`); no shoulda or rspec
+  - Clean up the source
+  - Add some executables (mcal, etc.)
+  - Add extensive documentation to this README
+  
+<center>
+<h2>The Martian Common Era Calendar</h2>
+<br>
+Hal E. Fulton
+<br>
+<br>
+</center>
+
+<p>
+It may seem odd, but people have been designing Martian calendars
+for over a century now, despite the fact that not even one human
+has set foot there yet. By one count, over <i>seventy-five</i> 
+different calendars have been created for the red planet. One 
+reason for this, of course, is that it's a challenging and fun 
+problem.
+
+<p>
+But it goes beyond that. We are closer than ever to a human presence
+on Mars; and it is never too early to think about such details.
+
+<p>
+In this "me too" spirit, I give you the <i>Martian Common Era</i>
+calendar (MCE). Like many, it is a variant of Darian Calendar of 
+Thomas Gangale, perhaps the best known calendar for Mars.
+
+<p>
+The purpose for this particular calendar, however, is more than just
+a proposal for later use. It's also for fun. Want to celebrate the
+Martian New Year? You can, if you and your friends agree on when it
+is. Want to celebrate your Martian birthday? Those are special because
+they are twice as rare as regular birthdays.
+
+<p>
+The purpose of this article isn't to give exhaustive technical details.
+If you are interested in Martian calendars, I can refer you to the 
+work of Thomas Gangale (mentioned previously). Others have done 
+interesting work as well &mdash; a web search will give you a wealth of
+material within minutes.
+
+<p>
+Bear in mind when I talk in this article about "Earth's calendar," I am
+being a little unfair. We have had many different calendars in our history,
+and there are different ones in use even today. The calendar I refer to as 
+the "Earthly" calendar is of course the Gregorian, probably the nearest 
+thing to a universal calendar that humans have ever had.
+
+<p>
+I'll give you just a little background information here. Consider this
+a short introduction to Mars and to calendars in general.
+
+<p>
+Mars is farther out from the sun than the earth is, and so it travels
+around the sun more slowly. In fact, a Martian year is nearly twice as
+long as one of ours &mdash; twenty-three of our months, give or take a little.
+
+<p>
+The Martian day, on the other hand, is very nearly the same as an earthly
+day &mdash; roughly 24 hours and 39 minutes. A day on Mars is called a 
+<i>sol</i> (as in <i>solar</i>)to distinguish it from an earthly day. This
+usage came about in the 1970s when NASA first placed a lander on the Martian
+surface; the scientists on the Viking lander team naturally needed to talk
+about the "day and night" patterns of the world they were studying.
+
+<p>
+There are some people who want to use alternate terms for a Martian year. 
+These terms include <i>mir</i> and <i>orb</i> (as in <i>orbit</i>). Others, 
+like me, are content to say "year" both for Earth and for Mars (and may 
+sometimes disambiguate by saying "Martian year," "Mars year," "M-year," 
+or something similar).
+
+<p>
+One thing that makes calendars hard (even on Earth) is that the year is not
+made up of an exact number of days. When the Earth orbits for a full 365
+days, it is still almost 6 hours short of going all the way around the sun.
+That, of course, is why we add a leap year roughly every four years.
+
+<p>
+Mars, on the other hand, orbits the sun in about 687 earth days or 668 sols.
+But again, when the 668 sols have passed, there is still a fraction of a sol
+left before it goes all the way around the sun. In this case, the fraction 
+is nearer to a half, so it makes sense for a Martian calendar to have a leap
+year almost every other year.
+
+<p>
+So these facts are pretty much beyond our control. Things that are more within
+our control are: Where do we start counting? How do we group the days together
+into months, and what do we name them? Do we keep a seven-day week? Where do 
+we put the leap day when we need it? 
+
+<p>
+First, let's look at the question of where the year starts. On Earth, the day
+most of us call "January 1" is considered to be the first day of the year. 
