Reminds me of one of my favorite posts from back in 2013: You advocate a ____ approach to calendar reform: https://qntm.org/calendar
Specifically (omitting a lot for brevity) :
You advocate a
( ) solar ( ) lunar (x) lunisolar
approach to calendar reform. Your idea will not work. Here is why:
(x) solar years are real and the calendar year needs to sync with them
(x) solar days are real and the calendar day needs to sync with them
(x) the solar year cannot be evenly divided into solar days
(x) having one or two days per year which are part of no month is stupid
(x) your name for the thirteenth month is questionable
(x) the solar year cannot be evenly divided into seven-day weeks
Specifically, your plan fails to account for:
(x) rational hatred for arbitrary change
(x) unpopularity of weird new month and day names
and the following philosophical objections may also apply:
(x) good luck trying to move the Fourth of July
(x) the history of calendar reform is insanely complicated and no amount of further calendar reform can make it simpler
Furthermore, this is what I think about you:
(x) sorry, but I don't think it would work
> This is perhaps the weakest of the objections, it has been done and it could be doable, if we wanted it.
It's not at all a weak objection. I mean, sure, months are arbitrary in the first place and it's not physically impossible to set up a system where there are days outside of months. A system set up that way might not be any more complicated than our current system since it would simplify other aspects of dates. But making that change would fundamentally break a lot of our assumptions about dates in ways that would require a ton of effort to fix (good luck getting any existing software dealing with dates to work properly after that) for very little benefit, so it's an incredibly bad idea.
The article also says that the extra days wouldn't even be assigned days of the week, which is even worse.
Getting existing software to change is equivalent to the general switching cost problem which is the real argument.
Software modeling would be easy. For software it is an additional month with only 1 or 2 days. However because it would be a holiday most humans wouldn't need to talk about it as a month.
> Software modeling would be easy. For software it is an additional month with only 1 or 2 days.
You will also have to consider actions that are performed on the first of the month and possibly special case these months. This also means that effectively you have just made the length of months even more variable and confusing.
It's really only "easy" in the sense that it is possible to fit it into existing date data structures, but it's not necessarily that "easy" in terms of updating business logic.
> However because it would be a holiday most humans wouldn't need to talk about it as a month.
It can't be a holiday for everyone and it will still need to be something that people think about. Hospitals, fire departments, police, and websites that are expected to operate 24/7 will still need to operate and handle the special day. It will probably also need to be a new special case in literally every contract (leases, employment contracts, billing terms, etc.).
The human aspects are perhaps even harder than the computer aspects because it won't be possible to simply think of it as a 1 day month, it will HAVE to be special cased.
The original article also said that the extra day would not be assigned a day of the week. This is even more confusing. Currently, every 7 days is one week regardless of years, but if this change is made, for any action that needs to be performed once every seven days, either a decision will have to be made on whether 1) to perform/not perform the action on the special day depending on what day it WOULD have been if it wasn't the special day and then shift the day of the week for the next year, or 2) allow an interval of 8-9 days once a year.
Even though there is currently a slight inconvenience from 1) months having different numbers of days and 2) days of the week not being the same every month, these reflect the realities that months are arbitrary and don't match up to solar years/days.
The proposed change is actually arguably worse because it has to add even more complexity in the form of edge cases to try to pretend these problems don't exist.
a week is just a quarter month... or that is, it should be a quarter month. months as a time unit are several sorts of screwed up. my advise, if you can get away with it, is to never use months in your accounting.
Two small things I would love to see changed with the current calendar, but will never happen:
* February deserves to have more days, so that every month has either 30 or 31 days. Taking away 1 day from August and January, and giving them to Feb would do it. In leap years, Feb can have 31 days.
* The year ought to start on March 1st. So that September regains its place as the 7th month, October the 8th month, November the 9th month, and December the 10th month. Because that's what "Sept", "Oct", "Nov", and "Dec" mean.
In addition to making the number months make sense again, it would also make leap day at the end of the year (again), which probably also makes more sense.
I think your point about February is probably good, but I don't know that pushing two changes is possible.
> Because that's what "Sept", "Oct", "Nov", and "Dec" mean.
Please keep that idea to whatever languages that assign those names to months. East Asian languages have been calling them literally "month 1/2/3/etc." without problems, just like how most languages don't have 31 different words for each day in a month.
Alternatively, if we have to give months aliases, why not fix those languages by e.g. reassigning "September" to the 7th month?
TEC has many unique properties. It breaks down into many mathematical models. One week can be divided into whole days by 2, 3, and 6. They can be divided by 4 as well, with half days. One month can similarly be divided into weeks. One month can be divided into whole days by 2, 3, 4, 6, 9, 12, and 18. The 60 week year, not counting the last week, can be divided evenly by 2, 3, 4, 6, 12, 15, and 30.
