Talk:Date of Easter/Archive 1

This page is an archive of past discussions. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page.

Name of the Article

I do not like the title of this page. It is too long and I never get the small words and capitalization quite right; the capital 'E' is un-Wikipedia anyway. I propose to re-name this "Computus", for the medieval term of this craft. -- Tom Peters 25-Jul-2003 21:52 UTC

What was the original title of this page? The term Computus is hardly known, even by those who take an interest in these things. The case-sensitivity of Wikipedia's search function (apart from the first word) is one of the major bugbears in its use.—Copey 2 13:42, 3 May 2006 (UTC)

"Calculating the date of Easter" still links here, but I believe other re-directs existed with a different capitalization of Date or easter; anyway, I never got it right. "Computus" is the technical term and has been used since the early middle ages. "Calculating ..." is a descriptive phrase. IMNSHO an encyclopedia should have lemmata of words, not awkward phrases. Tom Peters 18:01, 3 May 2006 (UTC)

Gauss's Algorithm

I found a source for Gauss's Algorithm, Blackburn & Holford-Strevens pp. 864–866. However, the Gregorian exceptions are described differently:

if M = 0, 3, 6, 8, 11, 14, 17, 22, 25, 27 and 19a + M (mod 30) = 29, d = 28.

if M = 2, 5, 10, 13, 16, 21, 24, 29 and 19a + M (mod 30) = 29, d = 27

Does anyone know if the exceptions are actually equivalent? They don't seem equivalent to me. Alternatively, which is right? --Gerry Ashton 17:14, 9 October 2006 (UTC)

I'm not sure what you mean: the algorithm provided on this page does not give a list of exceptions like H-S&B do, but provides a pre-computed table of (M,N) values that is identical with what H-S&B compute; and gives two heuristic rules. Do you have an example where the two descriptions give a different date? Tom Peters 21:00, 9 October 2006 (UTC)

Gauss spent several years on his definitive calculations for Easter and published three (similar) algorithms. I cant give a definitive source for this but Algorithm 1 presented here works for the 1980-2024 dates listed elsewhere and I had the impression always worked without exceptions. It looks quite similar to the algorithm posted on the page but precedes it by 150 years !

Numerator	Divisor	Quotient	Remainder
year	19		A
year	100	B	C
B	4	D	E
8B+13	25	G
19A+B-D-G+15	30		H
A+11H	319	M
C	4	I	K
2E+2I-K-H+M+32	7		L
H-M+L+90	25	month
H-M+L+month+19	32		day

A1jrj (talk) 13:17, 1 May 2008 (UTC)

I may have been misled here - a little research suggests that this algorithm was popularised in O’Beirne, T.H. “Chapter 10 Ten Divisions Lead to Easter” in Puzzles and Paradoxes. London: Oxford University Press, 1965. 158.234.250.71 (talk) 17:57, 1 May 2008 (UTC)

That happens to be correct for all years starting with Year 0. But it is not particularly quick.

Definitive algorithms should be directly traceable to the prime authority - that means Clavius or the Calendar Act, since those are what a Pope and a King authorised - or to the person whose algorithm it is said to be. To do otherwise introduces doubt. Faster algorithms may be preferred for actual use, but should be compared against something known good. For Easter, that means comparing for at least 5700000 years.

For use by people, Easter algorithms need to give Month and Day. But for computer purposes it is often better to give Day-of-March instead - h-m+l + 22 in this case - since that can be directly converted into such as a JavaScript Date Object or a VBScript CDate. And, correspondingly, it can be useful to have the Day-of-March indicated in comment if Month and Day are returned. Indeed, if the language provides an easy way of moving to the Sunday next following, it can be useful to have the Day-of-March of the Paschal Full Moon returned or indicated.

82.163.24.100 (talk) 17:52, 1 May 2008 (UTC)

References and Table

This article is weak in the way sources are cited. There are long passages of complex text that do not indicate which source they came from.

In particular, two edits have just occured concerning the epact table, but there is no indication which source the table was taken from, so one is forced to do some laborious calculations to see which version is right (unless you happen to be a computer programmer). --Gerry Ashton 22:14, 4 October 2006 (UTC)

I have found sources for the table in question and added appropriate footnotes --Gerry Ashton 23:13, 4 October 2006 (UTC)

The tables follow from the procedures as explained. The readers should be able to construct them themselves. The procedures come from the documents involving the Gregorian reform, which are not easily accessible: the site of Rodolphe Audette [1] appears to be the only reliable source on the Internet. I believe all computations are explained in sufficient detail that they can be checked, so no further reference needed; indeed, I've found nowhere in writing how the lunar calendar REALLY works, why things were done this way, and what all the consequences are. Much of the literature have not thought things through (including apparently Clavius!). The draft by Denis Roegel [2] comes closests but is very technical. Dr. H. Lichtenberg is misguided. Tom Peters 10:59, 5 October 2006 (UTC)

The Blackburn & Holford-Strevens source that I added explains the mechanics, but does not go into the accuracy of the result. The fact that the text is self-consistent is better than nothing, but the information in the article really should be verifiable. --Gerry Ashton 16:15, 5 October 2006 (UTC)

There is a Reference to Weisstein, for PFM. That page is unexceptionable but trivial; it links to an Easter page, which is in error. It seems to me that the Weisstein material is no longer maintained, and is better not cited. 82.163.24.100 (talk) 12:29, 18 April 2008 (UTC)

In References + Links, I see nothing for the original authoritative materials. The Papal Bull and the Calendar Act give details of the new secular calendar (the Leap Year Rules), and give authority for, but no description of, the revised Easter rules : those were in Annexes. References and links are much needed. Clavius is imaged at http://mathematics.library.nd.edu/clavius/. For the Anglican version, I suggest the Preface to the printed Book of Common Prayer of the Church of England. 82.163.24.100 (talk) 12:29, 18 April 2008 (UTC)

I've Googled for "intermediate Years betwixt" which should appear in any true copy of the BCP, and found nothing authoritative ; "eskimo" has it, but has Gregorian material in a URL containing 1662. 82.163.24.100 (talk) 12:29, 18 April 2008 (UTC)

Google Books currently has, via http://books.google.com/books?id=zr8PAAAAIAAJ&printsec=frontcover&lr=#PPA24,M1, images of the relevant pages of the 1815 Prayer Book; see pp. 24 ff. It largely matches a c.20 copy of the Book (limited-period matter differs but corresponds). It can be considered as a including a faithful representation of the Annexe to the 1751 Act. Page http://justus.anglican.org/resources/bcp/ The Book of Common Prayer at the http://justus.anglican.org/soaj.html Society of Archbishop Justus has http://justus.anglican.org/resources/bcp/1789/Tables&Rules_1789.htm TABLES and RULES from the 1789 U. S. Book of Common Prayer, in HTML, matching the 1815 Book. I suggest that at least one of those should be a reference or link in the article, at least until/unless a good direct copy of the original Annexe itself is found on the Web. 82.163.24.100 (talk) 10:45, 22 April 2008 (UTC)

Whose mistake?

I wrote an Excel spreadsheet using Jean Meeus' formulae; it put this year's Gregorian Easter on 2 April, two weeks early, and the Julian Easter on 10 April, a Monday. I haven't found a mistake in my own copying. Could someone else please verify the Meeus algorithms given here? NakedCelt 16:45, 10 March 2006 (UTC)

Found the mistake in my Gregorian calculation, but the Julian Easter still seems to be on Monday.

There is currently a 13-day difference between the Julian and Gregorian calendars. April 10 Julian is April 23 Gregorian, a Sunday. Checking another source, I find that this is the correct Julian Easter for 2006. Indefatigable 20:57, 10 March 2006 (UTC)

Western Easter only?

Do I understand correctly that this page pertains only to the Western calculation of Easter? If not, the two methods need to be explained distinctly. If so, then this page should include this specificity very early, and a corresponding article be created explaining Eastern Easter computation. — Xiong 熊 talk * 05:57, 2005 August 18 (UTC)

The article does discuss both Easters, but not under the terms "Western" and "Eastern". Eastern corresponds to Julian, and Western to Gregorian. Perhaps an explanation of the different nomenclatures should be added. Indefatigable 16:46, 20 August 2005 (UTC)

There are in fact three cases of significance, at least - Gregorian Easter on the Gregorian Calendar (in general current use), Julian Easter on the Julian Calendar (used in Europe in the first half of the second millennium) and Julian Easter on the Gregorian Calendar (currently, I think, what Russians need). The calendar change is a simple function of the centade. 82.163.24.100 (talk) 21:47, 9 April 2008 (UTC)

Easter Sunday (Western)

If I am reading the Computus page correctly, it is stating that according to the Meeus/Jones/Butcher Gregorian algorithm, Easter Sunday will be April, 11, 2007. My 2007 calender has April 8, 2007 as Easter Sunday. On Jan. 1, I checked the page again and the correct date of April 8, 2007 was in the table named Worked Example Year(Y) = 2007

Did you look at the table with worked examples? The last three lines of the table give the results, and show April 8, 2007 as Easter, just like your calendar (and mine too). I entered the formulas in the table into Excel, and my results agree with the table. Notice that the formulas that appear in the section after the table are for Julian Easter, not Gregorian Easter. --Gerry Ashton 14:54, 30 December 2006 (UTC)

Chronology

This article has the potential to become a really great article for a complicated subject as the Easter calculation. First, I would discuss the Julian calculation BEFORE the Gregorian calculation.

The reason why I put it after the Gregorian one is that it is now obsolete and I thought people should not read through the old stuff before finding out how it is actually done now. Tom Peters 14:42, 29 March 2007 (UTC)

Julian Easter is a straightforward implementation of the basic principles. It uses a Spring Full Moon date of maximal regularity easily calculated from the Golden Number, then a move to the following Sunday. That, described in modern arithmetical terms, would be an easy summary introduction to the tabular methods (in either order) that were actually decreed. http://www.merlyn.demon.co.uk/estr-bcp.htm#JES. 82.163.24.100 10:36, 18 May 2007 (UTC)

Second, I would expand the article to include the epacts, new moon dates, and full moon dates in the same table.

I disagree. The article, which is already long in its current state, shows the general principle. The current page has the bare workable minimum of needed information, and all the rest can be constructed from that. No need to clutter the page with more and expanded tables; those can be found among the numerous references. Tom Peters 14:42, 29 March 2007 (UTC)

The Easter calculation is based on the Metonic cycle. That is, we assume there are 235 lunar months in 19 solar years. Because the Julian year has an average of 365.25 days, the lunar cycle drifts backwards with the Julian calendar by an average of one day every 312.5 years. Thus by 1582, the lunar cycle fell four days behind the Julian calendar over a 1250 year period of this calculation. Morever, the Julian calendar was too long by 10 days over those 1250 years. Thus, in 1582, 10 days were removed from the calendar to correct the Easter calculation, a new Gregorian calendar had an average of 365.2425 days, and new full moon dates were established for Easter. Thus, with the Gregorian calendar, the lunar cycle drifted later with the calendar by one day every 233 years, or 43 days every 10000 years. To compensate for this drift, the epacts were decreased by an average of 43 times in 10000 years using a specific pattern. The decrease in the epacts caused the lunar dates to appear later by one day except where full moon date April 18th was reset back one month to March 21st.

The Julian calculation really used epacts of 0, 11, 22, 3, 14, etc. but with a specific solar date of March 22nd and a full moon date 14 days later. I would show these epacts and the new moon dates along with the full moon dates on the same table.

When the Gregorian calculation was established, the specific solar date to use for the epact calculation became January 1st (and March 31st and April 29th). Thus the epacts for prior to 1583 were restated to 8, 19, 0, 11, 22, 3, etc. The full moon now came 13 days after the new moon. Thus the restated epacts still produced the same full moon dates. I would show this table for the Gregorian section. I would also show the epacts, new moon dates, and full moon dates from 1583-2599. Thus in 1583-1699, the epacts were 1, 12, 23, 4, 15, etc. In 1700-1899, the epacts were 0, 11, 22, 3, 14, etc. In 1900-2199, the epacts are 29, 10, 21, 2, 13, etc. In 2200-2299, 2400-2499, the epacts will be 28, 9, 20, 1, 12, etc. In 2300-2399, 2500-2599, the epacts will be 27, 8, 19, 0, 11, etc. I would show all these tables along with the new moon and full moon dates.

No please don't: too much information can confuse too. Tom Peters 14:42, 29 March 2007 (UTC)

--Trust101 04:38, 29 March 2007 (UTC)

Traceable Easter Sunday Date Calculation

The true authorities for the Date of Easter Sunday are :-

In the British/Anglican tradition - the annexe to the Calendar Act (now on the Web), and the copy in the Prayer Book;
In the Roman Catholic tradition - as far as I know, just the Explicatio of Clavius, expanding the Papal Bull.

The Explicatio is some 500 pages of 16/17th century Church Latin, and the Web copy at ND.EDU seems in poor focus. It is likely that the Parliamentary advisers of the mid-18th Century were good at Latin, and had a better copy. And it appears certain that the method decreed by the resulting Act of Parliament gives results nowhere disagreeing with the Roman Catholic practice.

Granted that we have the algorithms presented by Zeller, Gauss, Richards, Oudin, Butcher, Meeus, Feather, USNO, Mallen, and others; and that bug-free implementations of those all agree within the range of the internal arithmetic used. Nevertheless, it seems desirable that there should be an algorithm directly traceable to the ultimate authorities; and sufficient that it be done from the Prayer Book or the Act annexe.

Such an algorithm is derived and presented, in Javascript, in Web page http://www.merlyn.demon.co.uk/estr-bcp.htm, and tested in that page and in http://www.merlyn.demon.co.uk/estrdate.htm - it can be shown by test forms on those pages to agree with several other algorithms for the whole of a 5,700,000 year cycle. Or more.

Tables I & II in the Prayer Book are for the years 1600 to 8500 etc. The pattern of Table I, for the Sunday Letter, is obvious. The pattern of Table II is not obvious; guidance from other sources implies a simple form fitting the data given. A corresponding expression, valid in perpetuity, can thus be derived from each of those Tables; and Table III can also be reduced to an expression.

One of the implementations on the cited site should cover all years, positive and negative, for which the integer range suffices - ±9E15 in Javascript; another covers only the 32-bit unsigned range, and is rather fast. They are shorter than ones written for simple calculators.

With a copy of the pre-1753 Prayer Book, it should be easy to generate corresponding algorithms for Julian Easter.

82.163.24.100 21:43, 5 May 2007 (UTC)

That Julian Easter algorithm has now been included. It is deduced, rather than traceably derived from proper authority; but tracing should be easy once authority is located.

The work of Zeller deserves a mention in the main page; a link would suffice.

82.163.24.100 11:18, 29 August 2007 (UTC)

Logical Basis

May an ignoramus say, I came to the page wanting to understand why the date of Easter is so variable, and so complicated to compute. I think it would help other ignorami if this could be set out here. And why it is so important for the title of this page to be such an obscure word.

Afterbrunel (talk) 22:04, 6 January 2008 (UTC)

Reversion of Excel date for Easter

I have removed the Excel formula for the date of Easter that was added on January 15. I compared it to values given on two pages of the United States Naval Observatory. It matched the dates given from 1980 through 2024 at http://aa.usno.navy.mil/faq/docs/easter.php but when I tried the year 3055 in the form at http://aa.usno.navy.mil/data/docs/easter.php the USNO gave the date of Easter as April 8, while the Excel formula gave April 1.

This formula has been added once before, and somone reverted it without a detailed reason. Now that a detailed reason has been given, I believe I have established that the site http://www.cpearson.com is not a reliable source and algorithms from that site should not be added to this article. --Gerry Ashton (talk) 15:40, 15 January 2008 (UTC)

It is valid for 1900 to 2099 only; see http://www.merlyn.demon.co.uk/vb-date2.htm#ESu. The source of that page contains a valid function for 1900-2199, and another for all years, which could presumably be used in Excel, and indicates where their derivation from an authoritative source may be found. 82.163.24.100 (talk) 12:20, 20 March 2008 (UTC)

Note that Anglican and British Easter are technically only defined up to 8599, the limit of Table I in the Church of England Book of Commpn Prayer, authorised by the Calendar Act. Continuation of the Table can however be deduced using other sources. http://www.merlyn.demon.co.uk/estr-bcp.htm refers. 82.163.24.100 (talk) 12:20, 20 March 2008 (UTC)

The "on or after" saga

The web is littered with mistakes on the definition of Easter dating. This article, however, is of a very high standard in my view.

One of the most frequent mistakes is to abbreviate "on or after March 21" to "after March 21", which of course leads to incorrect results.

In fact, I've just corrected one of these mistakes in this very article!

I advocate that we change all "on or after March 21" phrases to "after March 20", to avoid perpetuating this mistake, and seek your views on this.

NO. We do not have the authority to rewrite the Calendar Act or Papal Bull Annexes, even without change to meaning. Your suggestion would have been acceptable in 1582 or 1751. 82.163.24.100 (talk) 16:33, 17 April 2008 (UTC)

I think this is consistent with the change in meaning of the word "after" from use up to about 300 AD where it meant "including and afterwards", then through an ambiguous phase up to aroung 1750 when it needed clarification, to modern use where it is well understood to mean the same as "following".

All input welcome within Wiki guidelines ;-) Ron Mallen (talk) 13:12, 15 April 2008 (UTC)

Good catch. I assume your AD 300 limit refers to Roman inclusive counting. But that phrase did not exist before AD 300, nor did it exist until much later. I suspect it is a modern phrase meant to succinctly state the original rule given either in tabular form or in more verbose terms. We do not know what the equivalent Greek term was because no Alexandrian texts survive from that period, only tables. The next earliest description is in Latin by Dionysius Exiguus in 525 (I discount Victorius of Aquitaine's confused description in 457). But he did not describe lunae XIV at all—he only gave a few argumenta on how to calculate each parameter, not what they meant. The first thorough analysis is by Bede in 725. Moreover, your suggested change from "on or after March 21" to "after March 20" is not appropriate for Wikipedia because we must use the terminology found in the literature. We are not allowed to invent new terms because this is an encyclopedia, a digest of known knowledge, not a research paper. Any such change is original research forbidden on Wikipedia. — Joe Kress (talk) 22:13, 15 April 2008 (UTC)

I don't agree with Joe Kress that there is a difference between "on or after March 21" and "after March 20". We are writing in modern English and may use any ordinary English phrase with the appropriate meaning. Now, if we were to make up a neologism for the concept of "on or after" as it applies to dates that are not tied to any particular time zone, that would be inappropriate. --Gerry Ashton (talk) 22:24, 15 April 2008 (UTC)

Correct Joe, on the reference to 300 AD Roman inclusive counting; also I agree that historic text uses the "on or after" phrase, I believe due to the need to clarify the ambiguous and changing meaning of the word "after".

However, I agree with your view Gerry, that we need to write this Wiki reference in modern and clear text. We have abundant* evidence that the phrase "on or after March 21" is misused, and I'm strongly in favour of replacing it with "after March 20" to prevent perpetuation of this simple but serious mistake.

A check now on Google returned 10,400 references to "Easter" and "after March 21", which includes only 717 references to "Easter" and "on or after March 21" (I hope it's not really that bad!)

I propose that we, as a group, let this topic run for a few weeks, then review collective input. I'll refrain from further comment for now to promote that input. Ron Mallen (talk) 01:13, 16 April 2008 (UTC)

Your Google count is flawed because Easter is indeed "after March 21"—only its Paschal full moons or lunae XIV are "on or after March 21". — Joe Kress (talk) 02:55, 16 April 2008 (UTC)

Searching Google for “Easter Dating”, I have taken the first 10 unique samples looking for misunderstanding of “Ecclesiastic Full Moon date on or after March 21”. Excluding my own site and Wikipedia, 4 have incorrectly abbreviated “on or after March 21” to “after March 21”, while a further 4 have used the misleading phrase “full moon after the Vernal Equinox” without giving a nominal date for this Equinox. Only 2 are correct in regard to dating. For example, Encyclopaedia Britannica states incorrectly: “Easter should be observed on the first Sunday following the first full moon after the spring equinox (March 21)”. Using this false definition, the earliest Easter Sunday date would be March 23! This mirrors my research on published references printed over the last 150 years which overwhelmingly have made this same mistake. Given this obvious confusion, I think an authoritative reference such as Wikipedia should use the clearer phrase “after March 20”. I propose to make these changes in the main article, while maintaining context, after waiting a few weeks for feedback from anyone with good reason to oppose this change. I have documentation of the 10 Easter definitions found above. Please email me at rwmallen@chariot.net.au if you would like a copy. (Ron Mallen (talk) 10:45, 14 June 2008 (UTC))

In the previous paragraph, the sentence ending "March 23!" is mistaken. The Spring equinox is an instant. If that occurs on March 21, the full moon can be later on March 21; and if that is a Saturday, Easter is March 22. Britannica did not have "date of the spring equinox", without which your deduction is unsound. 82.163.24.100 (talk) 14:44, 3 August 2008 (UTC)

One must distinguish very clearly between a "definition" and any corresponding "description", however careful the latter may be.