+To understand why is beyond the scope of this article. Suffice to say that it
+has more to do with history, religion, and politics than it has to do with 
+astronomy or any other science. 
+
+<p>
+On Mars, we could pick any day to be the start. One <i>logical</i> choice might
+be the first day of spring; after all, we think of spring as a time of "new
+beginnings" and so on. As a matter of fact, this choice is widespread among
+planetary scientists (for Mars and for other planets as well). In addition, most
+Martian calendars have made that same decision. So let's stick with that.
+
+<p>
+The astute reader may ask: <i>Which</i> vernal equinox? After all, the northern
+hemisphere's vernal (or spring) equinox is the same as the southern hemisphere's
+autumnal (fall) equinox. But here, as in many things, there is an arbitrary 
+convention; we assume that the <i>northern</i> hemisphere is the standard. When
+we draw a typical diagram of the solar system, we are looking downward onto the
+Sun's north pole (and that of the Earth and Mars as well). So the first day of
+the Martian year will be the spring equinox in the northern hemisphere, but the
+<i>fall</i> equinox in the southern hemisphere.
+
+<p>
+Next let's look at the question of months. We divide time on Earth into months
+because that is roughly the length of the lunar cycle. (Of course, once again
+Nature does not cooperate &mdash; the number of lunar months in a year is not 
+exactly twelve.)
+
+<p>
+But on Mars, Earth's lunar cycle is pretty meaningless. Yet we have grown used to
+the convenience of dividing a year into months &mdash; a period much larger 
+than a day but much smaller than a year.
+
+<p>
+It's not <i>necessary</i> to divide a year into months at all, of course. One 
+possible Martian calendar would just number the days from 1 to 668 (or 669 in a 
+leap year). This has the virtue of simplicity, but it does sound rather boring.
+To date a letter "Day 226" makes it sound like an entry from a prisoner's diary.
+Birthdays would also sound boring &mdash; "My birthday is 348, and hers is 529."
+Very practical, maybe, but not much fun.
+
+<p>
+It's tempting to look to the moons of Mars as a way of dividing a year into 
+months. Unfortunately, these are much smaller than ours and orbit much more 
+closely &mdash; so they move very fast. The outer moon Deimos goes around Mars 
+in less than a day and a half. The inner moon Phobos actually orbits in less 
+than 8 hours. So they're completely unsuitable for this purpose.
+
+<p>
+The usual approach is to make the months more or less arbitrary. Some Martian 
+calendars have divided the year into 12 months (as we are accustomed to); the 
+disadvantage is that the months then have more than 50 days each, much more 
+than we are used to. The other common choice is 24 months that are very near in 
+length to our earthly months. One problem then, of course, is naming the months. 
+
+<p>
+At least one calendar has actually compromised on a sixteen-month year, with 
+each month being 40-something days in length. This is an interesting and 
+innovative approach, but it doesn't appeal to me so much.
+
+<p>
+So I decided to go the "conventional" route and have 24 months. The next 
+question obviously is: How do we name them? 
+
+<p>
+Some have suggested a scheme where there is a "First January" and a "Second 
+January" (and so on for all the months). Others have suggested elaborate 
+culturally-neutral naming schemes.
+
+<p>
+My choice was to use the conventional twelve month names and intersperse the 
+names of the signs of the zodiac. This solution is not "culturally neutral" as 
+such; but virtually every language in the world has names for the twelve months 
+and the twelve signs. For me, that is close enough.
+
+<p>
+So I start with the usual "January" and then follow with Gemini. Why Gemini, 
+you may ask? Well, if you view the solar system as a clock with the sun at the 
+center, the constellations are arranged around it so that Mars is "in" Gemini 
+at the time of its vernal equinox. 
