Wesnoth calendar

[Simons Mith]

(See also: http://wesnoth.org/forum/viewtopic.php? ... t=calendar)

Consensus on the earlier thread seems to be to give mainstream Wesnoth day and year lengths approximately equal to those of Earth, with the finer details still To Be Agreed.

The Fate of a Princess campaign has a couple of dated journal entries mentioned in it. As I am not aware of any names for Wesnoth months, I sidestepped the problem by using a date format of DD MM YYY with Roman numerals for the month. This gives dates of the form 12th III 967YW, 2nd II 122YW and variations thereon. As long as we avoid day numbers above 30 and month numbers above 12, er, XII, we don't have to decide the finer details of the system. Using Roman numerals is still a fudge, but affects suspension of disbelief less than using the same month names as Earth. Still, it would be nice if the months and possibly the seasons did have names. Naming the weekdays is much less important, I think.

Consensus seems to be emerging elsewhere that the Saurians have a talent for astronomy (and possibly astrology and mathematics too) so might it be fun to have a set of widely-used saurian month-names? But, whatever, could we nail the Wesnoth calendar down now?

Edit: Oh yeah, and perhaps produce some constellation names as well, as the calendar and the night sky tend to be related together, because most civilisations base their calendars around astronomical observations.

[silene]

Still, it would be nice if the months and possibly the seasons did have names.

A bit of History: During the French Revolution, the Gregorian calendar was replaced by the Republican calendar. This brand new calendar was designed by some clever people (e.g. Lagrange), hence the rules for naming of the months are quite interesting. I'm not saying we should use it for Wesnoth, but some of their ideas may be worth considering when designing the Wesnoth calendar. Link to Wikipedia.

[Cloud]

Personally I see different races using different calenders.

Elves - One of the oldest races of Wesnoth and with their longer lifespan I can see them having a more complex calender (i.e. one which repeats every 7-12 years or along those lines). Perhaps something linked to plants or lifecycles, something elaborate, elegant and frustratingly confusing for outsiders.

Dwarves - Another of the oldrer races, but being a predomonantly subteranian race. Lower units of time (like an hour) would probably be measuers by the burning of candles - potentially something humans could 'borrowed' when they met the dwarves. A calender I can see being simple and somewhat inaccurate.

Drakes - Unknown age, possible longer than the above. Probably based around the movements of the sun.

Saurians - Unknown age, middling to old. Lunar based obviously. Possible geometrical based on their maths.

Wose - Old, possible just a knowledge for how old related to the rings of a tree.

Orcs and Goblins - Middling. Very little need for a calender to begin with, but when they begin to work with humans they could possible take a simplified version just so they understand the dates humans talk of.

Undead - Have no need or desire for a calender (who'd want to know how long they had been enslaved after their death for?)

Gryphons - Knowledge of season but nothing written.

Mermen/maids - Something related to tides?

Nagas - Tidal but different to Mermen?

Ogres - Some seasonal knowledge but nothing written.

Trolls - As Gryphons.

Just throwing ideas out, I haven't mentioned humans for the obvious reason.

[thespaceinvader]

Saurians would be more complex than simple lunar, to my mind - they'd probably know of the existence of all the other planets in wesnoth's solar system, any leap days etc, and have a long solar system cycle based calendar, like the mayans.

[melinath]

Humans would use a calendar based on the founding of Wesnoth. Undead would use a much older calendar, probably, dating back to the times when the Wesfolk were still on the Western Continent. (When you think undead, don't think skellie - think really smart lich. They want to know exactly how long they've cheated death, I'd imagine.)

[Simons Mith]

I'm fine with all these, as long as someone somewhere in Wesnoth uses a calendar system approximating to our own...

My first concrete suggestion is to do with the day-measurement:

A day is divided into six watches, which do correspond to the six watches used in the Wesnoth game, each being about four hours long. An individual watch may be subdivided further - commonly into a 'half watch', a 'quarter watch' or quarter (about an hour), an eighth (~half an hour), a twelfth (~20 minutes) or a fifteenth (~15 minutes). Smaller fractions are usually fractions of twelve or fifteen; 1/15, 1/30, 1/60, 1/120, and 1/240 (a minute) being most commonly used. If some fraction of time smaller than 1/240 is needed, we switch to the same fractions of the minute. A 'quarter' is the Wesnoth equivalent of an hour, while the Wesnothians would say 'a fifteenth' where we would say 'a quarter of an hour'. (Their measure is actually 16 minutes long.) 1/240 of a watch is called a minute, and is the nearest Wesnoth equivalent of an Earth minute.

If we set the Wesnoth minute to 60 Earth seconds, then their day length exactly matches ours; I'm proposing we use that as a workable rule of thumb while assuming the exact value is slightly different, but not enough to matter. Maybe somewhere in the range of 58-62 seconds.

How's that for a start?

[Edit: Added 'eighth' as another commonly-used fraction.]

[Simons Mith]

Year measurements: I'm intentionally separating this into its own post.

I set up the time measurements to be different to Earth but with certain overlaps, and I think 1, 2, 4 hours and 1/2 hour, 1/4 hour, and one minute are probably the most useful overlaps to have.

Assuming people like the time measurements for a day, I'm proposing the year is divided up in a similar way to a single watch. For cultural reasons, some races might prefer to use a system of 15ths, while others may use 10ths, 12ths, 20ths, or 60ths, but on the whole most systems should be able to generate a useful yardstick close to an hour and close to a minute without too much difficulty.