There are only two proper authoritative definitions of Gregorian Easter. One, for the Catholic Church and adherents, is in Clavius' Explicatio - that's hard to read, as Clavius wrote in a long-winded fashion in Latin (and the only Web-accessible copy I know of is in poor focus). The other, for Britain and the then & subsequent colonies, is the Annexe to the Calendar Act of 1751, as passed, visible in Statutes at Large 1765 and Prayer Book 1815 at Google Books, but not as in the Statute Law Database.

Only those should be cited as definitions; as authoritative definitions, they must not be re-worded.

For a description, give the exact Act/Book wording : "EASTER-DAY ... is always the first Sunday after the Full Moon which happens upon, or next after the Twenty-first Day of March ; and if the Full Moon happens upon a Sunday, Easter-day is the Sunday after." with in addition a statement that said Full Moon is a calculated approximation.

I consider "upon, or next after the Twenty-first Day of March" to be perfectly clear.

But, to allow for the semi-literate, one might add, outside the prime description, that since the Paschal Full Moon is after March 20th and Easter is after that, the earliest Easter date is the 22nd. But it would be better to put that since the Paschal Full Moon is on or after March 21st and Easter is after that, the earliest Easter date is the 22nd.

Or one might explicitly state that some have incorrecly omitted "on or".

Be aware, though, that British Gregorian Easter is truly defined only for years 1600 to 8599.

See http://www.merlyn.demon.co.uk/estrdate.htm and http://www.merlyn.demon.co.uk/estr-bcp.htm.

82.163.24.100 (talk) 13:55, 15 June 2008 (UTC)

Is the correct definition: "Easter is the first Sunday following the first full moon on, or after the Ecclesiastical Vernal Equinox. The Ecclesiastical Vernal Equinox occurs on 21 March"? jonathon (talk) 18:38, 14 June 2008 (UTC)

NO. The definition, in British Law inherited by the then and subsequent Colonies, makes no mention of the Equinox. It uses March 21st, which is the only fixed date which should be mentioned. Also, while the Law at first refers to the Full Moon, it later uses Paschal Full Moon. Any short description should include "Paschal". The Law does not use "Ecclesiastical", as far as I recall.

The definition requires and includes Tables I-III in the original Act, from which the full algorithm can be deduced : anything less can only be a description.

82.163.24.100 (talk) 13:55, 15 June 2008 (UTC)

Every Lunar Month, there is an Actual Full Moon, at a particular instant worldwide. There is an Ecclesiastical Full Moon Date within about a couple of days of every Actual Full Moon; it is the same local date world-wide. The Ecclesiastical Full Moon on or after March 21st is the Paschal Full Moon. Easter Sunday is the following Sunday. 82.163.24.100 (talk) 13:55, 15 June 2008 (UTC)

Thankyou all for your guidance, and in particular, clarification of appropriate use of definitions and descriptions. I have made several clarifications and minor corrections to improve the quality of this topic. Ron Mallen (talk) 04:04, 23 July 2008 (UTC)

The article has been altered recently to use "after 20th March" instead of "on or after 21st March". That is entirely wrong. It is well-known that the nominal Equinox date, March 21st, is used in the calculation of Easter. Introducing a form of description other than that in the Calendar Act and the Book of Commpn Prayer of the church of England is sheer folly, and will lead to confusion. The form "on or after 21st March" is both customary and entirely clear; that is what must be used. 82.163.24.100 (talk) 14:44, 3 August 2008 (UTC)

I have taken care to not touch the definitions, leaving the correct wording "on or after 21 March". I have added a succint, clear and accurate description of the definition that includes, among other things, the clearer phrase "after 20 March". Sadly, as above, it is a fact that "on or after" is frequently incorrectly abbreviated. This has been wrong for years in many published encyclopaedia, and this mistake continues to be repeated in online definitions including an instance in the main article here until I corrected it ;-). I think we all agree that "on or after 21 March" is historically correct; I also believe that a clear desription including "after 20 March" is the best way to minimise the ongoing confusion. Ron Mallen (talk) 14:11, 10 August 2008 (UTC)

Compare the original version, 10 August 2008, with the subsequent corrected version of 13 August. It very clearly indicates the error and confusion that can be caused by any reference to March 20th, when that is combined with the well-established "on or after" for the Paschal Full Moon. Eschew the 20th, except maybe for putting "(i.e. after March 20th)" after one non-definitive reference to on or after the 21st. 82.163.24.100 (talk) 10:48, 13 August 2008 (UTC)

Structure of Article

Perhaps the Article should be more clearly subdivided :-

1 Considerations leading to the Rules

2 About the actual, official Rules annexed to Bull and Act, and how they are decreed to be used

3 Algorithms derived from the Rules, suited to modern use, with particular reference to efficient, traceable ones

4 Consequences of the Rules each subdivided into Dionysian (Julian) and Gregorian.

In each case, Julian should precede Gregorian; and the introduction should say so.

If copyright permits, there should be legible images of the necessary material as in the Book of Common Prayer : for Gregorian, that amounts to one paragraph plus three pages. The paragraph is EASTER-DAY (on which the rest depend) is always the first Sunday after the Full Moon which happens upon, or next after the Twenty-first Day of March ; and if the Full Moon happens upon a Sunday, Easter-day is the Sunday after. The pages are the one defining the Golden Number, and the two containing Tables I to III. Images too small to be read, but recognisable, of those three pages are under 33 and 38-39 in http://www.merlyn.demon.co.uk/estr-bcp.htm .

82.163.24.100 (talk) 16:33, 17 April 2008 (UTC)

Perhaps, between section 2 The Official Rules and 3 Algorithms, one could have a section on rules equivalent to the official rules as an aid to understanding and seeing that the following algorithms are correct. The March 20 rule could be put there if deemed necessary. Karl (talk) 08:12, 18 April 2008 (UTC)

Call it pro tem "Section 2.5, Alternative Expressions". IMHO, any mention of March 20 will lead to confusion. Those who cannot understand "on, or next after" are beyond hope. The formal statement above is now imaged on the cited page. The Sunday Letter, given by Table I and explanation, can be readily found from a larger, but simpler, Table, needing less explanation, loc. cit.. It should be noted that the Epact, while no doubt considered in the construction of the Tables, plays no part in the use of the Tables; it exists in my copy of the Book merely as an unexplained column of page 35. Immediately adjacent to the first use of "Full Moon", the description "ecclesiastical" should appear; that is sufficient to indicate that it is not exactly the one in the sky. 82.163.24.100 (talk) 11:38, 18 April 2008 (UTC)

14th Day or Full Moon?

Consider the second paragraph of the Article, before the History section.

The Book of Common Prayer of the Church of England, in a Preface section decreed by the Calendar Act, does not use the fourteenth day as such; it refers to the Full Moon. I suspect that Catholic sources do explicitly refer to the fourteenth day.

It seems inappropriate to say that the canonical version uses the fourteenth day, without qualification. I suggest a transposition to : The canonical rule is that Easter day is the first Sunday after the nominal full moon (the 14th day of the lunar month) that falls on or after 21 March (nominally the day of the vernal equinox).

82.163.24.100 (talk) 19:58, 20 April 2008 (UTC)

There are at least two "canonical versions", the 1751 Anglican version already quoted and the original 1582 Catholic version, which is the first sentence of the sixth canon attached to the papal bull [3]: "Pursuant to the order of the holy Council of Nicaea, Easter, on which other movable feasts depend, is to be celebrated on Sunday immediately following the fourteenth day of the first month (the Hebrews called the first month the lunar month in which the fourteenth day coincides with the vernal equinox, i.e. March 21, or the nearest following)." Other canonical versions probably appear in the German and Swedish laws implementing the Protestant reformation astronomical and Gregorian Easters. — Joe Kress (talk) 07:39, 22 April 2008 (UTC)

Then let The canonical rule is that be omitted : Easter Day is the first Sunday after the nominal full moon (the 14th day of the lunar month) which falls on or after 21 March (nominally the day of the vernal equinox). 82.163.24.100 (talk) 10:26, 22 April 2008 (UTC)

Tabular Methods

The current Tabular Methods section describes the historical arguments by which the dates of Easter were decided. As far as I know, the detailed authority in the Catholic tradition is in the Explicatio of Clavius.

But, for the British Empire and (then) Colonies, the authority is the Calendar Act Annexe of 1751; and, as its method is required by the Act to be in the Prayer Book, that seems also to be the general Anglican authority. Naturally it is chosen to agree with the Catholics.

So, in practice, the method of determining British (and American, etc.) Easter from Authority is quite different from what is currently described in the Article. In general, one determines the Golden Number, then uses three Tables to determine the Sunday Letter, a Cypher, and the date of the Paschal Full Moon, from which the date of Easter Sunday follows; but a simpler Table can be used for 1900-2199. Clavius' details were employed in the construction of the method, and play no subsequent part in its use.

British Easter before 1753 was also defined by a Table in the Prayer Book, the use of which is not described in the Article.

http://www.merlyn.demon.co.uk/estr-bcp.htm, in deriving an algorithm traceable to the Book, verifies its processes by computing matching Tables; but assumes that a description of how to use them is at hand.

I don't yet know whether there is, on the Web, a nicely-legible representation of the Easter material in a current C of E Book of Common Prayer.

It might be held that such methods do not fall unter the title Computus; but they should at least be mentioned herein. Perhaps Computus should be reserved to be (mainly) for material from before, say, 1750, with another page on calculation and consequences.

82.163.24.100 (talk) 12:07, 26 April 2008 (UTC)

On page 240 Butcher states that before 1662, the English Prayer Book did not have a definition for Easter day. On the next page he states that in 1662 [through 1752] the Prayer Book included the definition: "Easter Day is always the first Sunday after the first Full Moon which happens next after the one and twentieth day of March. And if the Full Moon happens upon a Sunday, Easter Day is the Sunday after." He notes that this is wrong because it does not allow Easter to occur before March 23—it does not allow the Full Moon to occur on March 21. — Joe Kress (talk) 07:31, 27 April 2008 (UTC)

I cannot swear to "not before 1662", but what I've heard supports that. The Prayer Books of 1662 & 1683 do have the bad version. The Act (in S@L 1765) cites in S.III the bad version as being in the then-current Book; its Annexe gives, for inclusion in the Book, the correct words, used to this day.

The Books of 1662 & 1683 have a Table indexed conventionally by Golden Number and Sunday Letter giving the Easter date; Mar 22 is at XVI,D and Apr 25 at VIII,C. The merlyn page reproduces that by calculation using the JPFM algorithm deduced from other sources, amd. has a copy of its instructions. 82.163.24.100 (talk) 18:57, 27 April 2008 (UTC)

Eastern Orthodox Easter and Passover

I just removed a phrase asserting that Eastern Orthodox Christians "observe the additional rule that Easter may not precede or coincide with the first day of the Jewish Passover." This bit of folklore is widely believed among Orthodox Christians, but it is spurious. The canons clearly prohibit us from taking the Jewish Passover, as currently calculated, into account. (The Jews adopted a new method of determining the date of Passover sometime around the fourth century, and Christians were instructed to ignore the new Jewish practice.) BALawrence (talk) 05:25, 30 March 2008 (UTC)

Late Jewish tradition attributes the Rabbinic calendar to a mid-4th century sage, but there isn't a scrap of contemporary evidence for this tradition. The Rabbinic calendar in its present form dates from the age of the Geonim (9th/10th century).

As I read the history, Christians in the 4th century were not being instructed to ignore "a new Jewish practice," but to discard the traditional Christian practice of consulting their Jewish neighbors to learn when the week of Unleavened Bread would fall, and setting Easter for the Sunday in that week. This was considered unsatisfactory because the Jewish calendar at the time was not unified. Jews in one town might have the week of Unleavened Bread at a different time from Jews in another town.

Maybe we should address the eastern church's Passover superstition explicitly, rather than passing over it in silence. It goes back at least to the time of Johannes Zonaras. People reading this article might expect to find in mentioned.--Mockingbird0 (talk) 01:01, 30 August 2008 (UTC)

Jitter?

This statement:

However, there is some protection of the lunar calendar against the errors of the solar calendar. The leap days are not inserted in an optimal way to keep the calendar synchronized to the solar year. The corrections to the leap day scheme are limited to century years, and add two nested intercalation cycles (100 and 400 years) around the four-year cycle. Each cycle accumulates an error, and they add up to more than two days. So in the Gregorian calendar, the actual dates of the vernal equinox are scattered over a time window of about 53 hours around 20 March. This may be acceptable for a calendar period of a year, but is too much for a monthly period. By separating the "solar equation" from the "lunar equation", this jitter is not carried to the lunar calendar. If we were to combine the solar and lunar equations and spread the net 4×8 - 3×25 = 43 epact subtractions in 10,000 years evenly, then the solar jitter would also affect the lunar calendar, which would be unsatisfactory

Needs to be supported, or deleted. It is not obvious that the lunar side of the computus needs to do anything but maintain agreement between the ecclesiastical and empirical moons. And it is not obvious that the disagreement between the ecclesiastical and astronomical equinoxes would have any effect on the lunar-side agreement.--Mockingbird0 (talk) 01:01, 30 August 2008 (UTC)

iCalendar/RRULE

I'm working on an ical specification for easter and have been unable to find any notes on this. Since iCal is the standard method of recording events these days is there any chance we could add information about icalendar implementations of easter? Nickjost 23:57, 23 October 2007 (UTC)

Your meaning is not entirely clear to me. If a definition of the date is required, one cannot do better than the original British Calendar Act - that was inherited by the American colonies and will still be in force there unless replaced by later legislation. If an algorithm for the date is needed, then that derived by merlyn from the Act seems unlikely to be beaten; but it can only be implemented for iCal by someone with a knowledge of an applicable language and with test facilities. 82.163.24.100 (talk) 11:34, 4 May 2008 (UTC)

Knuth's algorithm having recently been read, merlyn's best is now slightly faster. It does not rely on Knuth, though, as the changes are self-evident in hindsight. 82.163.24.100 (talk) 11:46, 8 September 2008 (UTC)

Age and epact

Tom has changed the text on the assumption that the age of the moon must be real (not whole) number. This need not be the case in the English language at least. Applying this assumption causes the text to be more complicated as a result.

How old is Tom? What is Tom's age? 35.767 or 35? (not actual age).

So I think the age of the moon in days as a whole number is a valid concept (in English at least) and could be retained on this page for simplicity. Perhaps the Dutch language does not allow whole number for its equivalent to age. -- Karl 1-Aug-2003 09:45 UT

Uhm, where does this apply? -- Tom Peters 12-Aug-2003 21:07 UTC

I'm referring to the change on 31 July "TP: elaborations on epact<->dayte in lunar year<->age of Moon. Could be explained without reference to "age" (continuous real number), but only "day of month" (ordinal number).) " concerning the section Some Theory. I've just made a small change there that should clarify the moon age references. -- Karl 13-Aug-2003 09h UT

I disagree with your insertion "(moon age 0 at solar new year's day)". The intention is to start the lunar year at the same time as the solar year, i.e. the date (ordinal number) is 1. The epact is 0. The epact is the difference in date between the lunar and solar year. This is equivalent to the age of the Moon, but that is secondary. Anyway, the age of the Moon on 1 Jan. is 1, not 0 (hence the tedious explanation of epact as the age of the Moon on the day before the solar year starts). Think of it as the crescent Moon, which marks the 1st day (ordinal number) of the month, being already (at least) 1 day old. This is also what I state in the 2nd par. in the section on the Gregorian computus; your insertion is in conflict with that text. -- Tom Peters 12:31, 13 Aug 2003 (UTC)

I disagree with Tom's assertion that the age of the moon is 1 on new year's day when the epact is 0. This contradicts the first paragraph of the Catholic Enclopaedia article. Referecenced in Epact. Also the New Moon in this context is always the First Cresent and the astronomical new moon does not come into this. SeeTalk:Epact. -- Karl 15 Aug 11h UT.

Tom is right. The Epact is the age of the moon on January 1st diminished by one. If the Epact is 0, the age on January 1st is 1.--Mockingbird0 (talk) 04:43, 30 August 2008 (UTC)

Algorithms: Meeus/Jones/Butcher (Gregorian)

Implementing this in a calendar program shows that it is correct, but the names of the variables on this page make it hard to understand what is actually going on. I've given it a try (read as: mostly guess-work) given my limited knowledge, maybe someone with more knowledge about Computus can fix my incorrect assumptions - and then put it into the article (I realise some of it is trivial, e.g. the century bit, but for completion):

a = golden_number
b = century
c = year_in_century
d = gregorian_cycle_number (as in, the howmanieth cycle it is)
e = gregorian_cycle
f = ?
g = ?
h = epact
i = gregorian_cycle_number_of_year_in_century - or basic attempt at figuring out a leap year?
L = gregorian_cycle_of_year_in_century - or basic attempt at figuring out a leap year?
m = ?

I've tried to wrap my head around the constants whilst trying to figure this out, too:

451 = ?
114 = MAX_EASTERDATE_AS_DAYOFYEAR - am I right that this is supposed to represent the 25th of April?
100 = YEARS_IN_CENTURY
19  = MAX_GOLDEN_NUMBER
32  = ?
30  = MAX_EPACT
25  = ?
22  = ?
15  = ?
11  = LUNAR_OFFSET
8   = ?
7   = EMBOLISMIC_MONTHS
4   = YEARS_IN_A_GREGORIAN_CYCLE
3   = ?
2   = ?

I would love to see my mistakes corrected and the bits I don't understand explained. I know this is not much to work with, but I'd really like to see the algorithm explained. Notably, I'm aware that often in purely mathematical calculations of dates, an "offset" has to be defined as a constant so the result is not shifted by some amount - so I realise that not all numbers used must neccessarily have another meaning but for, shall we say, 'error correction' in the algorithm; but even that, I find, ought to be documented.

-pinkgothic 14:35, 4 April 2006 (UTC)

There are 8 adjustments every 2500 years so that is where 8 and 25 come from. 15 represents the year 1500. We are counting adjustments from 1500. There are 3 adjustments every 400 years due to the Julian/Gregorian conversion. The number 114 represents April 19th. That means 95 represents March 31st and 96 represents April 1st. If we take this number, subtract 6, then divide by 30 (greatest integer), this will represent the month minus 1. 32 represents 32 adjustments every 10000 years. 10000 years is a complete adjustment cycle with the 8 every 2500 years and 3 every 400 years for the Gregorian/Julian conversion. 22 sounds like it's 25 minus 3. --Trust101 18:41, 28 May 2007 (UTC)

Spelling of Meeus corrected in heading. Knuth has a more efficient algorithm, which deserves at least to be cited - and perhaps someone could discover its traceability to formal authority (i.e. Bull or Act). It's about as fast as the best of merlyn. 82.163.24.100 (talk) 19:20, 24 September 2008 (UTC)

Leap year clarification

The article reads as follows:

Note that leap days are not counted in the schematic lunar calendar: The cycle assigns to the first day of March after the leap-day the same age of the moon that the day would have had if there had been no leap-day.

Would it therefore be unacceptable to treat the second lunation as long instead of short in a leap year, i.e. splitting the xxv/xxiv label for February 5 so that xxv is February 5, xxiv is February 6, xxiii is February 7 and so on (making the last day in February always i) and marking February 5 with 25 instead of February 4? I understand the deal with Dominical letters and the fact that you must skip February 29 when assigning them (I'm assuming the 1752 Anglican rule for the purposes of my question). I am unaware of a situation whether the second lunation being long or short in a leap year matters, but then again I'm not an expert on the subject. 70.120.227.18 (talk) 17:24, 27 September 2008 (UTC)

In the Article, care should be taken to discriminate clearly between the Computus as it stood immediately before/after the publication of the 'Explicatio' and any more modern developments. 82.163.24.100 (talk) 18:02, 28 September 2008 (UTC)

"The 1752 Anglican rule" must mean as in the Annexe to the Act (dated 1750) of 1751, and the consequent part (Tables I II III and associated text) of the beginning of the Prayer Book. The Act and Book give an algorithm, designed to match the 1582 Catholic Rule, but without derivation. So such details as lunations are really nothing to do with the Anglican Rule. 82.163.24.100 (talk) 18:02, 28 September 2008 (UTC)

Notes

The present Note 11 says : "the [Golden Number] of a year AD is found by adding one, dividing by 19, and taking the remainder (treating 0 as 19)." Blackburn & Holford-Strevens p. 810."