+
+<p>
+This, by the way, is a <i>heliocentric</i> or sun-centered viewpoint. It is 
+therefore 180 degrees opposite to the approach followed by ancient astronomers, 
+who thought according to an Earth-centric standard, calculating in which 
+constellation the sun fell. By that kind of thinking, Martians would see the 
+sun in Sagittarius at their vernal equinox, so we could name that month 
+Sagittarius; but now that we know that the planets orbit the sun (not the 
+Earth!), let us think and act accordingly.
+
+<p>
+By now, you may be thinking of the old rhyme, "Thirty days hath September..." 
+How many days should each Martian month have?
+
+<p>
+There is more than one way to partition them, of course. We could adjust them 
+so that they fit the Martian seasons; these are "lopsided" or irregular because 
+of the highly elliptical orbit Mars has.
+
+<p>
+I prefer a simpler approach. Let's give each of them 28 days. The exception, of 
+course, will be the final month of Taurus, which will have only 24 (giving a 
+total of 668 sols). When there is a leap year, I would add the day naturally to 
+the end of the year, giving Taurus a 25th day (rather than adding to the end of 
+the second month as we do on Earth).
+
+<p>
+Wait a minute! Leap years? Yes, as I mentioned before, they are an issue on 
+Mars &mdash; more so than on Earth, in fact.
+
+<p>
+You may think leap years in our Earthly calendar are "easy." In a sense, they 
+are. But there are more rules than most people realize!
+
+<p>
+The Earth orbits the sun in roughly 365.2422 days &mdash; the leftover 0.2422 
+days, of course, is why we have leap years. In four years, we accumulate 0.9688 
+days, which is very nearly an entire day, so we add a day to "re-synchronize." 
+So far, it's easy.
+
+<p>
+But this is a little too much correction. We've over-corrected by about 0.03 
+days (or about 42 minutes). So the next rule is: When the year is a multiple of 
+100 (like 1800 or 1900), we will <i>skip</i> a leap year. I won't do the math, 
+but this results in a slight under-correction. We then add yet another rule: If 
+a century year (a multiple of 100) is <i>also</i> a multiple of 400, then it 
+<i>will</i> be a leap year. This is why 1800 and 1900 were not leap years, but 
+2000 was. 
+
+<p>
+The bad news is, Martian leap years will be just as complex. The good news is, 
+we can manage that complexity much the way we do with our Earthly calendar.
+
+<p>
+I'll spare you the math. We'll use the same scheme Thomas Gangale used with his 
+Darian calendar. Essentially, each decade will have six leap years and four 
+non-leap years; the leap years will have 669 sols and the others will have 668.
+
+<p>
+As a side note: We are accustomed to leap years occurring fairly rarely &mdash; 
+but in this calendar, leap years are actually <i>more common</i> than non-leap
+years. For this reason, I recommend introducing the more intuitive terms "short
+year" and "long year." A <i>long year</i> is then the same as a leap year, of
+course.
+
+<p>
+The easy way to allocate these leap years is to say: Odd numbered years have an 
+odd number of days (669), and even-numbered years will have an even number of 
+days (668). Since this is only five leap years per decade, we will add the rule 
+that a multiple of 10 is a leap year.
+
+<p>
+To make the math come out better, we'll have to add two more rules: A century 
+year (multiple of 100) is <i>not</i> a leap year, whereas a millennial year 
+(multiple of 1000) <i>is</i> a leap year.
+
+<p>
+This raises yet another question: How do we number the years? In calendar 
+jargon: When is our <i>epoch</i>? On Earth, we number according to a scheme 
+devised by a medieval monk who tried to estimate the birth year of Jesus, using 
+that for Year 1. (Most authorities believe he was off by at least four years.) 
+This is the Common Era, now commonly abbreviated "C.E." (as in 2010 C.E.); for 
+centuries, of course, the abbreviation was "A.D." (<i>Anno Domini</i> or "the 
+year of our Lord").
+
+<p>
+For Mars, there are several logical choices. One common suggestion has been to 
+use the date of the first robotic landing on Mars (by the Viking lander, in 
+1976). This has the disadvantage that many common dates (such as the year of 
+someone's birth or some twentieth century event) might be negative dates. As of 
+early 2009, it has been roughly eighteen Martian years since Viking landed.