Triangular Earth Calendar is named so because of its triangular properties. Unlike most calendars which are viewed only on a square grid, TEC is viewed in its most natural form, the triangle. The following are representations of TEC in triangular form:
Here is the view of a single week.
1
2 3
4 5 6
That is the same for weeks in a month. Here is the view of days in a month (using single numbers only).
I encounter this in UI/UX all the time. You can think of ways to make things easier for people who haven't learned it yet, but in the process it makes things difficult for the people who already learned it, because they need to learn something over again and also unlearn what they already learned. And in many situations, most people who need to learn it already have.
In short, no one will approve something like this calendar change unless only 5 year olds get to vote.
It seems like several companies have adopted something similar without trying to change the months. Take for example Intel's calendar where the manufacturing process follows a weekly calendar - their dates are represented by a year, a work-week number, and a day number[0]. They even make their suppliers use their calendar.
The calendars on MacOS and iOS also have a toggle (under advanced settings) to display week numbers.
The low-key success of week-number calendars points to what I think was the big reason the Comté calendar never took off: Months are, unless you're working in a maritime environment, an unnecessary intermediary unit of measure, and for most are only valuable for their legacy cultural references. Redefining them should have been a non-starter.
A week-number calendar is great because it's simpler than trying to retrofit a year-month-day structure, yet still allows legacy Gregorian months to be seamlessly overlaid on it. The Intel PDF is a great example of this.
I don't fully understand the day number part of the linked document. If I want to say July 12th, 2021 in Intels weekly calendar, do I write:
2021-29-12 (29th week of the year, 12th day of the month)
Or do I write:
2021-29-02 (29th week of the year, second day of the week)
The linked PDF makes it seem like the former. But the latter makes more sense in my head because it doesn't rely on months (which is seemingly the whole point)
> all the Pope Gregory the 13th did was account for leap years by adding in an extra day every few years to keep the year aligned with the astronomical year.
The Julian calendar has leap days every four years. What Pope Gregory the 13th did was remove it from every 100th year (except those divided by 400, hence why 2000 was a leap year). Otherwise the difference between the Gregorian and Julian calendars would be far more than the 12-13 days they are now.
I worked at Kodak in Financial Information Systems, and 1989 was spent getting ready to convert all of the batch jobs from 13 periods with 28 cycles to monthly.
Lots of figuring out when to run things… I had that exact poster up on my wall in my cube that’s shown in the article.
I would imagine that it was very sifficult to use a non-standard calendar with the pc-based commodity software that was becoming ubiquitous in the mid/pate eighties. Was that part of the reason to migrate back to the Gregorian calendar?
“A hotel that did a business of $10,000 per week in room sales found that its receipts from room sales were less in May than those in April. It looked as if the business was dropping off. May was one day longer than April and yet its room sales were less. The figures, however, proved to be very misleading. As a matter of fact business was actually better in May than in April – ten dollars a day better – but the monthly comparison seemed to show that it was worse.”
So revenue/day was up in May, and there are more days in May, but total revenue was down in May? What's the explanation?
I would assume the alignment of the months to the days of the week. Assuming they make more at the weekend, if May had a one more weekend then April (I think that’s possible) then it would have greater revenue.
The Kodak calendar fixes that as all months have four weekends.
Having the same weekday a specific day of each month sounds so boring. I like the variation with our current calendar. Sometimes your birthday is on a weekend, sometimes not, sometimes Christmas is on better or worse days than other years. Easter may be early or may be later. It makes the year more interesting.
If you're making a 13 month calendar, you have to rename all the months so that there is no ambiguity. If you plan an event for September 15th, it would be insane for people to have to ask the question "which September 15th"?
Better yet, let's just get rid of months. They make no sense as a unit of time. There is no other measurement that can have multiple different values depending on the time of year. Imagine if the length of a meter varied, life would be madness.
I have worked on a widely deployed product that _internally_ uses something similar: 28 month days, 13 months to a year. Data is aggregated in the product into daily, weekly, monthly, and yearly aggregate files. It does two things differently from the article: 1) it doesn't add a day at the end of the year to get to 365 (28 * 13 == 364, 1.25 days short of a normal year), and 2) it considers the beginning of time to be Jan _4_, 1970, not the more typical Jan 1. This was chosen so that each of the weekly data files started on a Sunday, as opposed to a Thursday which is what Jan 1, 1970 was. The most fun consequence of this odd scheme came up when I was trying to figure out what date a yearly data file should start on. Eventually I figured out the pattern, and realized that the date is slowly drifting backwards, starting from Jan 4, 1970, at a rate of 1.25 days per year from 1970 due to only having 364 days per year. It's been 52 years, and so the current year file starts in late October this year.
There are industries that employ 13 period accounting. I encountered it in the bar and restaurant sector. Weeks are accurate, repetitive measures for them. Months and quarters are too variable to be of much use. However the practice does make for a lot of adjusting entries, prepaids and accruals to align with the actual calendar and billing cycles. Somewhat of a pain...