It would be better to use a reference saying that the Golden Number of a year AD is found by dividing by 19, taking the remainder, and adding one. 82.163.24.100 (talk) 15:59, 19 October 2008 (UTC)

The Book of Common Prayer provides a more authoritative reference for the value, but expresses it like B & H-S do. 82.163.24.100 (talk) 16:07, 19 October 2008 (UTC)

Dionysius Exiguus states in Argumentum 5 of his Easter tables (525) that to find the year of the nineteen year cycle add one before dividing by 19. Bede also states in chapter 58 of De Temporum Ratione (725) that one is to be added before dividing by 19. The term "Golden Number" was invented by Alexander de Villa Dei in 1200 in his computistic poem "Massa Compoti". — Joe Kress (talk) 07:30, 22 October 2008 (UTC)

One should discriminate between a definition and a recommended method of calculation. The definitions - Calendar Act of 1751, Prayer Book, Explicatio, etc. all add one first. But that cannot be recommended as a way of calculation, for which adding one last is better. Just before the citation of the present Note 11, the good method for Golden numbers is shown; B & H-S is already in the References. That is why making Note 11 include a reference for the "Good" method would be an improvement. Nore that Gauss and Meeus (as shown in the Article) and others all use (Y mod 19). 82.163.24.100 (talk) 16:15, 23 October 2008 (UTC)

Au contraire! Adding one before dividing by 19 was the recommended method of calculation throughout history, when words were used, not symbols. Only with the advent of electronic computers can the alternative be adventatious because it eliminates an if-then construct. — Joe Kress (talk) 00:22, 24 October 2008 (UTC)

Please re-read what I wrote. The Article should give the "official" definition, but where it recommends a method that should be the best method. The link after the equation in the "Theory" section should go to an authority which can be quoted as giving that equation. If there is no such "bookish" authority available, the link can go to the original Calendar Act and/or to the relevant part of the Explicatio and/or to some much earlier "Julian" source, and with (for non-mathematicians) a statement of equivalence. The alternative is better because it eliminates an operation; the fact that the operation may involve a break in flow is less important. I hypothesise that the original form was employed because it avoided the use of zero; "no remainder" is not the same as "remainder of zero" until zero is known about. 82.163.24.100 (talk) 19:20, 25 October 2008 (UTC)

Adding one before dividing by 19 was not an official definition—it was the recommended method of calculation. Of course its source should be mentioned, but don't call it an official definition. At the time, humans were performing the calculation, who immediately recognized whenever there was no remainder, replacing it with 19. Only when dumb machine calculators are used can the if-then clause be regarded as an extra operation. I don't believe that zero was not known to Dionysius Exiguus—the first epact within the table of epacts was indicated by the Latin word nulla. All other entries were Roman numerals, implying that nulla was also a number, so some translators translate it as zero. (However, no remainder was nihil, which is always translated as nothing.) Furthermore, Dionysius was fluent in Greek and Greek astronomers were regularly using zero. Its symbol was a small circle with a long overbar, see Greek numerals#Hellenistic zero. The Alexandrians, who wrote in Greek, would have used this Hellenistic zero for the first epact. Roman numerals did not have a symbol for zero, so Dionysius was forced to use a Latin word, nulla, in place of the Greek zero. — Joe Kress (talk) 05:41, 26 October 2008 (UTC)

Date of the equinox

The Article indicates clearly enough that March 21st is the date, fixed on the Julian and Gregorian Calendars, which is the starting-point for the Easter determination. And it indicates clearly enough that March 21st refers to the Equinox; and that it can do so only nominally, because of the slight motion around that calendar date of the actual Equinox. (The drift of the Equinox through the calendar, because of the mis-match between the Solar and Calendar years, can be ignored.)

But it is not entirely clear when the nominal Equinox date was fixed at March 21st. Was it fixed on the Roman Calendar in Caesar's time or later, so that the date was considered unquestionable at Nicaea? Or did Nicaea decide on March 21st? Or was it quietly fixed by people deriving the numerical solution based on what the Council had said? Or when? Of course, we can easily see that it was fixed in 1662, and no doubt it was fixed many centuries before that; the Article indicated that it was probably fixed before the time of the East/West Schism.

82.163.24.100 (talk) 20:36, 21 October 2008 (UTC)

The traditional date of the Roman equinox was March 25 as described by Pliny the Elder about AD 77. March 21 was chosen by the Church of Alexandria (not Rome) sometime between 277 and 311, well before the Council of Nicaea. Either in 258 or 277 Anatolius of Laodicea, originally of Alexandria, chose either March 18 or 22 as his equinox (Eusebius is not clear on this point). The first year of the earliest Easter table which uses March 21 as its equinox is 311, preserved as an Ethiopean copy described by Otto Neugebauer in Ethiopic Astronomy and Computus (1979). Venance Grumel in Chronologie (1958) thinks that it was chosen by Bishop/Pope Peter of Alexandria during the first decade of the fourth century. The Council never chose an equinox. That is a false statement made by the Roman Catholic Church in its 1582 papal bull Inter gravissimas, based on a false statement by Dionysius Exiguus that the Alexandrian method of computing Easter was sanctioned by the Council. The Church of Rome did not accept March 21 as its equinox until about 342. — Joe Kress (talk) 07:30, 22 October 2008 (UTC)

In fixing the date of Easter, the date chosen for the role of the Equinox is as important as the dates of the Ecclesiastical Full Moons. So I suggest that the essence of the choice of March 21st needs to be worked into the Article. 82.163.24.100 (talk) 15:57, 23 October 2008 (UTC)

Although we can identify approximately when March 21 was chosen, we do not know why. The best source for the equinox at the beginning of the fourth century was Ptolemy's Almagest, whose tables show that March 22 should have been chosen, using the Almagest Ephemeris Calculator. Modern back calculations showing that the equinox was March 19 in the Julian calendar at the beginning of the fourth century also don't help. — Joe Kress (talk) 00:22, 24 October 2008 (UTC)

The approximate date of the choice of March 21st as the "base date" can be indicated, even if the reasoning is lost and/or it is not known whether that was then thought to represent the correct best Equinox date fot that or some earlier time. They could have wanted the Julian date of the Crucifixion-year Equinox. 82.163.24.100 (talk) 19:29, 25 October 2008 (UTC)

Both Eastern and Western Christian writers of the third century (Africanus and Hippolytus) agree that the Julian date of the equinox was March 25, at least at Creation. The synchronism of Creation to the Passion of Christ meant that March 25 became the date of either the Crucifixion or the Ressurection. Because this is not March 21, I don't understand your last sentence. I agree with your first sentence. — Joe Kress (talk) 05:41, 26 October 2008 (UTC)

The method given by the British Calendar Act and the Church of England Book of Common Prayer

Inserted Section for Article, present position 3.1.2

The Tabular Methods section describes the historical arguments and methods by which the present dates of Easter Sunday were decided in the late 16th century.

In Britain, where the Julian Calendar was then still in use, Easter Sunday was defined, from 1662 to 1752 (in accordance with previous practice), by a simple Table of dates in the Anglican Prayer Book (decreed by the Act of Uniformity 1662). The Table was indexed directly by the Golden Number and the Sunday Letter, which (in the Easter section of the Book) were presumed to be already known.

For the British Empire and Colonies, the new determination of the Date of Easter Sunday was defined by the Calendar (New Style) Act 1750 with its Annexe. The method was chosen to give dates agreeing with the Gregorian Rule already in use elsewhere. It was required by the Act to be put in the Book of Common Prayer, and therefore it is the general Anglican Rule. The original Act can be seen in the British Statutes at Large 1765.

The method is quite distinct from that described above in Gregorian calendar. For a general year, one first determines the Golden Number, then one uses three Tables to determine the Sunday Letter, a Cypher, and the date of the Paschal Full Moon, from which the date of Easter Sunday follows. A simpler Table can be used for limited periods (such as 1900-2199) during which the Cypher (which represents the effect of the Solar and Lunar Corrections) does not change. Clavius' details were employed in the construction of the method, but they play no subsequent part in its use.

J R Stockton shows his derivation of an efficient computer algorithm traceable to the Tables in the the Prayer Book and the Calendar Act (assuming that a description of how to use the Tables is at hand), and verifies its processes by computing matching Tables.^[1].

If arguments to the contrary do not soon appear, I intend to put the above in the Article.

I have now done that. 82.163.24.100 (talk) 19:02, 17 October 2008 (UTC)

Addition to that insertion made. 82.163.24.100 (talk) 19:25, 10 November 2008 (UTC)

Some of the "Details" section which the above now follows is solely dependent on the dates which have been chosen and not on the reasoning or calculation leading to those dates. The frequency plot is an obvious example. I suggest a new section, 3.1.3 as things stand, with a name like "Consequences", into which all such would be moved from other parts of Section 3. Another example might be the 5700000 years part, though adaptation would be needed as the figure 5700000 is not derived in the same way for each method. 82.163.24.100 (talk) 19:02, 17 October 2008 (UTC)

The Discussion is rather long now, and some is old. The original of the transplanted section can go quite soon. How do Discussions get prevented from tending towards infinity? 82.163.24.100 (talk) 19:02, 17 October 2008 (UTC)

The Prayer Book Julian Table should go in the Article. There is a copy at http://www.merlyn.demon.co.uk/estrtbls.htm#OT. The Julian table presently in the Article can be deduced from it since its dates are a week before the latest Book Table entry for the Golden Number.

ASIDE : The original Calendar Act and the Prayer Book tabulate, but do not define, the Epact. What algorithmic definition of the Epact is there applied?

82.163.24.100 (talk) 17:53, 4 May 2008 (UTC), 82.163.24.100 (talk) 11:30, 22 September 2008 (UTC)

I discovered that "1662" refers to any version of the Book of Common Prayer published after 1662, not to the date it was printed, which could be in the 17th, 18th, 19th, or 20th centuries. I found a Google copy of the 1765 Book of Common Prayer with its Easter tables. The Anglican site has the old 1549 and 1552 versions of the Book of Common Prayer. They only include an Almanac for 19 or 30 years without any instructions on how to use it. I haven't found any version printed between 1662 and 1752 that include the Easter tables online. Epacts tabulated in a table in the 1750 act and in the Book of Common Prayer are not needed to determine any holiday, so they are not defined. — Joe Kress (talk) 20:20, 25 September 2008 (UTC)

1662 : I only half agree. There was a liturgical upheaval in 1662, and all Books with substantially that liturgy can be called 1662. But as regards Easter it is different - pre-1662, no Easter rules; 1662-1752 Julian Rules; post-1752, Gregorian rules. And in this Discussion and this Article, it may be necessary to refer to a more-or-less specific printing. Example : I found a newly-printed 1662 Prayer Book in our local Library - it had the 1662 Liturgy, but Easter Tables as in the current Prayer Book.

1765 Prayer Book link - noted, accessible, thanks.

1662 and 1683 Easter pages are on-line, but privately.

Epacts for the 1765 Book are copied from the Act. But later Books have later Epacts; and those require definition where not listed by Law, quite regardless of whether thay are of any use. merlyn derives an Epact algorithm, but not from True Authority. 82.163.24.100 (talk) 22:04, 25 September 2008 (UTC)

-

I think I've dealt with the following; feel free to adjust residuals. "above" - material should go above the Julian section, so it would have been clear (the "Tabular Methods" section had already been mentioned). Existing article section "Gregorian Calendar" cites 1582, so is I suppose Catholic-derived. "Catholic" could be added, for reference. 82.163.24.100 (talk) 22:04, 25 September 2008 (UTC)

Don't refer to another section within the article unless necessary. Simply state the Anglican method. I don't know what method you are refering to when you say "that described above" within "The method is quite distinct from that described above." You may be referring to the 1662-1752 Anglican method or to the Catholic method. Please be specific. I only see minor problems, including Wikipedia formatting problems, which cannot be handled on this talk page, only within the article. One of these is a naked URL, not allowed on Wikipedia. Furthermore, even when properly named, all external URLs should be treated as references within footnotes, never within the text of the article. Within the text they are simply identified, for example, "J R Stockton derives …"<ref>[http://www.merlyn.demon.co.uk/estr-bcp.htm The calculation of Easter Sunday]</ref>. — Joe Kress (talk) 20:20, 25 September 2008 (UTC)

Do not refer to any section by number. Those numbers are automatically assigned within the table of contents but do not appear within the text of the article, and will change if someone adds another heading. You can hyperlink any heading within the article by prefixing it with a '#' as in [[#Gregorian calendar|Gregorian calendar]], for example. I was flummoxed by the '1662' used by many sites, including the site which has the complete text by Lynda Howell, entitled The 1662 Book of Common Prayer Website. I assumed that '1662' meant the year of printing. Although she claims to show all differences between the various printings, she omits any differences for the Easter tables. Howell's tables do not include any epacts. My local university library catalog has links to online copies of many printings between 1632 and 1710, including the directory used during Cromwell's 'reign', most of them within Early English Books Online, as well as a modern printing. — Joe Kress (talk) 03:29, 26 September 2008 (UTC)

That will be a "liturgical" 1662. The body of the C of E B of C P is due to the Church; much of the preliminary material (including Easter) is due to the State - Sovereign/Parliament (under advice, no doubt). FYI, apart from the choice of the "52 years", the 1815 Book is textually equivalent to the 1765 Book, but much better printed. Caveat : I note that I've been misusing "preface" to refer to all up to the end of the Easter Tables. 82.163.24.100 (talk) 20:03, 26 September 2008 (UTC)

For what its worth, one of the sites I found used the term "front matter" for everything before the "Morning Prayer" section. — Joe Kress (talk) 03:45, 27 September 2008 (UTC)

Ugly!

I've unstruck "above in"; it reads better, and that should be above. For other tweaks, see History Differences. 82.163.24.100 (talk) 16:57, 27 September 2008 (UTC)

Early English Books Online (EEBO) has hundreds of copies of the Book of Common Prayer from 1549 to the nineteenth century. Between 1645 and 1660 A directory for the publique worship of God forbade Easter. In keeping with its name, the period 1549–1700 was well represented by EEBO, with fewer copies post-1752. A gap with no copies exists between 1701 and a few years after 1752. Many/most of the years 1549–1700 are represented, with a few years represented by ten or more copies. The latest pre-1662 copy of the Book of Common Prayer was for 1660. The latest copy I found anywhere is still the 1987 copy on the "1662 Book of Common Prayer Website". One e-book of a 1978 Episcopal Church version had Easter tables near the end of the book, but they had lost their tabular format so were quite garbled.

The 1549 version had no Easter tables.
The 1552 version had an almanack for xix years (MDlii–MDlxx) giving Easter, plus the golden number, epacte, circle of the sonne, and dominicall letter, without instructions.
The 1588 version had an almanacke for 36 years (1578–1610) giving Easter and six other moveable feasts, plus the golden number, epact, and dominicall letter, which, unlike most other copies, had instructions on how to calculate all of them (the EEBO copy is damaged).
The 1596 version included an almanacke for xxvj years (1578–1603) giving Ashwednesday and Easter, plus the golden number and Sund. letter, without instructions.
The 1625 version had an almanacke for xxxix yeeres (1603–1641) giving Easter and six other moveable feasts, plus the golden number and dominicall letter, without instructions.
The 1629 and 1630 versions had an almanack for xxxii years to come (1629–1660) giving Easter and six other moveable feasts, plus the golden number and dominicall letter, with instructions on how to calculate the golden number but not the dominicall letter (although not tabulated, the epact is mentioned in a note).
The 1640 version had an almanack for xxvj years (1622–1648) giving Easter and two other moveable feasts, plus the golden number and dominicall letter, without instructions. It also had a table to finde Easter for ever via the prime (golden number) and Sunday letter, without instructions.
The 1660 version had an almanack for xxxviii years (1648–1685) giving Easter and six other moveable feasts, plus the golden number and dominical letter, with instructions for the golden number.
The 1662, 1684, 1693, and 1700 versions had a table and rules for the moveable and immoveable feasts; a table of [Easter and six other] moveable feasts calculated for fourty years (1661–1700 for 1662, 1684, and 1693; 1700–1739 for 1700), plus the golden number, epact, and dominical letter, without instructions; and a table to find Easter for ever via the golden number and Sunday letter, without instructions. — Joe Kress (talk) 03:18, 1 October 2008 (UTC)

I don't see a clash with the proposed insertion. One probably should not assume that every printer of the old Prayer Book actually got everything right.

I suppose that in those days anyone literate enough to read the Prayer Book would have been smart enough to see how to use the Julian Table, if previous material indicated what the Number and Letter were. The merlyn page which computes that Table also shows its inſtructions.

In http://books.google.com/books?id=aeCZahwsCioC&q=act+of+uniformity+1662&dq=act+of+uniformity+1662&lr=&ei=9tjjSNKrHIquywSL-7neBA&pgis=1 (The BCP from the Act of Uniformity 1662) I get only the "About this book" tab; but for those more fortunate it seems to provide a fairly definitive 1662 Book. EEBO seems inaccessible to me - links for the Article need to be checked by a domestic non-American user.

82.163.24.100 (talk) 20:48, 1 October 2008 (UTC)

The list of EEBO copies was meant to record their individual descriptions, not as a suggestion that they need to be added to the article. To access EEBO, you must use a terminal physically inside a university library that subscribes to EEBO—check the catalog of the library before traveling to that university. Only one of many EEBO copies of the 1662 printing was used for this comparison.

The Book of Common Prayer from the original manuscript attached to the Act of Uniformity of 1662, and now preserved in the House of Lords is a faithful transcription of the 1661 manuscript into modern (1892) Roman and Italic fonts. All punctuation and spelling errors were faithfully reproduced, including the scribe's choice of u/v and i/j forms, but the long s was replaced by the modern 's'.

Spelling is more modern in the printed form, for example "finde" in the 1661 manuscript is "find" in the 1662 printing. All numbers in the 40-year table were Arabic numerals in both forms. Although the values of the numbers in the "Easter for ever" table were the same in both forms, almost all were Roman numerals in the 1661 manuscript but Arabic numerals in the 1662 printing. The Easter for ever table belongs in the article. The January to December calendar was after the Easter tables in the 1661 manuscript but before them in the 1662 printing. — Joe Kress (talk) 07:30, 22 October 2008 (UTC)

Meeus Julian

I tryed calculating this in Excel and it doesn't seem to work... —Preceding unsigned comment added by 141.24.191.221 (talk • contribs) 13:07, 26 October 2008

The algorithm is essentially identical to the algorithm appearing in Astronomical algorithms (1991) by Jean Meeus (p.69). Note that it only calculates the Julian Easter, not the Gregorian Easter used by Western churches. Furthermore, it gives its Easters in the Julian calendar, not the Gregorian calendar. To derive the Eastern Orthodox Easter normally given in the Gregorian calendar, 13 days must be added (1900–2099). — Joe Kress (talk) 00:10, 27 October 2008 (UTC)

Zeller's algorithms for Julian Easter on the Julian Calendar are visible as images via http://www.merlyn.demon.co.uk/zeller-c.htm in the "82px" to "86px" pages, and demonstrated as JavaScript in the "1882" to "1886" pages. They could be used as a cross-check.

And http://www.merlyn.demon.co.uk/estrdate.htm#OE has a non-authoritative calculator including Julian Easter on the Gregorian Calendar.

I think the Article might say a little more, if known, about Orthodox Easter, even if only by adding links. 82.163.24.100 (talk) 21:34, 28 October 2008 (UTC)

The Papal Bull and the Explicatio

It seems strange that the strings "Papal Bull" and "Inter Gravissimas" and "Explicatio" do not appear in the Article. The first two are indeed in the linked "Gregorian calendar" page. Also, apart from in the "British" part, no "Clavius".

Note 12 links to certain canons in Latin and French, but it's not obvious what authority they have, especially for monoglots. 82.163.24.100 (talk) 21:34, 28 October 2008 (UTC)

FYI : The merlyn site now has, at http://www.merlyn.demon.co.uk/estr-xpl.htm, a literal implementation of the algorithm in Christopher Clavius's six Canons. Representations of CC's Tables are shown, and the algorithm actually reads them and uses them in the manner prescribed by CC. The code is therefore moderately slow - Easters appear at only about 3E10 times the rate of real life - but that is much faster than the immediate intended readership could have determined them.

I expect that readers will, correctly, have assumed that the previous paragraph was intended to refer to Gregorian Easter. The same can now be said of Julian Easter. 82.163.24.100 (talk) 12:35, 26 December 2008 (UTC)

CC's actual Tables cover various spans of time; but they can be used for greater periods in a manner that CC would probably have accepted. Results are in full agreement for Gregorian Easter from AD 0 up to above AD 5,700,000 with other respected methods and with the Book of Common Prayer. 82.163.24.100 (talk) 22:05, 8 December 2008 (UTC)

AD 0 is not a date.
What is the first meaningful date for Easter? Perhaps no dated record of early celebrations exist, but AD 20 or so seems more realistic than 1 BC. --Gerry Ashton (talk) 22:49, 8 December 2008 (UTC)

The meaning of AD 0 is perfectly clear to those who understand zero. It is a convenient lower bound for testing. AD years before the middle of the First Millennium are in any case proleptic
By the Bull, Gregorian Easter was valid only from AD 1583. But one can ask when Easter would have been if the Gregorian Rules had applied in earlier years, or one can note that the repeat interval is 5700000 years and so consider "AD N" as a representation of "AD M×5700000+N". Pedants can instead test from 1583 to 5701582.
Easter used the Julian Rule from about AD 325 to 1582; the Canons may well include a sufficient definition of that. 82.163.24.100 (talk) 12:28, 9 December 2008 (UTC)

Date Format

It is clear from reading this Discussion page that the favoured date format is as "March 21" (or sometimes "March 21st") rather than "21 March". The former agrees with other Wikipedia practice; for example, there is an article entitled "February 30". The logical order is "[Year] Month Day" in diminishing order of significance. The Article should be changed throughout, after establishing agreement. 82.163.24.100 (talk) 19:34, 10 November 2008 (UTC)

The format of the article is, with a few mistakes, "21 March". Since there is a consensus for this format, it should not be disturbed. (If this were an American topic, it could be changed, but it isn't an American topic, it's an international topic). Also, the format "2008 November 10" is seldom used in English and is not one of the acceptable formats described in Wikipedia:Manual of Style (dates and numbers).