+
+<p>
+Another choice, used in the Darian calendar, is to mark the first year that 
+Mars was observed in a telescope, the year we call 1609. In this system, the 
+Earthly year 2008 would roughly correspond to (or overlap) the Martian year 212.
+
+<p>
+The choice I made was to anchor the Martian epoch as close to our accustomed 
+one as possible (hence the name "Martian Common Era"). As it happens, Mars had 
+a vernal equinox about three weeks into January of the year we call the Year 1. 
+(Our calendar has no Year 0, which explains why the 21st century actually began 
+on January 1, 2001).
+
+<p>
+So let's call this early equinox the Martian epoch. Choosing the epoch near our
+accustomed one gives the Martian calendar a moderately interesting property: If 
+you take the ratio of the Earthly and Martian year numbers, you will get a 
+familiar ratio. It is 1.88 &mdash; the same as the ratio of the length of a 
+Martian year to the length of an Earthly year. If you think about this for a 
+moment, it becomes obvious. While Mars orbits the sun 100 times, the Earth will 
+orbit the sun about 188 times.
+
+<p>
+This, of course, makes it possible to quickly estimate a Martian year from an
+Earthly one, or vice versa. Since Mars orbits the sun in 1.88 of our years,
+the Martian year 1000 would be approximately our year 1880.
+
+<p>
+So we know where our calendar starts and where each year starts. We know the 
+names of the months and how long each year is. We know when the leap years are.
+
+<p>
+One question we haven't asked is: How do we deal with days of the week?
+
+<p>
+Again, there are multiple possibilities. One possibility is to forget them 
+altogether &mdash; throw away the idea of a "week" as a division of time. But 
+in nearly all countries and cultures, we have been used to them for a long time.
+
+<p>
+Another possibility is to use a week that is some length other than 7 days (or 
+sols in this case). This is not unheard of even in Earthly calendars; but 
+again, most people worldwide, in recent centuries, recognize a seven-day week.
+
+<p>
+Some Martian calendars use weeks of seven sols or other lengths. Most of them 
+use sophisticated or colorful naming systems for the sake of cultural 
+neutrality, novelty, or other reasons.
+
+<p>
+In keeping with my theme of simplicity and intuitiveness, I have decided to 
+keep the "normal" seven-day cycle. Likewise, I will keep the ordinary names, 
+which after all may be translated into many different languages.
+
+<p>
+Some Martian calendars have the days of the week fixed with respect to the 
+year. That is, the year always starts on the same day of the week; and thus 
+(for example) the 20th day of the third month will always be the same day of 
+the week from year to year.
+
+<p>
+This is appealing in many ways. But personally I have two small issues with it; 
+first of all, the end of the year introduces a "jump" where, for example, a 
+Wednesday on the last day of the year might be followed immediately by a Sunday 
+on the first day of the next year. I find the discontinuity to be a little 
+jarring.
+
+<p>
+Second of all &mdash; and some will disagree &mdash; I don't entirely like the 
+concept of a date occurring on the same weekday every year. I find it to be a 
+little boring.  Do you want your birthday to fall on Monday <i>every year</i>? 
+Should New Year's Day (or the first sol of spring) <i>always</i> be on a 
+Sunday? Let's have some variety instead. Let's start the cycle at some known 
+point and let it stretch forward and backward, the seven-day cycle running 
+independently of the rest of the calendar, more or less as it does on Earth.
+
+<p>
+Of course, the Martian Common Era calendar <i>does</i> offer a few small 
+advantages in that respect. Notice that each month (except the last one) is 28 
+days long, which is a multiple of 7. This means that the first of each month 
+will be the same day of the week for the entire year. If January 1 starts the 
+year off on a Tuesday, then Gemini 1, February 1, and all the others will also 
+fall on a Tuesday. In fact, each month except Taurus will have the same pattern 
+of weekdays all year long. (Even Taurus will follow that pattern until it runs 
+out of days.)