Interesting how the 1920s seem to be an era of social management experiments everywhere, with basic things like calendar being questioned all over the world. Just a few years later after Eastman's push for a reform in the United States, the Soviet Union actually changed its calendar. The reform affected the definition of a week rather than the month. The reform was not successful, and the changes were only in place in 1929–1931.
My grandfather, who turned 10 in 1929, vaguely recalled how the country briefly abandoned universal weekends, which meant a mess scheduling anything family-related. Everyone had different work schedules.
https://en.wikipedia.org/wiki/Soviet_calendar
Can someone make a solid case as to why the calendar has to follow astronomical events in a post-literate, largely non-agrarian, society?
When literacy rates were low, and many people were farmers, it was important to be able to easily anchor schedules to the seasons. Having May 30 always be the right time to start planting your crunchberry crop was a viable oral-tradition thing.
But in an industrial society, we can fairly easily publish "peak season for crunchberries this year starts June 6" even if that's five days earlier than last year.
Then we can pick, say, a 360-day year with no leap-nothing for ease of record keeping.
There are some activities that might legitimately be seasonal-- you don't want the kids playing $active_sport when it's 45C outside, so we might need to realign their calendars each year, but not a huge deal breaker. Reggie Jackson could be "Mr. When The Afternoon High is 28C" instead of "Mr. October."
I have long thought that the uninterrupted series of 7-day weeks will be one of the hardest things to change that humanity has. I don't see how it can be done. It is so embedded in our culture. I think it will never happen as long as civilisations exist.
I think it can change and in fact has already started to change. The widespread adoption of remote work is eroding the concept of a "workday" and I expect this to continue.
The treasurer of my old volunteer fire department was a manager at Kodak and he always brought the Kodak calendar to put on the wall of the board of directors room. Kodak had moved back to a 12-month year, but as shown in the image in the article, the week started on Monday instead of Sunday. If you glanced up to look at what day of the week a particular date would be, you'd be off by a day. We hated it.
Messing with the definition of the week is always an issue I have with these calendars. An unending cycle of days is more useful than fixing the months to the week. The week-based calendars like https://en.m.wikipedia.org/wiki/ISO_week_date are a better fit, though they clearly have more of an issue with seasonal drift.
I had an hour-long debate with coworkers about the International Fixed Calendar (Sol Calendar). There are so many benefits! I particularly like that every 1st of every month is a Sunday, every 2nd of every month is a Tuesday, etc. No more "What day or date does _____ holiday fall on _this_ year?"
What you're saying is apparently true for the International Fixed Calendar, but not the Sol Calendar. According to [1], the first (and second and third...) of every month in any given year is the same day of the week, but the mapping changes every year.
The International Fixed Calendar may start every 1st on a Sunday but the European Fixed Calendar will probably make each 1st a Monday. Agreeing on things is hard.
Something I notice in the Kodak calendars image: rather than having extra days, the 1928 calendar simply has 364 days, but the 1989 calendar has an extra week in the last period. Did they ever explain the rule for adding an extra week?
But lunations are, as far as I can see, irrelevant to 99.99% of human beings. There is no reason a modern calendar has to align with what the Moon is doing.
Why do you think a calendar should be aligned (sort of) to them?
Because that is the way nature divides time. Anything else is arbitrary. The first thing our ancestors would have noticed, before the invention of time-keeping devices, would have been the phases of the Moon and how they relate to Sun/Earth cycle, ie. the seasons. Lunations so soli-lunar events, not merely lunar.
Not so arbitrary really as there are 7 planets visible to the naked eye. Well, 5 planets and 2 "lights" to be precise but each day is named after one of these. The reason we don't get all this is that we're so out of touch with planetary cycles. We obsess over micro-measurements on terminals with ever more powerful technology yet can't see what is right under our noses.
What is more fundamental to earthly time than the rotation of the earth and its orbit around the Sun. From there it follows that the cycles of the other planets may also be important. I don't see how any of this is arbitrary. Isn't it amazing, for example, that Jupiter's cycle is 12 years and there are 12 lunations in a year? The cycle of Uranus is 84 which is 12 x 7. There's also a subtle mathematical relationship between 12 and 7. 12 = 3 x 4, 7 = 3 + 4. You have to have eyes to see these patterns.
> What is more fundamental to earthly time than the rotation of the earth and its orbit around the Sun.
True.
> From there it follows that the cycles of the other planets may also be important.
Not particularly, no.
> Isn't it amazing, for example, that Jupiter's cycle is 12 years and there are 12 lunations in a year? The cycle of Uranus is 84 which is 12 x 7. There's also a subtle mathematical relationship between 12 and 7. 12 = 3 x 4, 7 = 3 + 4. You have to have eyes to see these patterns.
None of that is amazing in the slightest, and your base numerology won't make it amazing.
You can find coincidences everywhere if your definitions are loose enough.
Specifically (omitting a lot for brevity) :