Also, I don't see any discussion of date format on this discussion page, so don't understand your comment that "it is clear from reading this Discussion page that the favoured date format is as 'March 21'". Just because people use the date format they personally prefer to make informal comments, that does not mean they advocate changing the format of the article. --Gerry Ashton (talk) 19:56, 10 November 2008 (UTC)

Perhaps not; but it does mean that they ought to advocate, support, or accept such a change. (The date/time immediately following ought to appear as YYYY-MM-DD hh:mm UT, which is everywhere unambiguous [ISO 8601]; UT not UTC, since hh:mm cannot represent leap seconds) 82.163.24.100 (talk) 12:14, 26 December 2008 (UTC)

The canonical rule ...

I am not entirely happy with a sentence near the top of the article : "The canonical rule is that Easter day is the first Sunday after the 14th day of the lunar month (the nominal full moon) that falls on or after 21 March (nominally the day of the vernal equinox)."

Firstly, such a sentence should say what it means, grammatically, if the parenthetic parts are removed. But then it means that the month, not just its 14th day, falls on or after March 21st.

Secondly, just what is the canonical rule?

For Anglicans and those in the then Empire, the Rule is in the Calendar Act Annexe, and in the Prayer Book; those talk of the Full Moon, but not of the Lunar Month or the 14th day.

For those who in 1582 should have heeded Pope Gregory XIII, the Rule is as expressed in the Six Canons. Those do talk of the New Moon and the 14th day, in justifying the choice of the Rule. But the actual Rule is the tabular algorithm : look up the Golden Number, look up a Letter, use them to look up the Epact, look up the Dominical Letter, and use those two either with the Calendarium or with Tabula Paschalis Nova Reformata.

How about : "The Full Moon is taken as the 14th day of the Ecclesiastical Lunar Month. Easter Day is the first Sunday after the Full Moon which falls on or after 21 March (the nominal date of the Vernal Equinox)." ?

82.163.24.100 (talk) 22:32, 5 January 2009 (UTC)

Weisstein reference

I think that the Weisstein reference should be removed. It seems to add nothing, is in part wrong about Easter, and the site ignores corrections. 82.163.24.100 (talk) 14:28, 26 January 2009 (UTC)

Gauss Reference

The Article lacks a reference for Gauss's publication of his final algorithm. And so do I.

The Gauss algorithm in the Article has been converted directly to JavaScript and tested successfully for years AD 0 to 123456789 in http://www.merlyn.demon.co.uk/estrdate.htm against an algorithm derived from the Book of Common Prayer. The latter algorithm uses the usual form of the Lunar Correction, but the Prayer Book and Calendar Act do not justify that in perpetuo, only for AD 1600-8600, which is convincing but not formally perpetual. Is it known exactly where the distribution of 8 steps per 2500 years is finalised in the explicatio?

82.163.24.100 (talk) 21:02, 13 February 2009 (UTC)

According to the reference cited in the article (Reinhold Bien, "Gauß and Beyond: The Making of Easter Algorithms", Archive for History of Exact Sciences 58/5 (July 2004) 439−452), Gauss published his final version in C. F. Gauß, "Berichtigung zu dem Aufsatze: Berechnung des Osterfestes Mon. Corr. 1800 Aug. S. 121", Zeitschrift für Astronomie und verwandte Wissenschaften, Volume 1 (1816) (= Werke XI, Abt. 1, 201-202), which I have not consulted. — Joe Kress (talk) 05:39, 14 February 2009 (UTC)

I see that the algorithm in this Article is not exactly the same as that in http://de.wikipedia.org/wiki/Gau%C3%9Fsche_Osterformel - though they look equivalent. The Articles should give the form which Gauß himself gave. Also, I think that they should say whether Gauß followed Catholic or Protestant authority (printed Britannica says that he was religious, but not in what mode). 82.163.24.100 (talk) 17:02, 14 February 2009 (UTC)

The first external link in the German Wikipedia article has links to all of Gauss' Easter articles as they appear in his Works: Die Osterformel von C. F. Gauss - Quellen. The versions in his Works are formatted quite differently than are his original articles, as can be seen by comparing his original first article (1800) to the version in his Works. His 1816 article only corrects the single error in his 1800 article, so his 1800 article is still the basis of his final algorithm. I'm adding the citations to his articles to the Wikipedia article, as well as his Julian Easter. This English Wikipedia article is closer to Gauss' original 1800 article than is the German Wikipedia article, especially with respect to the to the "if" statements. — Joe Kress (talk) 04:02, 19 February 2009 (UTC)

Editing Request

I don't know how to achieve this myself.

In the final paragraph of "British Calendar Act and Book of Common Prayer", there should be a reference for such a description (footnote-type?), and that should refer to the 1765 Statutes and/or to the 1815 Prayer Book (which are currently the last two External Links) and/or to the current Book of Common Prayer of the Church of England (if the requisite part can be found on the Web); but NOT to the alleged current version of the Act cited by the previous External Link. "Easter Tables" of 1765 might be thought best. If that is agreeable and done, this section can then be removed from the Discussion. 82.163.24.100 (talk) 22:58, 15 February 2009 (UTC)

Done. I substituted a 1765 version of the Book of Common Prayer for the 1815 version. — Joe Kress (talk) 06:57, 19 February 2009 (UTC)

Unified system of corrections

Here is my attempt at treating the controversial subject matter:

It would be possible in theory to unify the solar and lunar corrections into a single system of lunar corrections, made generally at intervals of 200 and 300 years and operating in the same sense as the current solar corrections. This scheme would have the advantage that a line of epacts would almost always be valid for at least 200 years, while under the current system a line of epacts can expire after 100 years. The disadvantage of this unified scheme would be that errors in the lunations would accumulate more rapidly. Under the current system, the solar corrections ensure that the lunar calendar falls behind the astronomical facts by about 1 day in 324 years, and the relatively rare lunar corrections then correct for this. Under the unified scheme, the lunar tables would move ahead of the astronomical facts at a rate of about about 1 day in 235 years, which the unified lunar corrections would then compensate.

--Mockingbird0 (talk) 03:37, 12 March 2009 (UTC)

I don't understand the purpose of putting that in the article. The lunar aspects of the Gregorian calendar are used for the religious purpose of calculating Easter. In the past they may have had some other influence on religious observances, and maybe they were even used for the practical purpose of planing activities that required moonlight. The lunar calendar no longer has any practical use, so why would we speculate on possible improvements to the lunar calendar? Wouldn't it be better to just confine ourselves to reporting any reform proposals that come from notable religious entities? --Jc3s5h (talk) 12:44, 12 March 2009 (UTC)

Mockingbird's suggestion is silly, if not deceptive. He removed the explanation why this will not work, which I re-introduced recently, and now essentially supports Lichtenbergs proposition - how's that for NPOV? As for Jc3s5h's comment: that he has no practical use for a lunar calendar does not diminish the fact that a full-blown lunar calendar has always been part of the Gregorian calendar, and will be so forever by definition, until replaced by something else, and is still subject of bona fide academic study as exemplified by the papers of Lichtenberg and Roegel. Indeed the Gregorian calendar reform was motivated more by the problems with the lunat calendar and the Easter computation, than the seasonal drift (see the Vatican's symposium on the Gregorian reform back in 1982). Anyway this issue treats a puzzling aspect of the existing Gregorian lunar calendar, and therefore has a place on this page. The fact that Lichtenberg and now Mockingbird think that the calendar can be improved in this way, make it relevant to explain on this page why this is wrong. I really want this point (separate solar and lunar corrections) explained on this page. Tom Peters (talk) 01:05, 13 March 2009 (UTC)

I should add that Jc3s5h deleted Karl Palmen's re-introduction again for the wrong reason. Indeed we cannot predict the rotation angle of the Earth accurately 10000 years in the future. But this paragraph deals with the distribution of epact corrections as so many per 10000 (or 100000 or whatever) years: on average, once per 232 years. That is a timeframe that is both predictable and observed in the historical record, and indeed the Gregorian reform was introduced because the original Julian computus had outlived its validity of 2 or 3 centuries fivefold. Tom Peters (talk) 01:17, 13 March 2009 (UTC)

The Gregorian Calendar (including Easter) is perpetual; Gregory said so. The article must very clearly distinguish between the true Gregorian Calendar and any possible variations of it (there cannot be changes to it). 82.163.24.100 (talk) 19:59, 5 May 2009 (UTC)

Since the lunar calendar is now a religious matter rather than a practical matter, there is no practical reason to change the Gregorian lunar calendar (which I agree still exists within the religions concerned). I can see no reason to discuss reform proposals unless they are (1) under serious consideration by a notable religious group or (2) serve to illustrate some aspect of the existing system that would otherwise be difficult to explain. There are many hobbyists who like to invent new calendars, but such hobby calendars have no place in an encyclopedia. So pelase explain the significance of Lichtenberg and Roegel's work. Why should we explain the problems with Lichtenberg's work when the work wouldn't deserve mention in this article even if it were correct? (Not because I have anything against the work, but because the easy access to accurate moon data makes the accuracy of a rule-based lunar calendar moot, with the possible exception of religious observances).

I personally find the explaination of separate luner and solar corrections is rather dense; I've understood it in the past, but I don't retain the understanding and the present explaination does not quickly refresh my memory. So I agree an easier-to-follow explanation would be good. --Jc3s5h (talk) 01:39, 13 March 2009 (UTC)

Roegel's is an in-depth analysis of the Gregorian lunar calendar over its entire cycle, not a reform proposal. Lichtenberg takes his own shot at the inner workings of the calendar, and claims that its solar year is adaptable by design - so not an actual reform proposal. He does propose a simplification to the lunar calendar though by combining the two corrections. I think that is misguided for the reasons that I tried to explain. That there are two separate lunar corrections in the Gregorian calendar that sometimes cancel each other requires an explanation, and the obvious proposal to combine them is a step-up to such an explanation. My personal assessment is that the solar correction was an attempt to compensate for the Gregorian correction to the solar year, and bring the lunar calendar back to the original Metonic scheme with a Julian calendar; then the this original Metonic cycle was corrected every 3 or 4 centuries to compensate the inherent mismatch of that cycle. This was only a partially succesful attempt because the corrections take place after so many Gregorian years, not Julian years, leading to a different mean lunation length than may have been intended. Tom Peters (talk) 22:21, 13 March 2009 (UTC)

It sounds like both Lichtenberg and Roegel's work might be useful in explaining the present system, and I hope you are able to work out a helpful improvement to the explaination. --Jc3s5h (talk) 02:07, 14 March 2009 (UTC)

If we had a unified system of corrections, with unified corrections (UCs) in 1900, 2100, 2300, and 2600, then if the calendars were the same in 1800:

For the years 1800-2099 the modified lunar calendar would be the same as the Gregorian (UC and solar correction in 1900)
For the years 2100-2199 the modified lunar calendar would be 1 day behind the Gregorian (UC in 2100, solar and lunar corrections cancel)
For the years 2200-2399 the modified lunar calendar would be the same as the Gregorian (solar correction in 2200, UC and solar correction in 2300)
For the years 2400-2499 the modified lunar calendar would be 1 day behind the Gregorian (lunar correction in 2400)
For the years 2500-2799 the modified lunar calendar would be the same as the Gregorian (solar correction in 2500, UC and solar correction in 2600, solar and lunar corrections cancel in 2700)

I think these differences would be scarcely noticeable.--Mockingbird0 (talk) 18:47, 15 March 2009 (UTC)

Firstly, there is no doubt that the Gregorian calendar has a 53 hour jitter. This is shown in the graph in the accuracy section of the Gregorian calendar article. The 193 years from 1904 to 2196 inclusive have 49 leap days, while 193 years have 193*97/400 = 46.8095. This results a jitter of 49-46.8095 = 2.1975 days, which equals 52.74 hours. This this the 53 hours referred to. This affects the equinoxes of years 1903 and 2196. According to [Stellaphane], there are equinoxes on Sat Mar 21 19:15:19 UTC 1903 and Mon Mar 19 14:02:14 UTC 2096. Similar results can be obtained from year pairs (2303, 2496), (2703, 2896) etc.

Secondly, the unified lunar correction scheme ignores leap years completely, and so all this jitter is transmitted to the reckoned moon phase. The Gregorian scheme does not ignore a leap year at a century year. The moon phases occur usually one date earlier if the century year is a leap year than if it were not a leap year and so usually occur on the same day. This removes some of the jitter caused by the leap days. Within any of these jittery 193-year periods, the Gregorian Calendar rule never corrects the 19-year cycle by delaying the moon phases as must be done on average. If it did, there'd be considerable lunar jitter over the 193 years. I see that from the years (1800, 2100, 2300 and 2600) that this does not happen in the unified scheme for some time, but it will do later on, probably in 2800. If this were so, the Easter moon phases of 2703 and 2896 will show bad jitter in the unified scheme. You may check this with [Stellafane]. Karl (talk) 11:51, 16 March 2009 (UTC)

Mockingbird, you removed the explaning text because you require proof of the Gregorian jitter. The text explains what is going on, also see the graph Palmen refers too. What proof are you looking for? Compare actual Full Moon dates with those from the Gregorian lunar calendar with the two corrections and with the unified correction, for a period of at least 800 years. Tom Peters (talk) 22:34, 17 March 2009 (UTC)

Tom suggested comparing actual Full Moon dates with those from the Gregorian lunar calendar with the Gregorian corrections and with the unified correction, for a period of at least 800 years. Actually comparison is only necessary for those centuries that the Gregorian and Unified differ. These are 2100-2199, 2400-2499 and I believe 2800-2899. A comparison in the intervening centuries would not give a result, because the Gregorian and Unified would be synchronised. Also I'm not sure that 800 years is sufficient. Karl (talk) 09:48, 18 March 2009 (UTC)

I have already shown a comparison for a thousand years. The unified calendar in the century 2800-2899 will not necessarily differ from the Gregorian. There is a certain arbitrariness, in the unified scheme, about where the corrections occur. They could (and probably would) be placed to avoid falling in century years divisible by 400, such as 2800.

In any lunar calendar based on the Gregorian solar calendar, 235 lunations contain 6938, 6939, or 6940 days, depending on where the bissextile years fall within the 19 years. The 6939 and 6940-day cycles are the most common. A 6938-day cycle occurs whenever there is a lunar correction, whether cancelled by a solar correction or not. At the end of a 6938-day cycle, the beginning of the first day of the first lunation of the next cycle (nominally 6PM) moves ahead 1.6 days relative to the mean conjunction from where it was at the beginning of the outgoing cycle. If this is the lunar "jitter" you refer to, it is already a feature of the Gregorian lunar calendar. The 235 tabular lunations beginning on December 15th, 2088, and ending on December 14th, 2107, consist of 6038 days. The 235 tabular lunations beginning on December 17th, 2392 and ending on December 15th, 2411, contain 6938 days. In the Gregorian scheme, as noted, it happens whenever there is a lunar correction. In the unified scheme it would happen whenever there was no correction in a century year not divisible by 400. The question is whether it would be noticeable, given that the true conjunction hops around relative to the start of a tabular lunation anyhow. And if noticeable, would it be more noticeable in the unified scheme than in the Gregorian scheme?--Mockingbird0 (talk) 15:51, 18 March 2009 (UTC)

Mockingbird has pointed out that a period of 235 lunar months can have 6938 to 6940 days in either lunar calendar. If the uniform system were to do a corection in a leap century year a 235 lunar month period of 6941 days would occur. Mockingbird pointed out that a uniform system could avoid correcting on a leap century year. I don't know whether Lichtenberg proposed a uniform scheme that avoids correcting on a leap century year. Such a scheme would take account of some leap years and so not transmit the full 53-hour jitter to the lunar calendar. It would not be as effective at this as the Gregorian lunar calendar.

The Gregorian calendar behaves most of the time like the Julian Calendar. The corrections of the 19-year cycle appropriate to the Julian Calendar are the lunar equation corrections. The only time that Gregorian calendar behaves differently from the Julian calendar is across the common century years. The Gregorian calendar then corrects the Julian lunar corrections, by having a solar correction at each such time. This minimises the jitter, by allowing correction appropriate to the Julian calendar at other times. Karl (talk) 11:22, 19 March 2009 (UTC)

I am still interested in this discussion, but I do not have the time now to contribute. Later. Tom Peters (talk) 00:24, 20 March 2009 (UTC)

I found one important thing that the Gregorian computus does, but no uniform correction computus can do. For each triplet of common century years between the same leap century years, the Gregorian compuus provides at least two corrections. This ensures that the lunar calendar never runs short between leap century years. Karl (talk) 12:57, 20 March 2009 (UTC)

In a unified scheme, a non-correction year is equivalent in its effects to a lunar correction in the Gregorian scheme that is compensated by a solar correction. Indeed, a unified correction scheme could be devised that differs from the Gregorian only in those centuries in which the Gregorian scheme's lunar correction raises the epact relative to the previous century: 2400-2499, 3600-3699, 5200-5299, 6400-6499, 6800-6899, 8000-8099, and so forth. The unified scheme could be devised to agree with the Gregorian everywhere else.

I have always held that the Gregorian scheme is more accurate than a unified scheme. See my post at the beginning of this section. What I disagree with is the claim that "jitter" from the solar side is being "transmitted" to the lunar side by a unified scheme. The "jitter" on the solar side is the motion of the mean and true equinoxes relative to midnight (beginning of day) March 21 Gregorian. On the lunar side the difference between a fixed date and a fixed annual event is of no consequence. The important difference is between the beginning of a lunation and the mean and true conjunctions. So nothing is "transmitted" from one side to the other. Each side has its own accuracy issues. On the lunar side the mean conjunction, relative to the beginning of day of the first day of the 19-year cycle, moves about within a range of about 34.6 hours in the period from Jan 1, 1710 to Jan 2, 2014, taking its zero point as its position on January 1st, 1710. The biggest shift comes Between January 1st, 1786 and January 1st, 1824. These 38 years consist of two 6939-day cycles back-to-back, causing the mean conjunction relative to start-of-day January 1st to fall back by 1.3768 days over the course of the 38 years. The unified scheme will show the same variation if it has no correction in 1800 and a correction in 1900. But because the mean and true conjunctions shift about from lunation to lunation anyhow, the question remains of whether these long-term variations would be able to be detected by naked-eye observation.--Mockingbird0 (talk) 17:14, 21 March 2009 (UTC)

I do not know what exactly you mean by "mean" and "true" equinoxes, and "mean" and "true" conjunctions. In any case, this is not about the (small) variation of the precise moment of an aequinox compared to a pure cyclic mean progression, or similarly for the lunar phases (which do have a much bigger variation; see e.g. Full moon cycle). The calendar works with whole days (and tithis for the Moon, to be precise), and tries to pinpoint a day that a specific event (aequinox, New, or Full Moon) takes place. The nested Gregorian solar scheme allows the deviation to run up to more than a day on either side. The lunar calendar also suffers from mismatches between ecclesiastic and real Moon, up to 2 calendar days. Moreover, even within the calendar itself periods of 28 or 31 days between New Moons are possible, which do not occur in reality. This is caused by the epact play, and I predict that this will occur more frequently with a "unified" scheme. When I have time I'll write a program to compute all dates and get the statistics. Tom Peters (talk) 15:46, 23 March 2009 (UTC)

What I mean by true equinox and true conjunction should be clear enough. The true conjunction is the day and hour at which the moon has the same ecliptic longitude as the sun. The true vernal equinox is the day and hour at which the sun's equatorial latitude is 0 degrees on its progress northward from the winter to the summer solstice.

By mean vernal equinox I mean a day and hour that occur one mean vernal equinox year (currently 365.2424 days) after the previous mean vernal equinox. The astronomical definition of mean equinox of date is more sophisticated than this, but this simpler definition works for present purposes. So if the difference between midnight (beginning of day) March 21 and the mean equinox in a given year is d, after a year of 365 days the difference will be d + .2424 days (the mean equinox falls back relative to beginning-of-day March 21), and after a year of 366 days the difference will be d - .7576 days (it moves forward relative to beginning-of-day March 21). One could calibrate this mean equinox absolutely by taking a least squares fit of true equinoxes over some period of years.

I intend by "mean conjunction" the same sort of rough computation. Each mean conjunction occurs one synodic lunar month (currently 29.530589 days) after the previous one. After a lunar month of 29 days, the mean conjunction falls back by .5306 days relative to the first day of the incoming moon, compared to where it was on the first day of the previous moon. After a lunar month of 30 days the mean conjunction moves forward relative to the start of the first day of the moon by .4694 days. After a 19-year cycle of 6939 days the mean conjunction relative to the start of the first day of the first moon of the cycle falls back by .6884 days. After a 19-year cycle of 6940 days it moves forward by .3116 days. The mean conjunction can be thought of as corresponding to the molad of the rabbinic calendar, though Otto Neugebauer didn't like the identification of the two.

Though I haven't done so yet on this board, I could as well have defined a "tabular conjunction", a formal conjunction in the Gregorian lunar calendar, that takes place at 6AM on the 30th day of a 30-day lunation, and at 6PM (end-of-day) on the 29th day of a 29-day lunation. One could then compare lunar calendars by measuring the variance in the difference between the tabular conjunction and the mean conjunction.

I hope this helps make my posts clearer.