+
+<p>
+Let's look at terminology a little further. Obviously there is some ambiguity
+if we don't specify which calendar or planet we are talking about. If I say 
+"the month of Gemini," it's clear I mean Mars; but if I say "the month of 
+April," it is unclear. For that reason, I suggest the convention of an 
+optional "M" with a hyphen to be used in front of a name, a unit, or a number 
+as needed.
+
+<p>
+So I might say "Martian April" or "M-April"; I might refer to "20 M-years" 
+(meaning 37 Earthly years). I might refer to "the M-year 1071" or even "the
+year M-1071"; however, in the case of years, context will usually be enough to
+distinguish the two. To be formal about year, we can always say "the year
+1071 MCE."
+
+<p>
+So given what we have so far, we can calculate (at Earth midnight UTC)
+that January 1, 2010 was Sunday, M-February 9, 1069 MCE — MXT 09:44:24 /
+MTC 09:28:45. For another example: July 4, 2011 at 00:00 UTC is Tuesday,
+the 11th of Aries (still M-year 1069) — MXT 17:24:43 / MTC 16:56:46.
+September 13, 2011 at 00:00 UTC is Monday, Taurus 24, 1069 — MXT 19:53:11 /
+MTC 19:21:16. The next Martian New Year (1/1/1070 MCE at official midnight)
+falls on September 14, 2011, 05:25:58 UTC.
+
+<p>
+A Martian date still <i>overlaps</i> an Earthly date: even on Earth we
+don't agree on what day it is &mdash; for people in the USA, it's already
+"tomorrow" in some other parts of the world. The library no longer uses
+the old “same calendar day” pairing. It stores one MSD; 00:00 is Airy-0
+mean midnight (MTC). MXT is the same instant in SI hours. Timezones on
+Earth (and longitude on Mars) can still shift the Earth caption by a day.
+
+<p>
+A natural question is: What holidays might be associated with this kind of 
+calendar? Well &mdash; we can't predict what religious or civil significance
+(if any) may be attached to various dates on Mars. I assume that when there are
+humans on Mars, they will take care of such matters as they arise, regardless
+of what calendar they use. 
+
+<p>
+Take Christmas, for example. Conceivably we could pick a date in history and
+translate it to the Martian reckoning; but what date would we use? Our 
+December 25 falls on a different Martian day each M-year. Furthermore, we have
+never been certain of the year of the birth of Jesus, much less the exact date.
+(December 25 is purely a convenient fiction &mdash; the date was taken from
+the old Roman holiday of Saturnalia as a matter of convenience.)
+
+<p>
+So let's ignore things such as Christmas and Ramadan, as well as the numerous
+other holidays such as Thanksgiving and Mother's Day. This still leaves two
+categories of holidays we might celebrate &mdash; those based on astronomy and
+those based on modern events relating to Mars.
+
+<p>
+For example, celebrating the new year and celebrating the spring equinox are
+human traditions that have been observed in countless cultures for thousands
+of years. In this calendar, we kill two birds with one stone, as they are the
+same day (unless you are in the Martian southern hemisphere).
+
+<p>
+Earth will be the "real home" of mankind for quite some time in the future.
+Therefore any notable relationship between the planets in their orbits will be
+a cause for observance on Mars. I expect a "Conjunction Day" when the planets
+are across the sun from each other; and an "Opposition Day" when they are 
+aligned on the same side of the sun. These will be a matter of practical 
+concern, as communication with Earth will be easiest as we approach opposition,
+and will be more difficult or impossible at conjunction. They will also add
+variety to life, as they will occur fairly rarely and irregularly &mdash; not
+on a trivially predictable day. Thus they will "float" through the calendar,
+more like Easter than Christmas.