So if 1 year = 1 watch, then the same fractional system gives us 4 'seasons' (with 3, 6 or 8 seasons available for use to suit local culture and preference), 12 'months' or 15 'short months' and a 60th gives a 6 day week-equivalent. We can then cherry-pick which cultures prefer which subdivisions while still being able to convert between the different systems without too much difficulty. If we assume that the Wesnoth year is approximately 360 days plus a couple, then we have some spare days that can be used for festivals at solstices or equinoxes, for example. The Earth year length is particularly awkward. Can we make Wesnoth's year 362 11/12 days or something? Rounding up the extra would mean the days in a year are divisible by 3, while the missing 12th of a day would give a missing leap day every 12 years. That's different enough from Earth to be interesting while still being manageable, I think. Once the astronomers determine that the true length of a year is 362+331/360 days, they can arrange an extra-special leap day every 360 years

Start of year may vary from culture to culture; I'm proposing the winter solstice as the most common start-of-year point, followed by the spring equinox as the second most widely-used option.

Perhaps some races collect the stray days together for a several-day festival every few years, rather than having just a day here and there.

Again, is this a usable framework? If it's agreed, we can start slotting in preferred values for the different races while roughly retaining date compatibility between them.

[melinath]

The fractions seem too complicated. Have a look at current calendar systems on Earth.
You're planning on using a human calendar, right? So, it would probably have been made after their arrival on the Eastern Continent. Who would they name the months after? And how many do you think they should have?

[Simons Mith]

The fractions seem too complicated.

The system mainly uses three base fractions and successively halves them to
produce smaller subdivisions;
1, 1/6, 1/12
1, 1/2, 1/4, 1/8
1, 1/15, 1/30, 1/60, 1/120, 1/240

Earth equivalants include British pre-decimal currency, Imperial weights and measures, and the sexagisimal counting system used by a lot of ancients. It's no coincidence that 1/60th appears in the list above as one of the commonly-used fractions. Sexagisimal isn't a bad system, actually.

If that's too complicated, what fractions would you propose? I was aiming for something a bit more complex than just decimals, but I did want it to generate at least an hour-equivalent and a minute-equivalent.

In fact, among early cultures, simple multiples of two and occassionally three were preferred, because you could get successive subdivisions just by halving what you already had. That was easy. Decimal might well have been considered harder.


Edit: Further clarification: This is currently just a proposed calendar framework rather than something intended for a particular race. Once this (or something like it) is agreed, I then envision saying,

'The dwarves use eight 'seasons' called A-H, 15 'short months' in a year, called a-k, and have a 2 or 3 day beer festival centred around the winter solstice.' Their year starts at the winter solstice. They use a six-day week, with days called Ein, Deux, Three, Quattro, Go and Sex.

'The humans use four 'seasons' called Larry Curly, Moe and Shemp, 12 30-day months, and the holy days are spread at 4 month intervals throughout the year. In leap years, the winter holy day, halloween, is omitted.'

Etc.

So, roughly the same framework for all, but different races may empahasise different parts. Does that make my intention at this stage clearer?

[thespaceinvader]

Coming up with a firm calendar before we know exactly what the solar system is like would be premature, i think...

[Simons Mith]

Well, unless the solar system model is so accurate that it lets us derive year and day length from Kepler's laws, and that really would be overkill , I honestly can't see any show-stoppers. I haven't defined a length for the lunar month, which might be useful, but apart from that, what's the problem?

Besides, if we choose a lunar month of say, just over 24 or 30 days, we could use that to reinforce the choice of whatever calender system(s) we decide we prefer. So I'd be quite happy to decide the basics of both simultaneously, TBH.

[Gah. What's with my spelling these days? I know perfectly well how to spell calendar correctly, so how the hell do mistakes like that one keep getting through? I'm far too young to be going senile. I hope.]

[thespaceinvader]

I have hypothesised that the reason for a lack of seasonal variation in day length is due to an elliptical orbit, however, which would affect the calendar a lot

[Simons Mith]

If you want limited seasonal variation, you need a nearly-circular orbit, low axial tilt, and a country close to the equator.

The lower the axial tilt, the further you can go from the equator without getting strong seasons. At low degrees of orbital eccentricity, axial tilt is actually the bigger factor. Earth at perihelion is at 0.983 AU, at aphelion it's at 1.017AU; so the amount of heat we get from the sun varies by about 7% over a year. A highly elliptical orbit would make a far greater difference to the sun's brightness across the course of a year, and we'd be forced to pay more attention to seasonal variations.

[melinath]

Seasons and day length changes come (mostly) from a tilted axis, not an elliptical orbit. All orbiting bodies have elliptical orbits.

It could also just be that the change in day length is under four hours, in which case it wouldn't show up in our approximation.

As to complexity: Other cultures did indeed use fractions, and we still use them today. But notice that we don't say "a day is 1/365th of a year", we say "there are 365 days in a year". We don't say "A knut is 1/29th of a sickle", we say "there are 29 knuts in a sickle".

In other words, forget the poor Wesnothians - it's hard for me personally to understand your calendar system because there are so many fractions. Really, the only hard facts we need as a framework are: how many days in the year? how many days for the moon to go around?
Then, based on that, we can individually develop one calendar at a time.

[thespaceinvader]

A highly elliptical orbit without a tilted axis would also result in seasonal variations, as the whole planet moved different distances from the sun, but would result in seasonal variations without day length variations. I came up with that the explain why there is no apparent day length variation across seasons. I'm not wildly attached to it, though.