I consider the 28-day and 31-day intervals between formal Gregorian new moons to be a red herring. One can always eliminate a 31-day lunation by adding an extra day to an earlier 29-day lunation and starting the next lunation a day later if it begins before the extra day has been added. This is the traditional way of handling a bissextile year in the Julian computus: making the moon of February full instead of hollow, and starting the moon of March a day later if necessary. Likewise one can swipe a day from a nearby 30-day month to take care of a 28-day one, so that where one had a 30 and a 28, one has now two 29s. This can always be done without changing the date of Easter.

Though we may indeed discover that by starting from a Julian approximation the Gregorian lunar calendar reduces the variance of some variable relative to some other scheme, I think the reasons for the Gregorian system are more traditional and aesthetic than strictly arithmetical. The Julian calendar is the basis of the western calendar. The Gregorian is a reform of the Julian, so it rests on a Julian substrate. Starting the lunar-side computation from a Julian basis consists with this tradition. Besides, ordinarily when one works by successive approximations, one prefers to have the first term as close to the final answer as possible, and make the correction terms small. This isn't strictly necessary if one can get the same answer by other means. But it's elegant. The Gregorian scheme does this very well. The Julian approximation of 235 synodic months = 6939.75 days is a very good starting point. Even after a few millennia, when the synodic lunar month's value in contemporary days drops down to, say, 29.530565 days, 235 synodic months = 6939.75 days will still be a very good approximation, and a very good starting point for a lunar calendar. And then, if the epacts need adjustment, only the lunar corrections will need to be re-cast for the new value of the synodic month (7 in 2000 years, say, instead of 8 in 2500). The Julian (or near-Julian) underlay can go through "unchanged" in the sense that it will continue to start from the approximation 235 lunations = 6939.75 days.--Mockingbird0 (talk) 19:27, 23 March 2009 (UTC)

I've read from the above the Mockingbird agrees that the Gregorian Computus is more accurate than any uniform system. What he seems to disagree with is that the idea that a computus (especially a uniform correction computus) can transmit the jitter the solar calendar to the lunar calendar. One way of seeing that the solar calendar can transmit jitter to the lunar calendar, is by replacing the Gregorian solar calendar with a less jittery alternative such a the 33-year cycle of 8 leap years, which would do very well with a uniform correction computus. Karl (talk) 10:29, 26 March 2009 (UTC)

152-year artifact

Our discussion above reminded me of a discussion I had a couple of years ago on a different web site. Someone pointed out that Easter dates sometimes repeat after 152 years (Gregorian). I did a quick back-of-the-envelope computation and came up with the theory that this should happen whenever an interval of 152 years contains a century year which is a non-bissextile century year in which there is a lunar correction (such as 2100), and no other century year except it be a bissextile one. In the latter case there will not be a lunar correction in the bissextile century year since there is one in an adjacent non-bissextile century year.

When these conditions are fulfilled, on the solar side, the 152 years contain (365.25x152 - 1) = (55518 - 1) = 55517 days. On the lunar side, there are 2 cycles of 76 years containing 6x6940 + 2x6939 - 1 = 55517 days. Since this is a whole number of weeks (7931), the calendar dates will fall on the same days of the week in the 153rd year as in the 1st. Since the number of years is divisible by 4, the incidence of bissextile years in the 153rd, 154th, etc. years is the same as in the 1st, 2nd, and so on. Since it is a whole number of 19-year cycles and the epacts are unchanged (since the century year was non-bissextile it contained a lunar correction canceled by a solar correction, and any other century year is bissextile), the lunar dates will occur on the same solar calendar dates in the 153rd, 154th year as in the 1st, 2nd, and so on. Easter dates from 152 years previous then (so my back-of-the-envelope computation suggested) continue to repeat until either the year containing the lunar correction falls out of range, or until the epacts change, whichever is sooner. I've only spot-checked the theory with a few examples, though.

I don't know if we need to put anything about this in the article, but I thought it might be of interest to contributors.--Mockingbird0 (talk) 04:16, 22 March 2009 (UTC)

Jean Meeus, "Mathemetical Astronomy Morsels", mentions this period in Ch.60, p.363. Apparently this was first noted in 1951. Meeus does not explain this like you do, but shows that the recurrence lasts for 48, 52, or 100 years (so not 152 years again). The Easters of 1948 to 2047 are repeated from 2100 to 2199, so 100 years on a row. Tom Peters (talk) 15:29, 23 March 2009 (UTC)

The merlyn site agrees, and finds instances with other lengths and shifts. Examples : 356-499 = 1100-1243, 1941-2105 = 8341-8505, 16609-16803 = 23009-23203, 29697-30074 = 83697-84074, etc. The existence of repeats of stretches of many years seems worth mentioning in the Article. 82.163.24.100 (talk) 20:29, 7 May 2009 (UTC)

588 years, 44423-45010 = 188423-189010. 82.163.24.100 (talk) 20:45, 13 May 2009 (UTC)

2103, 1498-3600 = 427098-429200; 2123, 136490-138612 = 562090-564212. 82.163.24.100 (talk) 20:30, 16 May 2009 (UTC)

Thanks for that information.--Mockingbird0 (talk) 18:02, 23 March 2009 (UTC)

There is also an eleven year period in which Easter often occurs on the same date. This year Easter is on 12 April, which is the same date as 11 years ago and 11 years in the future. Also I've heard of a 95-year period. One notable example are the 23 March Easters of 1913 and 2008. Karl (talk) 16:27, 26 March 2009 (UTC)

State of the Article; various comment.

(A0) Since I last visited, the article appears, from History, to have been unwisely hacked about.

(A1) It should I think be divided into two or more articles.

(A2) Computus should refer only to the events leading up to the Explicatio and the Calendar Act, and perhaps when Gregorian was introduced. Julian Easter on the Julian calendar should precede Gregorian. Orthdox Easter - the Gregorian date on which the Julians celebrate - should be an entirely separate section.

(A3) There should be a clear distinction between the method specified by Clavius and simplified methods. For example, there's no copy in the Article of Tabula Paschalis Nova Reformata (Op. Mat. p.34) which is an essential part of Clavius' specified method as faithfully implemented at merlyn.

(A4) Consequences of the definitions should be on another page - intervals, frequencies, algorithms.

(A5) Alterations should be clearly separated. For example, Clavius apparently also presents a longer-cycle form, the Act of 1928, Aleppo 1997 might be put there.

(A6) I think that would make both maintenance and reference easier.

(B1) There should be an external link manifestly showing the Calendar Act as passed; and it should be adjacent to the link to statutelaw.gov.uk (current last link). Note 21 has a good link for it.

(B2) Note 23 is not authoritative; it would be better to present images.

(B3) The term "Book of Common Prayer" is insufficiently definitive; it should be "Book of Common Prayer of the Church of England" or "of whoever"; the C of E is not necessarily the only author of Books of Common Prayer.

(B4) The Clavius reference, to Op. Mat. at Notre Dame, should be replaced or supplemented by one to http://echo.mpiwg-berlin.mpg.de/ECHOdocuView/ECHOzogiLib?mode=imagepath&url=/mpiwg/online/permanent/library/YXK9FE9W/pageimg . Al, who told me of it, wrote "Compared to Notre Dame's copy, this version is better focused, in higher resolution, and in colour, and it also has the benefit of being the original 1603 typesetting, rather than the 1612 reprint in the Opera Mathematica. The page numbering and chapter numbering are different."

(B5) The Weisstein link is still not worthwhile; and the Easter page it links to contains errors.

(B6) The Note 32 link, Butcher at Google, does not show me text; I am not in the USA. Could such links be appropriately annotated? A normal Google Web search has failed to find two phrases which I'm told are in Butcher.

(B7) I have heard of a new, shorter, faster, full-period Gregorian Easter algorithm, details of which it would be premature to give here.

(B8) The merlyn site lists all the intervals in years between consecutive Gregorian Easters of the same date - common date, ordinal date, week-numbering date.

(B9) And "on or after 21st March" should be restored throughout, to be in agreement with all true authority and to better maintain the connection with the conventional Equinox.

(B10) A Julian Easter, Julian Calendar histogram should cover 532 years.

(C) I've noted in the Article that the Calendar Act method does not mention Epacts. Their appearance in its "Table of the Moveable Feasts" is unexplained.

82.163.24.100 (talk) 22:07, 4 May 2009 (UTC)

Butcher is also available on the Internet Archive. Several versions are linked, including the Google copy which you have not been able to access. Google books has a different version for each language, using different two letter codes (en=english, ru=russian, etc.), so Butcher and other books may be accessable elsewhere. For example, I found some Latin books on the Russian server which were not available on the English server. — Joe Kress (talk) 01:34, 6 May 2009 (UTC)

Thanks - a better link [The text version must be unreliable, including such as 2$. 6d. for the price!] [Why did a Protestant Bishop write on Easter after Catholic authority?] [why are writings of a British Bishop in Americana?]. My intentional sectionalisation above was destroyed; I've restored the indication in a different manner. 82.163.24.100 (talk) 13:19, 6 May 2009 (UTC)

You should consider use the Wikipedia codes for lists, see Help:Lists. Numbered lists are easy using the left-justified # sign, but do not separate the items with blank lines. If you insert a blank line between numbered items, the numbering system starts over from 1 again. So two successive individualy numbered lists can be created by separating the two lists by a single blank line. Each item in a simple bullet list uses a left-justified asterisk * rather than a #. You may indent the second numbered list by using two left-justified ##. I don't see an easy way to make alphanumeric lists. — Joe Kress (talk) 00:36, 8 May 2009 (UTC)

New section on algorithms

I propose a new section, after that referring to Gauss, Butcher, et al, saying that faster and more compact algorithms exist, and that some can be found at http://www.merlyn.demon.co.uk/estralgs.txt. 82.163.24.100 (talk) 20:06, 5 May 2009 (UTC)

I don't believe it!

The intro says:

The canonical rule is that Easter day is the first Sunday after the 14th day of the lunar month (the nominal full moon) that falls on or after 21 March (nominally the day of the vernal equinox).

which seems to be a garbling of something correctly intended but badly said. I've always been told that it occurs on the first sunday after the first "full moon" (nominal) after the "vernal equinox" (nominal), nothing about some lunar month. Now, using the definition in the intro, imagine a lunar month that starts 28 days before the vernal equinox, so that the months "falls on" the vernal equinox, but then the day 14 of that month happen at least 7 days (in fact 14 days) before the vernal equinox, and we have an easter that falls before the vernal equinox. This contradicts what the article Paschal Full Moon tells us. David Madore's site says like me, about the Gregorian calendar:

Specifically, Easter is the Sunday strictly following the first ecclesiastical full moon (called the Paschal moon) that occurs on or after March 21.

Jean Meeus in his Astronomical Algorithm says as David Madore and also a Swedish Encyclopedia (Bonniers Compact Lexikon), and foremost the world council of churches, which seems to shorten the “first full moon after the vernal equinox” to “first vernal full moon”. I think the article intro is confusing the Jewish calendar month Nisan 14 with any lunar month whatsoever. But Nisan is not a leap-number 1 of any sequence of 12 lunar months - instead the Jewish lunar year is sometimes 13 months to correct with vernal equinox, and so I believe the intro has been written to contain misunderstandings. ... said: Rursus (bork²) 19:59, 13 May 2009 (UTC)

Or, as an alternative hypothesis, the erroneous sentence is just garbled, by someone intending the "nominal full moon" instead of "14th day of the lunar month", while "that falls on or after 21 March" should refer to "the 14th day" instead of to "the lunar month" as it does by English language rules. ... said: Rursus (bork²) 20:11, 13 May 2009 (UTC)

Firstly, it is wrong to say "The canonical rule" without indicating which one. There are three primarily : "Catholic" (Gregory/Clavius), "British" (Calendar Act), "Orthodox" (Nicaea, more or less) ; the first two give the same days, the third differs. The wordings and algorithms differ significantly in all three. It seems possible that when other Protestant countries with State churches changed to Gregorian they may have legislated in different terms, generating different but equivalent Rules. RSVP.

Secondly, one should distinguish between the arguments leading up to the choices of the rules (in which the 14th day features), and the rules themselves. The "British" rule says nothing of the 14th day, the lunar month, or the new moon. Its principles are as quoted in the "British Calendar Act and Book of Common Prayer" section of the Article; its details are in its Tables and their instructions.

Do not attempt to obtain rules from any other than the original sources; even Gauss got it wrong, and one should not expect anyone else to be more reliable. And the "Catholic" rule, until His Holiness approves a translation, can only be obtained by one who can read Clavius's Latin with ease.

82.163.24.100 (talk) 21:06, 13 May 2009 (UTC)

The rule as stated is exactly correct for Gregorian Easter. "Full Moon" is too vague. The computus uses the 14th day of the lunar month, that starts with the first day according to the epacts and the calendarium that Clavius lists.

Tom Peters (talk) 14:08, 4 August 2009 (UTC)

Two major two-way distinctions need to be clearly made throughout the Article. One is between Roman Catholic and British/Anglican (Orthodox is a straightforward modification). The other is between the arguments leading up to the choice of the dates for Easter, and the actual rule as decreed algorithmically and its consequences.

The Rule as stated at the top of this discussion section is only correct for Catholic Easter; the Anglican Rule says nothing of 14 days, since it uses a computed Full Moon (no doubt 14 days after the Catholic New Moon). Unless there is complete clarity as regards which Rule is intended, there will be endless confusion between the contrasting "new" and "full" moon methods.

Since the article is rather long, it might well be split into three or four parts. One would be all leading up to and including the Catholic definition of Easter, i.e. to the end of Clavius. Another would be all leading up to and including the British/Anglican definition of Easter, i.e. to the Calendar Act as passed. A third could be, after asserting the two Rules to be perpetually equivalent (which requires extrapolation of at least one of them), all the "numerical" consequences. And a possible fourth (if not a section of the third) would be on algorithms implementing the calculation, traceable to one or other authority either by direct reasoning or by comparison of results.

82.163.24.100 (talk) 12:17, 16 August 2009 (UTC)

Ah. What I have written has been for the Gregorian calendar, and based on the original texts. Call that "catholic" if you like, but AFAIK it has been adopted "as is" by most protestant churches (and many non-Christian countries) as well - in Britain in 1752 or whereabouts if my memory does not fail me. You claim that the Anglican church uses a different computus or at least different canonical texts and terminology. Is that true, I wouldn´t know? Can you provide proper quotations and references? Moreover, does the Anglican church ever celebrate Easter at another date than the Gregorian one?

If the British Calendar Act talks about Full Moon instead of 14th day, but otherwise the result of the computation is fully equivalent to the original Gregorian one, then I consider that just a national formulation of the same thing: maybe mention it in something as little as a footnote or a sub-paragraph at most, but not a reason to split up the article. There will be a differently formulated law for every national church and possibly for any independent church. Spawning pages for each of them is A Bad Idea IMNSHO. I could support splitting the pages in one for history & old Julian computus, and one for the Gregorian computus and its intricacies.

Let me know, Tom Peters (talk) 12:30, 1 September 2009 (UTC)

I agree with Tom Peters regarding the computi: As long as the Catholic and Anglican computi result in the same date for Easter Sunday, regardless of the terminology used, then it is the same computus and should be described as such with the same canonical rule. I disagree that the article should be split. It is not overly long compared to many Wikipedia articles. — Joe Kress (talk) 23:51, 14 September 2009 (UTC)

Epact page

Is there any good reason why the page does not link to the Epact page? 82.163.24.100 (talk) 20:18, 20 May 2009 (UTC)

Epact is linked several times in the text of the article. According to repeated links, only the first use of an item should be linked, but if it is used far away from its first use, it may be linked again. According to See also, a term which is linked within the body of an article should not be linked within See also. — Joe Kress (talk) 00:44, 5 August 2009 (UTC)

Granted. I think that I did mean that it should be in the "See also" section. IMHO, links in the text suffice for general explanations, but pages closely associated with the main topic or title should be in "See also"; such links are much easier to see. 82.163.24.100 (talk) 21:00, 24 November 2009 (UTC)

Gregorian algorithm

"A New York correspondent" submitted this algorithm for determining the Gregorian Easter to the journal Nature in 1876.[34][35] It has been reprinted many times, in 1877 by Samuel Butcher in The Ecclesiastical Calendar,[36]:225 in 1922 by H. Spencer Jones in General Astronomy,[37] in 1977 by the Journal of the British Astronomical Association,[38] in 1977 by The Old Farmer's Almanac, in 1988 by Peter Duffett-Smith in Practical Astronomy With Your Calculator, and in 1991 by Jean Meeus in Astronomical Algorithms.[39] The Gregorian Easter has been used since 1583 by the Roman Catholic Church and was adopted by most Protestant churches between 1753 and 1845. German Protestant states used an astronomical Easter based on the Rudolphine Tables of Johannes Kepler between 1700 and 1774, while Sweden used it from 1739 to 1844. This astronomical Easter was one week before the Gregorian Easter in 1724, 1744, 1778, 1798, etc

What is exactly the "astronomical Easter [that] was one week before the Gregorian Easter in 1724, 1744, 1778, 1798, etc"? the astronomical Easter calculated by the following algorithm? The astronomical Easter based on the Rudolphine Tables of Johannes Kepler? Or they are the same?

The passage should be better:

"A New York correspondent" submitted this algorithm for determining the Gregorian Easter to the journal Nature in 1876.[34][35] It has been reprinted many times, in 1877 by Samuel Butcher in The Ecclesiastical Calendar,[36]:225 in 1922 by H. Spencer Jones in General Astronomy,[37] in 1977 by the Journal of the British Astronomical Association,[38] in 1977 by The Old Farmer's Almanac, in 1988 by Peter Duffett-Smith in Practical Astronomy With Your Calculator, and in 1991 by Jean Meeus in Astronomical Algorithms.[39]This astronomical Easter was one week before the Gregorian Easter in 1724, 1744, 1778, 1798, etc

The history of the Gregorian Easter should be in another part of the text. For the algorithm is irrelevant.--194.65.151.101 (talk) 17:01, 27 January 2010 (UTC)

Agreed. I'm moving all history to the History section. — Joe Kress (talk) 18:52, 27 January 2010 (UTC)

I've got a question. The Gregorian definition states that "Easter Day is the first Sunday after the 14th day of the calendrical lunar month which begins on, or next after March 8." When does this March 8 start? It depends on which timezone you're in, so there should be a precision of this term, whether it's the Jerusalem meridian, the Rome meridian, plain UTC, or whatever. leifbk (talk) 13:28, 12 April 2010 (UTC)

Easter is calculated for the local time of each place where it is observed. Jc3s5h (talk) 14:56, 12 April 2010 (UTC)

Citation needed. I don't believe it. leifbk (talk) 15:22, 12 April 2010 (UTC)

Well, you could just look on the internet for the times the various churches around celebrated the recent Easter and observe they won't all fit into any single 24 hour period. Also, the article does give a reference: Clavius, Christopher (1603): Romani calendarij à Gregorio XIII. P. M. restituti explicatio. In the fifth volume of Opera Mathematica (1612)

If what you're saying is correct, I think that the article might be improved with a section on how the various churches, eg. the Roman Catholic church, the CoE, North European Lutherans, American Protestants, etc. each define their specific Easter and what meridian they base their computations on. leifbk (talk) 16:09, 12 April 2010 (UTC)

I didn't mean each denomination (Roman Catholic, Greek Orthodox, Episcopal, etc.) I mean each church building. This past Easter, I understand both the western and Orthodox denominations all observed Easter the same "day". But if you look at the schedules for all the church buildings in the world, I expect you will find they all held their observances sometime between Saturday evening, local time, through Sunday afternoon, local time. Since all these buildings were in different time zones, the time period covered, measured in Universal Time would be about 48 hours.

The calculations use a calendrical lunar month, which only consider whole days, not time of day. Furthermore, the calendrical lunar months are rule-based, not based on actual observations of the moon (although the rules are intended to be an adequate model of the moon over the long term). Jc3s5h (talk) 16:44, 12 April 2010 (UTC)

We're talking of two different things then. My concern is that "March 8", or rather, the 14th day after, starts at different times around the world. If the new Moon appears at Rome March 22 00:10 local time, the same moment in Jerusalem will be March 23 23:10 local time, and thus there will be a week's difference between the Easters computed in Jerusalem and Rome. There must be a common meridian somewhere that the Easter time is computed from. leifbk (talk) 17:02, 12 April 2010 (UTC)

You are thinking astronomically. However, the church rejected astronomical calculations in favor of simple tables, which were known to deviate from their corresponding astronomical calculations by as much as two days. Kepler summarized this argument in his famous quote, "Easter is a feast, not a planet." Easter was calculated astronomically during the eighteenth century by Germany and the Nordic countries before they reverted to the tables. — Joe Kress (talk) 18:30, 12 April 2010 (UTC)

There is no meridian. If the calculation were done astronomically, one would note there was a new moon on March 15, 2010, at 21:01 Universal Coordinated Time, which would be the same day at 22:01 in Rome or the next day, March 16, at 00:01 in Jerusalem. But this does not matter. Using the tables in the article, or equivalent tables that can be found in Blackburn and Holford-Strevens pp. 821-828, one finds that the tablular new moon in 2010 is March 17. The tabular full moon is 13 days later, March 30. The Sunday after March 30 is April 4. The tables are concerned only with calendar dates, not with actual astronomical events. Jc3s5h (talk) 18:57, 12 April 2010 (UTC)

I get the point. However, the definition does not say that "Easter day is the day computed from such-and-such a table". The definition involves both a date and a new moon, and the mention of a date in this context is meaningless without an implied meridian. The Venerable Bede may not have known that the Earth is round, nor may he have been aware of the concept of different meridians. But the idea should certainly have occurred to somebody since then. leifbk (talk) 07:32, 13 April 2010 (UTC)

leifbk wrote "Easter day is the day computed from such-and-such a table". In effect, that is indeed the definition. If you read Inter Gravissimas you will see that while Gregory's bull contains explicit rules for computing leap years, it refers the matter of the lunar calendar and calculation of Easter to the explanation provided by Gregory's advisers:

Moreover, so that the fourteenth day of the Paschal moon is given with precision and that the age of the moon is presented with precision to the faithful in accordance with the antique use of the Church, to take note of it each day with the reading of martyrology , we order that once the Golden Number is withdrawn from the calendar, one substitutes the cycle of the epacts for it which, thanks to its very precise rules mentioned above for the Golden Number, makes so that the new moon and the fourteenth day of the Paschal moon always hold their place. And this is seen clearly in the explanation of our calendar, where are also presented Paschal tables in conformity with the ancient habits of the Church and which make it possible to find more surely and more easily the sacred date of the Easter.