+
+<p>
+I can imagine that "Viking Day" might be celebrated on Mars. The Viking lander
+was the first successful landing by a probe on Mars on July 20, 1976 (which
+itself was the anniversary of the Apollo 11 moon landing). That UTC date is
+still Virgo 12, 1051 MCE, a Martian Tuesday. At Earth midnight UTC the clocks
+read MXT 07:17:53 / MTC 07:06:10; at the landing time 11:53 UTC they read
+MXT 19:10:53 / MTC 18:40:05.
+
+<p>
+It might make sense to celebrate the winter solstice on Mars, as every culture
+on Earth does. Although it won't matter <i>quite</i> as much as on ancient
+Earth, seasonal variations will still affect the Martian settlers.
+
+<p>
+Another big day might be the anniversary of the first human landing on Mars
+&mdash; certainly a day worth celebrating hundreds of years into the future! 
+But of course, we don't know yet when that day will be.
+
+<p>
+But what about Martian versions of Earthly holidays? Will there ever be a 
+Martian Thanksgiving or Halloween or any of those? Only time will tell.
+
+<p>
+So far, we've discussed the calendar, but not timekeeping in terms of hours and
+minutes. There are two or three way of dealing with Martian time.
+
+<p>
+One school of thought says: Forget tradition; let's institute a totally new
+method of timekeeping. And there is some appeal to this. Most of our 
+measurements in modern times are "metric" or base-10 systems. Some have
+suggested dividing the day into 1,000 divisions that are like minutes (or
+100,000 that are like seconds). Such units might be called "ticks" or "beats"
+or other similar names; specifying a time of day might be as simple as 
+naming an integer, such as, "The time is now 592." Overall, however, this
+is too complex for my taste. It introduces totally new terminology and a new
+set of arbitrary units, and it's completely incompatible with what we're
+used to.
+
+<p>
+Another approach was taken by Kim Stanley Robinson in his Mars trilogy. In
+these novels, the Martian settlers measured time as we do from midnight to
+midnight -- but the extra 39 minutes were "off the clock." Time, in effect,
+"stopped" during the timeslip, which they used for relaxation or celebration.
+This is an entertaining solution, and my own concept is very similar.
+
+<p>
+The nearest thing we have to an "official" clock for Mars is the one used
+by many NASA scientists -- what is sometimes called <i>stretched time.</i>
+Because the Martian day is so near our own, it makes sense to divide it
+into 24 hours (each of which is about 4% longer than ours). So each 
+"stretched" hour is made up of 60 "stretched minutes" each of which is
+60 "stretched seconds." So this clock goes from midnight to midnight just
+as we're accustomed to.
+
+<p>
+One small problem with stretched time is that our common units now have
+different values. We use seconds and minutes to measure duration all the
+time; it's undesirable to have two separate standards for those (even
+though the difference is a little less than 3 percent).
+
+<p>
+What I propose is that we keep our usual hours, minutes, and seconds. We
+simply let the clock run longer (similar to Robinson's solution). So the
+day would start at 0:00 and run until roughly 24:39:35 (plus about a
+quarter of a second). If you're staring at the clock at the exact end of
+day, you will see a slight hesitation (or your clock will fall behind by 
+a fraction of a second).
+
+<p>
+This brings us to the issue of time zones. These can certainly be set up
+similarly to what we have on Earth, or we could dispense with them entirely.
+The Martian equivalent of Greenwich, by the way, is a little crater named
+<i>Airy Zero</i>, inside Airy Crater, naturally located on the prime 
+meridian, zero longitude.
+
+<p>
+At this point, someone will wonder about Daylight Saving Time. In my opinion,
+if there is one thing we could leave behind on Earth, this is it. (Even 
+discontinuing the practice on Earth might not be a bad idea.)
+
+<p>
+If we settle Mars, we will have countless challenges and opportunities 
+before us. Worrying about clocks and calendars will be only one of thousands
+of issues we have to face. So we do little projects like this one.  It is 
+wise, or at least fun, to take care of what we can ahead of time.
+
+