I take this to mean the work by Clavius. Although the version cited in the article was published in 1603, I believe some earlier, possibly incomplete, form existed when the bull was published. Clavius presented his results in the form of tables. Jc3s5h (talk) 16:05, 13 April 2010 (UTC)

Three possibilities exist for "the explanation of our calendar" mentioned in paragraph 10 of "Inter gravissimas" (reprinted by Clavius as pages 13–15 of Romani calendarii a Gregorio XIII P. M. restituti explicatio (hover cursor over "Go to Page", select "Work: Roman Calendar of Gregory XIII", and select the appropriate page): (1) The six canons attached to the bull and reprinted by Clavius on pages 16–31. (2) A book by Antonio Giglio, the brother of Luigi Giglio, which was finally finished after several years in 1585 but never published. Eventually, the only surviving member of the papal commission, Christopher Clavius, published the explicatio in 1603, which was reprinted in 1612 in volume 5 of Opera Mathematica of Christopher Clavius beginning on page 53. I do not know of any translation. (3) "A small book" mentioned in paragraph 5 of the bull entitled Compendium novae rationis restituendi calendarium (Latin and French - the French can be automatically translated by Google toolbar) printed and distributed to several universities and princes in 1577. It was reprinted by Clavius on pages 3–12. A good discussion is in The Gregorian reform of the calendar on pages 177–188 and 201–239. The canons are mentioned on page 220 via "[The bull] was printed together with the new perpetual calendar and the essentials of the new system". Giglio's book is mentioned on page 225. — Joe Kress (talk) 00:30, 14 April 2010 (UTC)

Week of Unleavened Bread

Why does this article consistenly refer to "the Week of Unleavened Bread"? The holiday is called Passover. Tad Lincoln (talk) 06:47, 29 January 2010 (UTC)

Leviticus 23:6, I should think. --Kay Dekker (talk) 16:17, 1 April 2010 (UTC)

Lichtenberg formula (released 1997)

k=floor(year/100)

m=15+floor((3k+3)/4)-floor((8k+13)/25)

s=2-floor((3k+3)/4)

a=year mod 19

d=(19a+m) mod 30

r=floor((d+floor(a/11)/29) (*)

og=21+d-r

sz=7-(year+floor(year/4)+s) mod 7

oe=7-(og-sz) mod 7

os=sz+oe

k,m,a,d as Gauss

r="Korrekturgröße" ~ correction {replace the d/e depent rules by Gauss}

og="Ostergrenze" ~ limit of eastern

sz="Sonntagszahl" ~ number of sunday

oe="Osterentfernung" ~ difference between limit and sunday

os="Ostersonntag" ~ easter sunday {values 32+ -> subtract 31 for day of april}

(*) shorter version of r calculation by unknown person - Lichtenberg formula is longer —Preceding unsigned comment added by 79.199.102.44 (talk) 17:23, 3 May 2010 (UTC)

Modern calendar

I quote from the introduction: "... the Julian calendar has a leap year every four years and the modern calendar omits seven of these leap days every 900 years". According to the rules stated in Gregorian calendar three of these leap days are dropped every 400 years: "Every year that is exactly divisible by four is a leap year, except for years that are exactly divisible by 100; the centurial years that are exactly divisible by 400 are still leap years." Rgds --Boobarkee (talk) 22:21, 1 June 2010 (UTC)

The change was introduced by sockpuppets of the blocked User:Vote (X) for Change, whether for the purpose of creating confusion between the Gregorian calendar and the Revised Julian calendar, or just to make trouble, I don't know. I have revert his edits back to January. I hope anyone who made any worthwhile edits since then, if they are applicable to a good-faith version of the article, will restore them. Jc3s5h (talk) 23:36, 1 June 2010 (UTC)

Hutchins' rule (2005)

Divide	By	Quotient	Remainder
y	19	-	a
y	100	b	-
y	4	c	-
3b + 3	4	d	-
8b + 13	25	e	-
19a + d - e + 15	30	-	f
a + 11f	319	g	-
f - g + y + c - d + 2	7	-	h
f - g - h + 120	31	j	k

Easter Sunday in Gregorian year y falls on day k+1 of month j.

Cell 1 in the "h" line needs also a "-d" term". 94.30.84.71 (talk) 11:28, 27 April 2011 (UTC)

Yes. Apologies for the typo. The table is now correct.

Note: In the table, 'quotient' and 'remainder' refer to integer quotient and integer remainder respectively, ie where p is the dividend and q is the divisor, quotient = floor(p/q) and remainder = p mod q.

—Preceding unsigned comment added by 86.16.187.179 (talk) 09:36, 10 February 2011 (UTC)

Where did you find this algorithm? — Joe Kress (talk) 04:44, 11 February 2011 (UTC)

The algorithm was published in Datafile v24n1 (January/February 2005), see http://www.hpcc.org/datafile/datafilev24.html.

I have tested Hutchins' Rule against against the Meeus, Butcher and other methods for the years 0 to 1,000,000 and it agrees completely. Could it be the smallest and simplest Gregorian Easter algorithm? — Preceding unsigned comment added by 94.116.119.215 (talk • contribs) 04:50, 11 February 2011

No. "Simplest" is a matter of opinion, but Al Petrofsky has produced remarkably short ones. See http://www.merlyn.demon.co.uk.estrdate.htm and linked pages. 94.30.84.71 (talk) 17:43, 25 April 2011 (UTC)

A proper test should cover 5700000 years (the repeat interval) and verify that it does repeat. As a quibble, years before 1583 or 1753 should not be counted, as the definitions of Gregorian Easter are not valid for those years (use years 5700000 later). 94.30.84.71 (talk) 17:43, 25 April 2011 (UTC)

Commonly, in testing, the Day-of-March is useful. Evidently that can be obtained by simplifying Hutchins' last line. That would be worth adding as a "footnote to Hutchins". 94.30.84.71 (talk) 17:43, 25 April 2011 (UTC)

So your source is Tony Hutchins, "Gregorian Easter on the HP-12C", V24N1 HPCC Datafile (Jan/Feb 2005) P36. — Joe Kress (talk) 08:26, 12 February 2011 (UTC)

Is there a current URL for the Hutchins code publication itself? The one above seems not to lead to it. My tests of the code above suggest a possible typo somewhere, granted possibly by me. 94.30.84.71 (talk) 18:29, 25 April 2011 (UTC)

Evaluating Easter Sunday by hand for years 2008 and 2011 using the above table gives the same incorrect results as my code translation did. The dates fail to meet the essential requirement of being Sundays, one being six days late and the other a day early. But the code results seem always in the right ball-park. 94.30.84.71 (talk) 20:32, 26 April 2011 (UTC)

I thought that the Table must have a typo, and looking at the form of the errors has prompted me to try a "d" term in the dividend that leads to the value of "h". Changing "f - g + y + c + 2" to "f - g + y + c + 2 - d" gives results correct for years 0 up to at least 6E8. Presumably 86.16.187.179 tested something which was transcribed into Wikipedia, but did not test the transcription itself. See http://www.merlyn.demon.co.uk/estralgs.txt for the code. 94.30.84.71 (talk) 11:40, 27 April 2011 (UTC)

Yes, the line should indeed be "f - g + y + c - d + 2". I made a typographical error when entering the table into Wiki. Apologies. —Preceding unsigned comment added by 86.16.187.179 (talk) 05:36, 28 April 2011 (UTC)

One should always test algorithms in the form in which they actually will leave or have left the publisher, since it is the final stage which is the most fertile of typos. Hutchins' rule is reasonably, but not exceptionally, fast. You should put four tildes after each of your discussion contributions, as it says below the editing box. 94.30.84.71 (talk) 14:59, 30 April 2011 (UTC)

March 21st

In the Gregorian Rule, the Paschal Full Moon has always been that on or after March 21st. In the Julian Rule as given in pre-1752 C of E Prayer Books, the same is true. It appears to me that, at least since AD 1000, that has been the case (Harvey (Drabble)) - and probably for much longer. I suggest that, if possible, the Article should say explicitly when, or by when, the 21st of March (rather than the true Equinox) was generally taken as the reference point. 94.30.84.71 (talk) 17:22, 25 April 2011 (UTC)

The equinox became important to Easter in the third century as mentioned in the first paragraph of the history section of the article. Initially, no date was specified. Anatolius of Alexandria was the first to specify a date, variously reported as March 22, March 18, etc. At the same time, a Roman computist specified March 18 or 25. The Alexandrian computus, developed during the first decade of the fourth century, was the first to specify March 21. This became the medieval rule and the Gregorian rule. Rome adopted March 21 about 342. Thereafter, both Rome and Alexandria placed the equinox on March 21 even though their Easter tables disagreed. Neither ever used the true equinox. As mentioned in the last paragraph of the history via "astronomical Easter", the true equinox was used at various times between 1700 and 1844 by Germany and the Nordic countries. — Joe Kress (talk) 23:59, 26 April 2011 (UTC)

I'm suggesting that the ARTICLE needs the information, but not the details - something like "March 21st has been generally used as the earliest Full Moon date ever since the first half of the fourth century". Obviously Gregory needed to inherit that date in order to get the same Easter date range from similar rules. 94.30.84.71 (talk) 11:06, 27 April 2011 (UTC)

The details complicate the issue. Before 342, Rome allowed both the equinox and the paschal full moon to occur after March 18 provided that Easter Sunday itself occurred on or after March 25 (IIRC). The Alexandrian rules after 311 precisely agree with the modern rule. — Joe Kress (talk) 00:53, 28 April 2011 (UTC)

"The details complicate ..." - indeed. That is precisely why I suggested that the article should indicate when March 21st became generally accepted for the determination of Julian Easter. From what you say, that appears to be 342 (though you don't say here whether the decision applied in principle to Easter 342 itself (April 11th)). An article should not bury solid facts within a morass of detail. 94.30.84.71 (talk) 10:42, 8 May 2011 (UTC)

I would like to consult my sources again, principally Charles W. Jones, "Development of the Latin ecclesiastical calendar", pages 1–122 in Bedae Opera de Temporibus (1943).

We do not know exactly which year Rome adopted 21 March as the date of the equinox, but it was within a couple of years of 342. The equinox was only one component of the tables used to calculate Easter. During the fourth century, Rome used an 84-year cycle while Alexandria used a 19-year cycle to calculate Easter. Even though they used different tables which often indicated different dates for Easter, they consulted each other and agreed on a common date for Easter, which was announced by Bishop Athanasius of Alexandria in his festal letters [4]. Before about 390 when Bishop Theophilus of Alexandria published his Easter tables, Alexandria usually accepted Rome's preferred date, but therafter Rome was forced to adopt Alexandria's date. Part of the period that Athanasius was bishop of Alexandria (328–373) he was in exil in Rome (339–346) due to the Arian controversy, so he may not have had a choice. Only once, in 501, did Alexandria and Rome celebrate Easter on different dates, as far as we know. — Joe Kress (talk) 03:48, 9 May 2011 (UTC)

So far I've only compared Alexandria and Rome. An 84 (14)-year cycle was used throughout the British Isles (Insular Easter) until its last general use in northern Britain in 664 (last holdout in 931). It did not use a single date for the equinox, but an equinox period, 21–25 March inclusive. That is, the Easter full moon (luna 14) could occur anywhere within that period, but if the first Sunday after a full moon was before the first allowable date for Easter Sunday, 26 March, then Easter would be delayed a month. See note 85 on page 57 of Immo Warntjes, "The Munich Computus and the 84 (14)-year Easter reckoning", Proceedings of the Royal Irish Academy 107C (2007) 31–85 (available online at larger libraries). 84 (14) means that a saltus lunae occurred every 14 years for six saltus lunae within this specific 84-year cycle. — Joe Kress (talk) 17:43, 10 May 2011 (UTC)

In Software

The Excel functions in the Article look too simple to be full range. The actual range needs to be given. 94.30.84.71 (talk) 11:19, 26 April 2011 (UTC)

Definitions

There are exactly TWO definitions of the date of Gregorian Easter Sunday. The Gregorian one is (for 1583 to 4999) in Clavius' Six Canons, and for perpetuity Clavius' Opera Mathematica Tomus V, both given authority by Gregory XIII's Bull Inter Gravissimas. The other one is the British Calendar Act of 1751 as enacted, of which the relevant parts are by law printed in the Church of England Book of Common Prayer. Both are arithmetic definitions, giving the Sunday AFTER a computed full moon date ON OR AFTER March 21st (they are very different in form, but agree completely in result).

The word "definition", or similar, should not be used to refer to arguments involving the 14th day of the moon, the Vernal Equinox, etc., etc. - those are merely considerations which led to the choice of the Papal definition (the British definition was chosen to give the same answers, but in an operationally-simpler manner and in English). Ideally, the Pope and the Archbishop (or the ISO 8601 people) would issue a superseding agreed executable form (such as JavaScript, which most computers can execute).

To facilitate this distinction, the Article and the Discussion need to be divided into three parts (by major headings, or on separate pages), to be read in turn : (1) arguments leading to the definitions, (2) the definitions themselves, (3) the consequences of the definitions. Each of those could have two parts, for Julian and Gregorian (and something about Easter on the Greek secular calendar with its different leap years).

There are other interesting consequences, apart from the histograms; the number of weeks between adjacent Easters, the minimum and maximum number of years between Easters of the same date, the spans of years containing all 35 possible dates, ... .

94.30.84.71 (talk) 21:25, 26 April 2011 (UTC)

This is a tempest in a teapot. Because both the rule in Inter gravissimas and the rule in the Calendar Act of 1750 specify the same dates for Easter, I do not support any further discussion beyond that already in the article at British Calendar Act and Book of Common Prayer. See my extensive discussion with 82.163.24.100 in the archive. — Joe Kress (talk) 23:59, 26 April 2011 (UTC)

P.S. That link should be to Calendar_%28New_Style%29_Act_1750 rather than to a redirector : Calendar (New_Style) Act 1750. Giving a false form is misleading. 94.30.84.71 (talk) 10:04, 27 April 2011 (UTC)

You have missed the main point entirely.

And the Calendar Act was not OF 1750, since the normal interpretation of that is that the Act was enacted in AD 1750. It was enacted in AD 1751 (OS/NS/Julian/Gregorian).

The TITLE of the Act is "Calendar (New Style) Act 1750", in which 1750 is the Regnal Year, and 1750 should only be used as part of the Title. Don't get confused by the Act having been passed in the (Annual) Parliament which started in January 1750, since that was an OS date with the year number changing in March. 94.30.84.71 (talk) 09:59, 27 April 2011 (UTC)

The original title of the Act was An act for regulating the commencement of the year; and for correcting the calendar now in use. The short title assigned in 1896 to the modern version is Calendar (New Style) Act 1750. A committee chose that title. Large sections of the original act have been removed so the modern version is "gutted". — Joe Kress (talk) 17:19, 28 April 2011 (UTC)

Therefore, one must be clear, both here and on the page about the Act, as to which parts refer to the original Act (I use "as enacted") and which to the modern version; and also as to which authorities are used for those wordings.

The best representation on the Web for the Act as enacted seems to be in Statutes at Large 1765 at Google Books.

In the version of the current Act presented on the Web by HMG at http://www.legislation.gov.uk/apgb/Geo2/24/23, the discursive Section 3 (in HTML) dealing with Easter is virtually the same as in 1765 ; and the text in the Easter Tables part (images), after the foot of Section 6, is corrupt.

The current Church of England Book of Common Prayer contains words and tables which would satisfy King George II's legislators if they were still alive. 94.30.84.71 (talk) 12:13, 8 May 2011 (UTC)

"As enacted" is not precise enough because all later amendments were also enacted. It might be best to refer to the original version "as enacted 1750/51" to indicate both the 25 March and 1 January beginnings of the year. — Joe Kress (talk) 03:55, 9 May 2011 (UTC)

Another printing of the original act with all tables and the first amendment is in Archibald John Stephens, The statutes relating to the ecclesiastical and eleemosynary institutions of England, Wales, Ireland, India, and the colonies, Volume 1 (1845) 828–847. — Joe Kress (talk) 21:05, 13 May 2011 (UTC)

Miscellaneous

Why is it not something more simple such as the 2nd sunday in April -N7I2S5 (talk) 20:48, 1 July 2011 (UTC)

The UK "Easter Act 1928" permitted, if agreeed world-wide, a similar rule. But, now that we have ISO 8601 week dates, it would be better to fix exactly one date, the same each year, yyyy-W15-7. 94.30.84.71 (talk) 11:30, 17 November 2011 (UTC)

The 19-year Cycle is Metonic

The 19-year cycle is the Metonic cycle or Enneadecaeteris; discussions of Easter commonly use "Metonic". That link should be added somewhere within the text of the article, perhapa as "the 19-year Metonic cycle". One respected, but not expert, site, refers instead to the longer Callipic cycle. 94.30.84.71 (talk) 11:53, 17 November 2011 (UTC)

Modulo operation when dividend is negative

There are different definitions for the modulo operation which give different results when one of the operands is negative, and different programming languages implement it differently. For example, if year was -1 (i.e. 2 BC. I know Easter didn't exist back then, but algorithms should have a consistent definition regardless), then "Y mod 19" would return -1 in C and Java, while it would return 18 in Python and Ruby. So my questions are:

Do the different modulo operations produce different results?
If so, which one would give the theoretically correct date ("theoretical", if the year is before Easter existed)?
If so, should the correct one be specified in the article?

-- 208.80.119.68 (talk) 02:52, 16 December 2011 (UTC)

See Modular arithmetic. Laguna CA (talk) 19:40, 24 March 2012 (UTC)

Computus

"Computus" is an overly obscure, incorrect word. See http://www.merriam-webster.com/dictionary/computus ; Webster's Third New International Dictionary (Unabridged) makes it clear that "computus" refers to the tables, such as the Paschal new moon table, not the computation itself. I learned the computation of Easter's date long ago and occasionally reminded myself of the rules; this is the first time I recall seeing the word "computus". I suggest the mathematical part of Easter and this article be combined in "Easter, computation of date" or some similar name which can be found more easily. The use of "computus" seems to be obscurity for its own sake. Laguna CA (talk) 19:39, 24 March 2012 (UTC)

"Computus" is correct according to English-language reliable sources as required by WP:TITLE. For example, an entire 28-page chapter in The Oxford Companion to the year, itself a 1,000-page tome, is entitled Computus (801–828). It defines: "Computus is calculation, especially of the calendar, in particular but not exclusively for the determination of Easter." So computus does refer to the computation itself. Furthermore, this article's title has been stable for ten years without objection. About 5,500 articles and books that include "computus" in the title can be found on Google scholar. Needless to say, I object to any change in this article's title. — Joe Kress (talk) 04:11, 25 March 2012 (UTC)

Additional reference: http://www.onelook.com/?w=computus -- "Computus" does not appear to be in current, common usage in English now, or in the last century. Note that One Look cross-references many authoritative sources, including English (strict sense) sources; compare http://www.onelook.com/?w=Easter to see how many sources do not consider "computus" current English usage; it may be proper Latin usage, but this is en.wikipedia.org. From WP:TITLE "Article titles should be recognizable to readers, [fail] unambiguous, [???] and consistent with usage in reliable English-language sources. [fail]" IMHO, of course. Laguna CA (talk) 08:22, 26 March 2012 (UTC)

Onelook ignores "unabridged" dictionaries, two of which include "computus", the Oxford English Dictionary and Webster's Third New International Dictionary. The OED states that the word entered the English language in the mid 19th century. Nevertheless, no "abridged" dictionary, like collegiate dictionaries, include it. However, WP:TITLE allows "Recognizability – Titles are names or descriptions of the topic that are recognizable to someone familiar with (though not necessarily expert in) the topic." This allows a title that is unrecognizable to the vast majority of readers who are not familiar with the topic, which appears to be your concern. Consider the many technical articles in Wikipedia that have titles not familiar to the vast majority of readers. — Joe Kress (talk) 23:04, 28 March 2012 (UTC)

Onelook includes at least "Webster's Revised Unabridged, 1913 Edition" (see Easter definition, above). My top comment about computus referring to tables was sourced from W3NID(U). — Preceding unsigned comment added by Laguna CA (talk • contribs) 03:38, 31 March 2012 (UTC)

rename proposal

This article should be renamed to Easter computus with computus as a redirect there'. While the subject is certainly notable, the references do not support this the current title. The word computus is used generically in other astronomical calculations as well. Any objections?--RadioFan (talk) 16:27, 8 April 2012 (UTC)

Oppose. Computus is an ancient name for this computation. I have never seen the term used generically for other astronomical calculations and don't accept this claim without citations. Jc3s5h (talk) 17:34, 8 April 2012 (UTC)

Meeus Julian algorithm examples for 2008 and 2009

2008 and 2009 examples in the Julian Algorithm section are not correct considering the Gregorian calendar. They are not even on Sundays. — Preceding unsigned comment added by Claytom (talk • contribs) 02:02, 6 May 2012 (UTC)

So what? Jc3s5h (talk) 14:00, 6 May 2012 (UTC)

Merge with Easter Controversy article?

I think that this article's history section should be fairly brief. Merger would expand it to unwieldy lengths. I vote for keeping the Easter Controversy article separate, and removing the tag proposing the merger.--Mockingbird0 (talk) 04:29, 8 July 2011 (UTC)

Better in my view not to merge them, and summarize Easter Controversy here. Tom Harrison ^Talk 11:31, 26 March 2012 (UTC)

This article is basically mathematical. It should indicate clearly for which flavor of Easter a particular formula is applicable, but the reasons historical or otherwise why there are conflicting formulas belong to Easter Controversy. (DPL 6 April 2012) — Preceding unsigned comment added by 41.5.23.67 (talk) 09:29, 6 April 2012 (UTC)

I've removed the tag proposing Easter controversy merge as per consensus. What about Ecclesiastical full moon? --- DoctorKubla (talk) 14:51, 4 May 2012 (UTC)

I'll take that as a no, particularly since you didn't give any reasons for. — LlywelynII 01:49, 4 February 2013 (UTC)

Addition to table, claiming regular pattern

Is the table addition in this edit by User:Q5968661 comprehensible and sufficiently connected to the remainder of the article that readers will understand why it is included? Is this table useful to readers, and therefore merit inclusion? Is "regular pattern" a correct and useful titile for that portion of the table? Jc3s5h (talk) 19:54, 3 March 2013 (UTC)

Simplify or explain. The table is confusing, but if additional explanation were offered it would make more sense. Alternatively, the article could include a simplified walkthrough of what all of those dates mean for a single year. Andrew³²⁷ 19:21, 5 March 2013 (UTC)
This makes sense. Perhaps split into two tables preceded by explanations? Comparisons should still be easy enough for reader, if tables are in proximity to each other. — daranz [ t ] 20:35, 14 March 2013 (UTC)

Notes to Excel formulas

Both formulas on article page work correct only from 1900 to 2203 in the 1900 date system and provide only the Gregorian Easter Sunday dates

Calculate Gregorian Easter date

The actual state is a reduction to <100 characters with full functionality from 1900 to 9999

date system	Excel formula
1900	=FLOOR((5&-A2)+97%MOD(18.998MOD(A2+8/9,19)+INT(68%INT(A2%)-INT(A2%/4)-5/9),30),7)-34 or longer =FLOOR((4&-A2)-5+97%MOD(18,998MOD(A2+8/9;19)+INT(68%INT(A2%)-INT(A2%/4)-5/9);30);7)+1
1904	=FLOOR((4&-A2)-6+97%MOD(18,998MOD(A2+8/9;19)+INT(68%*INT(A2%)-INT(A2%/4)-5/9);30);7)+2
190x	=FLOOR((4&-A2)-DAY(5)+97%MOD(18,998MOD(A2+8/9;19)+INT(68%*INT(A2%)-INT(A2%/4)-5/9);30);7)+DAY(1)

Excel community: http://www.online-excel.de/fom/fo_read.php?f=1&h=58861&bzh=72926&ao=1#a123x

The Gregorian Easter period began 1583 and not 1900 - and that's why it must be enhanced (1900 date system).

=A2&TEXT(FLOOR((5&-A2-(A2%<19)/5%%)+97%*MOD(18.998*MOD(A2+8/9,19)+INT(68%*INT(A2%)-INT(A2%/4)-5/9),30),7)-34,"-MM-DD")

Notice: This formula provides the same results like Gauss from 0001-9999, you could enhance it again with starting =TEXT(A2,"0000")&...
17:42, 23 June 2013 (UTC)

Calculate Julian Easter date

As date: The following works from years 1900 to 9999 (1900 date system)

=(A2&TEXT(FLOOR((5&-1908-MOD(A2,28))+MOD(19*MOD(A2,19)-14,30),7),"-M-D"))-35

As text: The following works from years 0001 to 9999 (1900 date system)

=TEXT(A2,"0000")&TEXT(FLOOR((4&-1908-MOD(A2,28))+MOD(19*MOD(A2,19)-15,30)-4,7),"-MM-DD")

Combination of Julian/Gregorian Easter date

As YYYY-MM-DD text: The following works from years 0001 to 9999 and any breakpoint from 1 to 9999 (here it is 1583), taking 1900 date system

=TEXT(A2,"0000")&TEXT(IF(A2<1583,FLOOR((4&-1908-MOD(A2,28))+MOD(19*MOD(A2,19)-15,30)-4,7),FLOOR((5&-A2-(A2%<19)/5%%)+97%*MOD(18.998*MOD(A2+8/9,19)+INT(68%*INT(A2%)-INT(A2%/4)-5/9),30),7)-34),"-MM-DD")

Frank Schneider, 08:55, 5 November 2013 (UTC)--

First two Excel formulas of main page have a lot of restrictions - is it possible to replace them? 22:10, 30 September 2013 (UTC) — Preceding unsigned comment added by 85.199.76.112 (talk)

I added the universal formula. Frank Schneider, 08:55, 5 November 2013 (UTC)--

Mental arithmetic

>> step1 needs to be bracketed to clarify it is the +29 that is conditi0onal on y mod 19 = 5 or 16. >> and some clarification is also needed on what mod 30 really means here, since mathematically it should be in the range 0-29 or (sometimes) -14..15 but in the samples 55 mod 30 = 25 but 88 mod 30 = -2. A1jrj (talk) 14:01, 19 May 2013 (UTC)

Step one: using 45 - (y mod 19 × 11) mod 30 + 29, if y mod 19 = 5 or 16, to determine the date of PFM (PFMd).

Step two: using (y mod 100 + [y mod 100/4] + c + PFMd) mod 7 to determine the day of PFM (dPFM).

Step three: using PFMd + 7 - dPFM to determine the date of Easter (if the result > 31 the month = April).

where c = 3 for years 1900 ~ 1999, c = 2 for years 2000 ~ 2099, and c = 0 for years 2100 ~ 2199.

Take a few examples:

The year 2000 mod 19 = 2000 - 1995 = 2000 - 2000 + 5 = 5 hence 5 × 11 mod 30 = 55 - 30 = 25, so PFMd = 45 - 25 + 29 = 49 (Apri 18), and 2000 mod 100 = 0 hence dPFM = (0 + 0 + 2 + 49) mod 7 = 0 + 0 + 2 + 0 = 2 (Tuesday), so Easter Sunday = 49 - 31 + 7 - 2 = 18 + 5 = 23 April.
The year 1992 mod 19 = 16 hence 16 × 11 mod = 26, so PFMd = 45 - 26 + 29 = 48 (April 17), and 1992 mod 100 = 92 hence dPFM = (92(8) + [92(8)/4] + 3 + 48) mod 7 = 1 + 2 + 3 + 6 - 7 = 5 (Friday), so Easter Sunday = 48 - 31 + 7 - 5 = 17 + 2 = 19 April.
The year 2117 mod 19 = 2100 - 2090 + 17 - 19 = 2100 - 2100 + 10 - 2 = 8 hence (8 × 11) mod 30 = 88 - 90 = -2, so PFMd = 45 - (-2) = 47 (April 16), and 2117 mod 100 = 17 hence dPFM = (17 + [17/4] + 0 + 47) mod 7 = 3 + 4 + 5 - 7 = 5 (Friday), so Easter Sunday = 47 - 31 + 7 - 5 = 16 + 2 = 18 April.

So it is very easy to determine the date of Easter! --Q5968661 (talk) 10:06, 8 March 2013 (UTC)

Hello !!! Try it the other way around. I'll give you the dates: 22 March and 25 April, and you determine the year (the last and the next occurrence). Thanks. :–D 195.91.110.132 (talk) 16:07, 23 March 2014 (UTC)

Still confused

I am still confused as to why the golden number (date of the full moon) can be only 19 different days when a month is 30 or 31 or 28 (or more appropriately a lunar month is 29 or 30 days). Are there some days in the range of Mar 21 - April 20 or whatever that cannot be a full moon? (The golden number of any Julian or Gregorian calendar year can be calculated by dividing the year by 19, taking the remainder, and adding 1). This leaves only 19 possible days for a full moon out of 29 or 30.

The golden number is note a date; I don't have time to explain right now. Vincent J. Lipsio (talk) 19:09, 17 April 2014 (UTC)

In the Julian computus, there are only 19 dates in each 29-or-30-day cycle that the full moon can occur on. By the time of the Gregorian reform, those dates were about 4 days in error compared to the actual moon. In the Gregorian computus, in any given century there are only 19 dates possible for a full moon, but at the start of some centuries those dates shift ahead or back by one day because of the "solar correction" and "lunar correction" components of the formula for the epact. Indefatigable (talk) 19:13, 19 April 2014 (UTC)

I don't know all the details of the computus, but I do know this much. First, don't confuse the actual astronomical events with the calendar. Predictive calculation of astronomical events was not possible when computus was first created, nor even when the Gregorian reforms occurred. One could only tell when these events occurred through observation of the occurrence, at or after the fact. But the church needed to predict the Paschal moon in order to tell when Pascha was supposed to be celebrated, and only after that date was set could it know when to begin preparations for Pascha, the season that developed into Great Lent. They needed not only to be predictive, but up to two months in advance! The church was practical. It used the calendar as the best means then known to look ahead. Then it set up the computations based on the calendar. And in earliest times, that was the Julian calendar in the west, and the Alexandrian in the east (at least for the easterners who weren't depending on the Jews to make determinations of when Passover occurred, on the basis of the Hebrew calendar. (See the issue about that which was considered at the First Council of Nicea.)

Ancient observations of what had occurred historically were much more sophisticated than is normally supposed today, and records had been kept over centuries of time. Alexandria's famous library was a primary repository for much of such ancient knowledge, and that is why Julius Caesar went there to consult the best astronomers in constructing the Julian calendar in the first place, and why that calendar was a practical breakthrough in calendar quality. (The Alexandrian one was no slouch either.) But the ancients knew that it took 19 years for the moon to run through a whole number of lunar cycles within a whole number of years, and that is what the 19 is about in the computus. Computus also has to consider what days of the week the Paschal moon falls on, since Pascha must come on a Sunday following that moon (not on that moon). The reason in the Gregorian computus that there are only 19 dates possible in a given century (if that's really true), would have to do with the fact that in the Gregorian calendar, only one century year out of four is a leap year (unlike the Julian calendar). The computus accounts for leap years, because it is based on the calendar, and you have to account for that in order to keep the days of the week right.

Both calendars, over a sufficient number of centuries, and the Julian calendar more so, contain "inaccuracies" compared to actual astronomical events. They accumulate those inaccuracies over time, and eventually the discrepancies add up to a certain number of days. Within the last 50 years we have become able to calculate astronomical events, both observationally and predictively, with such accuracy that we can notice fractions of seconds (today, even microseconds or less) of differences. And we know that unpredictable astronomical events can and regularly do introduce tiny fluctuations in all these measures. We even adjust for those on a fairly regular basis now - perhaps you've heard of the "leap second". Suffice it to say that no calendar will ever be able to make predictions with complete accuracy if you project it far enough into the future. Computus survives because, in principle, it is still practical. Hope this helps some. Evensteven (talk) 20:01, 19 April 2014 (UTC)

Also, this year in Illinois the Vernal Equinox was Mar 20th. The equinox happens at a certain instant and in Illinois it was on Mar 20 but in Tokyo it was already Mar 21 since they are like 12 hours ahead of Illinois. What time zone do they use for equinox?

The meridian of Jerusalem by default; there was no knowledge of the problem (of longitude) when this algorithm was devised. Vincent J. Lipsio (talk) 19:09, 17 April 2014 (UTC)

Since the authors of the Gregorian calendar were working in Italian cities, I expect their tables were calibrated to the observations of local Italian astronomers. Indefatigable (talk) 19:13, 19 April 2014 (UTC)

One last question. Is Passover always in the middle (14th) of a month (often coinciding with Easter but occasionally almost a month ahead due to the extra month added sometimes to the year)? So is Passover always on a full moon? — Preceding unsigned comment added by 63.84.231.3 (talk) 17:58, 17 April 2014 (UTC)

The month (Nisan, IIRC), and all other months on the Jewish calendar, start at a new moon, so the 14th of the month is, therefore, a full moon, Vincent J. Lipsio (talk) 19:09, 17 April 2014 (UTC)

Drift in ecclesiastical full moon

Per WP:TPO, closing section created by IP sockpuppet of banned User:Vote (X) for Change

The following discussion has been closed. Please do not modify it.

It would be more helpful to tweak the wording, rather than nuke the section entirely. The value of delta T over 2,000 years has reached about 3 1/2 hours. The shift in the time of full moon over the same period is much the same. So the one cancels the other out. 156.61.250.250 (talk) 10:44, 14 May 2015 (UTC)

The deleted section appears to be largely based on "original research" which doesn't belong here. AstroLynx (talk) 10:50, 14 May 2015 (UTC)

(ec)My mistake. I see the objection is not to this bit, it's to the bit above. Nobody claimed it was badly written when it was part of a stable version five years ago. 156.61.250.250 (talk) 10:54, 14 May 2015 (UTC)

You said it was badly written. Now you claim it's original research. You seem to be making this up as you go along. The fact is that the change in delta t and the secular acceleration of the moon has been known for years. 156.61.250.250 (talk) 10:58, 14 May 2015 (UTC)

Actually it is both (original research as there is no reference and it is badly written - the cited numbers 84 and 19 come from nowhere). The fact that it was there five years ago is not relevant - it was evidently removed for the reasons which I mentioned just now. Changes in Delta T and the secular acceleration of the moon are indeed well known but you do not provide references for your claim. AstroLynx (talk) 11:03, 14 May 2015 (UTC)

It was removed as part of a batch by an editor claiming the article had been edited by a sockpuppet. The content was never questioned. 156.61.250.250 (talk) 11:45, 14 May 2015 (UTC)

Then why is it unsourced and badly written (see my earlier comments). Such content does not belong in WP. AstroLynx (talk) 11:54, 14 May 2015 (UTC)

I don't see it's badly written. Lichtenberg theorised a mixture of 200 and 300 - year correction steps. The actual error to be compensated is one day in about 210 years, so there will be many more 200 than 300 - year steps. The actual ratio turns out to be 84 to 19, but that's a side issue that doesn't affect the structure of his scheme. 156.61.250.250 (talk) 12:09, 14 May 2015 (UTC)

If the section which I deleted comes from Lichtenberg's paper this should be indicated more clearly. Still I do not feel that this level of detail belongs here, perhaps it should be moved to Reform of the date of Easter. AstroLynx (talk) 12:26, 14 May 2015 (UTC)

The consensus is that this level of detail does belong here. 156.61.250.250 (talk) 13:45, 14 May 2015 (UTC)

Kindly provide some proof of your claim. AstroLynx (talk) 13:58, 14 May 2015 (UTC)

It was a stable version five years ago. 156.61.250.250 (talk) 15:21, 14 May 2015 (UTC)

That is a very long time ago. So apparently nobody objected to its deletion five years ago? You still have not given any evidence that the deleted section is from Lichtenberg's paper. AstroLynx (talk) 15:42, 14 May 2015 (UTC)

Obviously it's not word - for - word because that would be plagiarism. As for your other point, doubtless many editors objected to the loss of their work when Jc3s5h arbitrarily turned the clock back. 156.61.250.250 (talk) 15:55, 14 May 2015 (UTC)

Lichtenberg describes the replacement of solar equations (common centennial years) and lunar equations (drift of one tithi in 312 1/2 years) with a unified correction cycle which provides the same number of epact corrections over the cycle. — Preceding unsigned comment added by 156.61.250.250 (talk) 16:20, 22 May 2015 (UTC)

You must be referring to another paper by Lichtenberg or someone else. I have read the Lichtenberg (2003) paper twice and I don't see anything which matches your claim. AstroLynx (talk) 10:44, 23 May 2015 (UTC)

He most certainly does use unified corrections (200 and 300 year steps). How do you claim he makes the adjustments? 156.61.250.250 (talk) 12:49, 23 May 2015 (UTC)

Lichtenberg's (2003) paper indeed discusses the lunar and solar corrections which are necessary to compute the correct lunar age but the section which I blanked is not a part of his paper as you claim. If you think that the blanked section (which I repeat here)

Using a more accurate leap year rule (218 leap years in 900 years with AD 2500 a leap year) the present system has a "lunar jitter" of 2.60 days. The unified correction cycle repeats every 22500 years starting in 1 BC. In this period there are 175 Gregorian solar corrections and 72 lunar ones for a net correction of 103 days. There are nineteen 300 - year correction periods interspersed among eighty - four 200 - year ones. These numbers are unforgettable - 84 is the number of years in the British epact cycle before the Synod of Whitby and 19 is the number afterwards. The 300 - year periods end in 1100, 2400, 3500, 4800, 5900, 7000, 8300, 9400, 10700, 11800, 13100, 14200, 15500, 16600, 17700, 19000, 20100, 21400 and 22500. In the first three millennia of the Christian era the lunar jitter is 3.16 days, which cannot be reduced by moving corrections which would otherwise occur in leap years. The cycle is furthest behind the moon in the 11th century and furthest ahead in the 22nd century.

is a part of his paper, please indicate where. AstroLynx (talk) 13:57, 23 May 2015 (UTC)

That's not what I asked you. My question was, does he or does he not make use of unified corrections in spaced out 200 - and 300 - year intervals to replicate the 43 epact corrections in 10 000 years? 156.61.250.250 (talk) 14:16, 23 May 2015 (UTC)

You're sidestepping the issue. The issue is the section which I blanked because it is unsourced and you keep on trying to put back. Up to now you have not be able to provide a verifiable source. AstroLynx (talk) 14:26, 23 May 2015 (UTC)

It's you who is "sidestepping the issue". I happen to know that's exactly how he did it. 156.61.250.250 (talk) 14:43, 23 May 2015 (UTC)

Provide a verifiable source then. AstroLynx (talk) 14:58, 23 May 2015 (UTC)

The verifiable source is obviously the paper itself. 156.61.250.250 (talk) 15:03, 23 May 2015 (UTC)

Sorry to puncture your balloon, but as I already indicated above the blanked section is nowhere to be found in Lichtenberg (2003). How long do you want to keep on this farce? I can keep up just as long as you can. AstroLynx (talk) 15:08, 23 May 2015 (UTC)

As your parachute deflates, you can ponder that there is no requirement that ideas should be reported in the precise words of the person who originates them.156.61.250.250 (talk) 15:20, 23 May 2015 (UTC)

The problem is that most of the numbers and years listed in the blanked section are nowhere mentioned in Lichtenberg's (2003) paper. Nor is the Synod of Whitby anywhere referred to in the same paper. How can you claim that your text is a selective rephrasing of Lichtenberg's paper? I am seriously beginning to doubt whether you have ever read the Lichtenberg paper you are claiming to cite. AstroLynx (talk) 15:38, 23 May 2015 (UTC)

As I say, it doesn't have to be exact. To say that "most" of the years are different is not good enough. Please list specifically the years which you find to be different. 156.61.250.250 (talk) 15:44, 23 May 2015 (UTC)

OK, which table or page lists the years 11800 and higher? Where is the Synod of Whitby mentioned? Where are the values for lunar jitter mentioned? Unless you can provide specific page numbers from Lichtenberg's (2003) paper mentioning these numbers, I can only conclude that it is unsourced. AstroLynx (talk) 16:02, 23 May 2015 (UTC)

As I pointed out, Lichtenberg does the calculation over 10,000 years so you wouldn't expect to see years 11,800+ appearing in his data. You can still provide the data for earlier years. 156.61.250.250 (talk) 12:09, 24 May 2015 (UTC)

So you finally admit that most of the numbers and years listed in the blanked section are not found in the Lichtenberg (2003) paper but are based on original research (probably yours?). I still seriously doubt whether you have actually read Lichtenberg's paper.

As I became interested in the reason why you persist in having this section restored I had a closer look at the original edit on 10 June 2010 by IP 94.194.22.179, one of the known London-based sockpuppets of Vote (X) for Change.

Comparing the latter's editing behavior (and that of his numerous sockpuppets) with yours has been very insightful and revealing - it explains so much. AstroLynx (talk) 12:57, 24 May 2015 (UTC)

What a load of rubbish! I drew attention to the "sockpuppet" issue earlier in this thread. The section was declared kosher five years ago so you have a high bar to overcome to remove it. You're creating a diversion by throwing mud around in the hope that some will stick. If you're not prepared to justify your position by detailing the "discrepancies" you allege you can be dismissed as just a troll. I object to people removing good content for spurious reasons. This subject was extensively discussed in the archives - I suppose Karl Palmen, Tom Peters and the rest are all sockpuppets as well. 156.61.250.250 (talk) 13:17, 24 May 2015 (UTC)

Despite your whining you still have not provided a verifiable source for the blanked section - original research doesn't belong in WP. The blanked section never was 'declared kosher' five years ago as you claim - it was repeatedly removed after your sockpuppets tried to put it up again and again.

If I were you I would also start worrying about your recent additions to Iranian calendars. I have noticed that you have added quite a lot of unsourced text in recent days. If you do not start adding verifiable sources soon I will remove most of it - including the sections which appear to be original research. AstroLynx (talk) 15:24, 24 May 2015 (UTC)

Nah, it was added once and there it stayed. I think we can assume from your reticence that the dates in the first ten thousand years are in agreement with those in the paper. 156.61.250.250 (talk) 15:36, 24 May 2015 (UTC)

Why don't you check yourself - you claim to have read the Lichtenberg paper. AstroLynx (talk) 15:43, 24 May 2015 (UTC)

Information like that is not something you would commit to memory. 156.61.250.250 (talk) 09:00, 27 May 2015 (UTC)

You should have kept notes. If you had access to the Lichtenberg paper five years ago, you should be able access it again now. AstroLynx (talk) 09:49, 27 May 2015 (UTC)

This is typical of your attitude. Looking at your editing history I see you've tried this trick before. When someone wants to do an end run around Wikipedia policy it's they who have to do the legwork. We have a verified stable version which you want to change. In the previous incident you were trying to skate round WP:V. Editors pointed to sources which they wanted to include in an article and you pointed to a book which you claimed supported your viewpoint. Asked to support this by quoting the relevant passage you said you didn't have a copy and would order it through the library but

You will have to ask this nicely and include the magic word.

Later, when an editor referred to the book you quoted from it instantly, thus proving that you had had a copy of it all the time. You exhorted the editor to get a copy of the book:

Why don't you prove me [Vgent] wrong by looking it up for yourself?

That's not what good faith, collegial editing is all about. The incident related to an attempt by you to exclude the other editors' sources from a description of a picture of Muhammad. You have this picture on your web profile. You are arguably obsessed with it. 156.61.250.250 (talk) 10:55, 27 May 2015 (UTC)

You are the person who wants to put his original research up, so you must provide the correct references. You cannot expect that other editors will do this for you. That fact that you seem to be unable to do this clearly proves that you actually never saw the Lichtenberg paper which you claim is your source. Your editing behaviour during the past few days on this page has thus clearly shown that you are a liar and a fraud.

It is interesting that you refer to the discussion on the Islamic calendar page with IP 87.81.147.76. For any editor it should be clear that this IP's editing behaviour is exactly the same as yours, which, for me at least, makes it abundantly clear that you the same person. Yet another London-based sockpuppet of Vote (X) for Change. AstroLynx (talk) 11:21, 27 May 2015 (UTC)

Wow. Every editor in a university city of ten million with an interest in calendars is a sockpuppet. You say a stable version is original research. Proof? I see that 87.81.147.76 was supported by other editors, so I suppose that makes them liar(s) and (a) fraud(s) too. An editor (not talking of me) uses the word "disingenuous":

The RFC closing statement didn't deal with the NPOV concerns, hence shouldn't be relevant to this discussion. Also, Hillenbrand's "sectarian" conclusion is notable, and usually the sole reason why reliable sources refer to his paper and this image set. It is disingenuous to ignore his conclusion and pass this image/MS. as neutral.

Speaking of you he says:

Your assessment is not entirely accurate. Hillenbrand didn't see a need to examine the other two images (note the word "already" in his conclusion). Also, the third image that he used to prove the "strong sectarian feeling" (The Envoy of Musailama) has nothing to do with the investiture of Ali. Incidentally, his arguments concerning that image also applies to this image (the prominent placement of ahl al-Bayt; al-Hussien is favored over al-Hassan by being placed closest to the prophet; etc). I didn't get your last point though, since the image is still relevant in the book's own page or where an example of sectarian or polemical art is needed.

And again:

Well, it was clear enough that you misrepresented Hillenbrand. He actually noticed a "strong sectarian feeling" in an image that was not about the investiture of Ali, contrary to what you wrote above. Then based on how you misrepresented the source you assumed that the RFC image has no sectarian purposes, which is original research (and can be disputed by pointing at Hillenbrand's arguments about the Envoy to Musailama). In any case, I don't see a reason why I should change my !vote. Adding pages/images to a general article from a source that has been described as sectarian and agenda driven is not inline with npov.

And again (of one of your supporters):

You actually missed an important argument. The image is found in a book that was illustrated for sectarian and polemical purposes. I would expect the closing party to justify why polemical/sectarian imagery/works should be given space in this article.

And again (of another of your supporters):

Actually, the image is being used here in the same context that reliable sources have described as sectarian (i.e., illustrating a neutral text about calendars). So your claim that it's "being used in non-sectarian/non-polemical ways" is nonsensical. Your 2nd point shows that you have very little experience in content editing. We are not supposed to give preference for one source and ignore others. In this case, reliable sources differ on caption and pertinence. This article must reflect these differences to meet WP:NPOV.

As the discussion unfolded it emerged that none of your supporters had read any of the books which their arguments relied on but you did not call them "liars and frauds". 156.61.250.250 (talk) 13:42, 27 May 2015 (UTC)

I'm stepping in to this argument because it's getting really stupid. First of all, as per WP:BURDEN: "The burden to demonstrate verifiability lies with the editor who adds or restores material, and is satisfied by providing a citation to a reliable source that directly supports the contribution... The citation must clearly support the material as presented in the article." Since 156.61.250.250 is the one restoring the content they must prove that the citation supports the material, and based on what I have read above they have failed to do so. Secondly, past arguments shouldn't be intruding on this page.

At any rate, this discussion isn't moving anywhere. If 156.61.250.250 cannot provide a quote from Lichtenberg to support their case then as far as I'm concerned the material is unsourced. Arcorann (talk) 02:25, 28 May 2015 (UTC)

What happened back in 2010 is that the material was added and found to be policy - compliant. That's why it became a stable version. Then in the back - and - forth of editing it got removed and came back. Obviously this does not mean that the person who happened to restore the material at any point in time is lumbered with going through the whole verification process again. 156.61.250.250 (talk) 13:50, 28 May 2015 (UTC)

I will also step in to say that what happened 5 years ago is procedurally irrelevant. If the desired material was supportable then, then it is still supportable, excepting only additional work that overturned it. Whether or not it was backed up by citation then, it needs to be backed up by citation now. If numbers and calculation is required, the source must provide it. It comes down to put up or shut up, because editorial opinion is insufficient, sockpuppet or not. Evensteven (talk) 20:12, 28 May 2015 (UTC)

Just because the material was in a stable version does not imply that it is correct or policy-compliant - I know that factually incorrect material can persist in a supposedly stable version for a long time, having corrected a few such errors myself. As per the section I quoted of WP:BURDEN, if you're restoring material and it's challenged you must be able to go through the verification process and show that a source exists, regardless of whether it has appeared in the article before. As said by Evensteven, if it's supportable then it's supportable now, and your failure to support it now suggests that it wasn't supportable then either, and that its removal was a correct decision. Arcorann (talk) 00:51, 29 May 2015 (UTC)

More or less as an aside, when I was learning computer programming

(Specifically COBOL -- this was in the antediluvian days when the IBM 370 was popular), I had to write a program to determine which day of the week a given date was. The way I did it was first to work out by hand that January 1st of the year 1 was a Saturday, count the number of days since then (and I dropped the 11 days in 1752 when Britain switched from the Julian calendar to the Gregorian calendar), do modulo 7 on the number, and there we were. My instructor thought it was an ingenious way of doing it. JHobson3 (talk) 19:25, 18 January 2016 (UTC)

Do we need an algorithm for a revised Easter Day computation?

In the past weeks two editors have been adding material criticizing existing Easter Day computations and promoting a revised Easter Day algorithm recently published by one of the editors in a rather dodgy journal. I do not believe that such material belongs here but before I start deleting all this material I would like to hear the opinions of other editors. AstroLynx (talk) 13:25, 10 February 2016 (UTC)

Wikipedia cautions strongly against primary sources. Unless these methods have been peer-reviewed in reputable sources, they may be mentioned and linked to, but should not receive the extensive coverage they occupy now. -- Michael Bednarek (talk) 18:13, 10 February 2016 (UTC)

My understanding from reading the material is that the author believes that their algorithm produces "better" Easter dates than the Gregorian algorithm. If so, then not only are their supposed criticisms invalid (the algorithms given calculate Gregorian Easter, which is correct - see also Stockton who has tested many algorithms, including ones based on the original source), but also the content should fall under WP:FRINGE and therefore be eliminated. Arcorann (talk) 11:43, 16 February 2016 (UTC)

Or more likely WP:UNDUE. At any rate, I've removed the content. The person is free to make their case here if they disagree. Arcorann (talk) 12:33, 16 February 2016 (UTC)

No Full Moon

I notice that over and over again editors pervert the Easter rule by introducing an (ecclesiastic) Full Moon. This is misinformation. Surely popular and astronomical secondary sources mention a Full Moon, but canonical literature from Dionysius to Clavius has always only been based on the 14th day of the lunar month, following from the definition of Jewish Passover. That this happens to be close to the day of Full Moon for a properly aligned lunar calendar (which the ecclestiastic reckoning is not), is circumstantial. Referring to the "Full Moon" has led only to misguided attempts to "improve" the computation of Easter.

I noticed only now that my latest attempt (7 Jan. 2015) to clarify things was quickly (26 Jan.) destroyed by an anonymous editor without further comment. I may try again in the near future to repair this when I have time, but can all those editors who feel the urge to "improve" this article please first read up on the relevant literature and not put in misguided misinformation?

Tom Peters (talk) 12:41, 22 February 2016 (UTC)

The downside of an encyclopedia that everyone can edit is sometimes that everyone can edit it. That IP editor, 156.61.250.250, has been blocked soon after their edits here. If you think the current wording is wrong, the obvious recourse is to correct it – preferably with citation of reputable sources. -- Michael Bednarek (talk) 02:33, 23 February 2016 (UTC)

Ref Nr. 52 is broken

and it points to a domain selling site now.Douira100 (talk) 17:58, 21 November 2016 (UTC)

It would be useful to know what years the various algorithms are accurate over

I can't find a source for the years the Meeus/Jones/Butcher algorithm works. Can anyone supply? — Preceding unsigned comment added by 73.149.40.182 (talk) 17:12, 31 December 2016 (UTC)

Excel formulas

The Excel formulas under the Software heading deliver different results. It should be stated that they have different scopes. The formulas

=ROUND(DATE(A1,4,1)/7+MOD(19*MOD(A1,19)-7,30)*14%,0)*7-6
(1900 – 2203)

and

=FLOOR((4&-A1)-DAY(5)+97%*MOD(18.998*MOD(A1+8/9,19)+INT(68%*INT(A1%)-INT(A1%/4)-5/9),30),7)+DAY(1)
(1900 – 9999)

differ in 5802 of the 8100 calculations for the years 1900-9999. The latter formula is the correct for those years while the former is probably only intended for the years 1900-2199 for which it is also correct. The difference is always a (positive or negative) multiple of 7 days, i.e. they both get the day of week right but not the week itself. Presumably it omits correcting for the period of the orbit of the moon. The first error occurs in the year 2204 which is why it might go unnoticed for a while. Other well-known formulas that contain errors include the formula

=FLOOR(A1&"-05-"&DAY(MINUTE(A1/38)/2+56);7)-34

from "Excel 2010 Tips & Tricks" by John Walkenbach and other sources which makes 5790 errors in the year range 1900-9999, the first one occurring already in the year 2079. Thus, it is not even intended for the period 1900-2199.

A final comment about the 9999 year formula is that it is not compatible with international date standards - FLOOR((4&-A1) must be replaced by FLOOR((A1&-4) for non-US date formats. — Preceding unsigned comment added by 31.15.60.198 (talk) 17:08, 22 March 2016 (UTC)

The scope of the longer formula is not only the full range of years, it also works in Excel files with both file formats - the 1900 (Standard) and the 1904 file format.

In 1998/99 there was a contest in Germany, to get the shortest Excel formula for 1900-2078 in the 1900 file format. Most formulas you can find in the www, are from this and do not care about 2079-9999 or the 1904 file format.

13:55, 27 April 2016 (UTC) — Preceding unsigned comment added by 85.199.76.137 (talk)

The "Tips&Tricks by John Walkenbach" formula is the winner of the German contest in 1998/99. For 1900-2078 and 2080-2203 the formula works correctly.

To catch 2079 also, you have to precise it from "...)/2+56)..." into "...)*49%+56.57)..." but after 2203 you are lost.

85.199.76.86 (talk) 14:42, 13 September 2016 (UTC)

Bug Alert

At the main page
=FLOOR((4&-A1)-DAY(5)+97%*MOD(18.998*MOD(A1+8/9,19)+INT(68%*INT(A1%)-INT(A1%/4)-5/9),30),7)+DAY(1)
is correct for 1900/1904-9999 in the 1900 and 1904 date system in Excel.

But there is a new not fully tested 1900-9999 Excel formula, added 2017-04-24 by user Sigurdhu
~~=FLOOR(DATE(A1,3,27)+0.97*MOD(18.998*MOD(A1+8/9,19)+INT(0.68*INT(A1/100)-INT(A1/400)-5/9),30),7)+1~~
It is OK in the 1900 (standard) date system, but fails for all years in the 1904 date system
Please remove this.
09:46, 28 April 2017 (UTC) — Preceding unsigned comment added by 85.199.76.86 (talk)

The untested formula (for 1904 dates) has been removed.
18:20, 2 May 2017 (UTC) — Preceding unsigned comment added by 80.187.103.61 (talk)

=FLOOR(DATE(A1,4,97%*MOD(18.998*MOD(A1+8/9,19)+INT(68%*INT(A1%)-INT(A1%/4)-5/9),30)-DAY(4)),7)+DAY(1)
is a correct modification for 1900 and 1904, and is independent of country-specific date constructions like (4&-A1)

10:33, 8 May 2017 (UTC) — Preceding unsigned comment added by 85.199.76.86 (talk) 18:36, 8 May 2017 (UTC)

User Sigurdhu - Please stop !!!

Your new Excel Formula at the main page is also completely wrong in the 1904 date system of Excel - for all years 1904-9999

Please remove this and stop to insert untested Excel formulas. Why do you do this?

I cannot follow the thesis "similar to above, but probably easier to comprehend", because the first formula is correct and the new one only meets in the standard 1900 date system and for the 1904 system all results are wrong

13:20, 21 May 2017 (UTC) — Preceding unsigned comment added by 85.199.76.86 (talk)

The untested wrong formula has been improved, the DAY()-function helps to handle the differences between both date systems 1900 and 1904

22:13, 28 May 2017 (UTC)

Software - old school solutions and methods are dominating

Most of software solutions are based on the equation sets of Gauss and his epigones.
These sets were made for mental arithmetic.
Modern software products are able to calculate the Easter Sunday date from the year in one step, short, and efficient.

The "mental arithmetic" algorithms will long outlive what follows

Yes, but the "mental arithmetic" will be of use indefinitely, long after the ephemeral modern languages and applications that follow come to dwell amongst the doornails; then, the "mental arithmetic" algorithms will be needed to create programs that execute in the future.
Vincent J. Lipsio (talk) 01:19, 20 July 2016 (UTC)

If you are always right, all future programming has to support the "mod" functionality, which is not self-evident.

i.e. in MsQuery, the common driver for Access, Excel, and dBase does not provide a mod operator or a MOD( ) scalar function.

That's why the mental algorithms are not a universal reference
85.199.76.86 (talk) 18:34, 6 September 2016 (UTC)

Firstly, the remainder on division, aka, the "mod" functionality, can always be derived in any language that supports integer division. Also, it is most unlikely that any general purpose programming language would not directly support the remainder on division, the "language" you cited not being a general-purpose programming language but, rather, as you called it, "driver" (although methinks "engine" or "special purpose language" better terms). Vincent J. Lipsio (talk) 12:23, 7 September 2016 (UTC)

Open Office - Calc (Gregorian, 1583-9957)

=EASTERSUNDAY(A2)

EASTERSUNDAY(year) works in LibreOffice as well as OpenOffice. Roches (talk) 14:55, 10 July 2016 (UTC)

Visual Basic and VBA(Gregorian)^[2]

Public Function EASTERSUNDAY(y As Integer) As Long ' for 1900-9999 as serial number
EASTERSUNDAY=7*((CDate(y &-5)+.967*(18.99*(y Mod 19+732)+(y\100)*.6794\1-y\400)Mod 29)\7)-34
End Function

Public Function EASTERSUNDAY(y As Integer) As String ' for 1500-9999 as string
EASTERSUNDAY=CDate(7*((CDate(y &-5)+6*(y<1900)+.967*(18.99*(y Mod 19+732)+(y\100)*.6794\1-y\400)Mod 29)\7)-34)
End Function

MS Access (Gregorian)^[3]
a) for 1900-9999 as serial number (date)
=7*((CDate([year]&-5)+,967*(18,99*([year]Mod 19+732)+([year]\100)/1,4718\1-[year]\400)Mod 29)\7)-34)
b) for 1500-9999 as string
=CDate(7*((CDate([year]&-5)+6*([year]<1900)+,967*(18,99*([year]Mod 19+732)+([year]\100)/1,4718\1-[year]\400)Mod 29)\7)-34)

MS Excel (Gregorian)^[4]
a) for the standard 1900 file format (1900-9999)
=FLOOR((A2&-8)-MOD(30*INT(11*MOD(A2,19)-68%*INT(A2%-64)+INT(A2%/4)+1),29.032),7)-97

b) for the 1904 file format (1904-9999)
=FLOOR((A2&-5)-MOD(30*INT(11*MOD(A2,19)-68%*INT(A2%-64)+DAY(A2%/4)),29.032),7)-5

c) for 1900 and 1904 file formats (1900/1904-9999)
=FLOOR((A2&-5)-MOD(30*INT(11*MOD(A2,19)-68%*INT(A2%-64)+INT(A2%/4)+1),29.032)-DAY(6),7)+DAY(1)

79.201.192.95 (talk) 15:25, 18 December 2017 (UTC)

MS Query formula assistant with MS Add-In for Access-, Excel-, dBase- and text tables: Gregorian Easter Sunday for 1583-9999^[5]

as a (serial) number:	7((Cdate(y&-5)+6(y<1900)+Format(.5013(y/19-y\19+49)+(.6794(y\100)\1-y\400)/720,'ns')\205)\7)-34
as a date:	`Cdate(number)`
as a text:	`Format(number,'yyyy-mm-dd')`

The letter y is for fieldname or [field name] or ['alias']
If you only have years 1900-9999 then you can remove +6*(y<1900)
85.199.76.86 (talk) 13:56, 22 November 2017 (UTC)

References

Council of Nicaea and Easter

Copied from User talk:Joe Kress#Council of Nicaea and Easter

Joe

I saw your new edit to Computus citing Hefele's 1883 History of the Christian Councils. Reading between the lines of his treatment, there was dispute in the 19th century about what exactly had been decided at Nicaea. Although more recent scholarship has looked in some detail at the matter, there are still disputes about what was decided at Nicaea. Just a heads up that I'll probably rewrite your addition once I have time to dig into Mosshammer (2008) and Nothaft (2018) to clarify what the bishops at Nicaea did and didn't decide. --SteveMcCluskey (talk) 01:58, 26 October 2018 (UTC)

Thanks for the Mosshammer (2008) reference (The Easter Computus and the Origins of the Christian Era), which I have. He states (p. 52) "This denial to Nicaea of any Paschal rule may have gone too far." Concluding, "It is therefore fair to say that the Council 'apparently' or 'implicitly' endorsed the rule of the equinox, even if it published no rule as such." Earlier (p. 51) he states that Constantine said that the council's "prohibition against keeping Passover 'with the Jews' was interpreted in ancient times as directed specifically against Quartodecimanism and as precluding the observation of Easter on the 14th day of the moon, even if that day was a Sunday." Thus Mosshammer attributes Hefele's rules to the Council, in some fashion. I do not have access to Nothaft (2018) (Scandalous Error). — Joe Kress (talk) 02:35, 27 October 2018 (UTC)

That's pretty much what I see in Mosshammer, although just before his comment on the Council implicitly endorsing the rule of the equinox, he says that "By deferring to Rome and Alexandria the council endorsed whatever [my emphasis] methods were in use by those churches and relied on the bishops of Rome and Alexandria to resolve any differences". After digging into Nothaft, I find that he has nothing to say about the question of Nicaea. He mainly talks about the later influence of Dionysius's false attribution of the 19-year Alexandrian cycle to the council. If this article's historical component is to be brought up to date, it should discuss that claim and its influence.

Maybe we should copy this discussion to Talk:Computus and continue it there.--SteveMcCluskey (talk) 19:32, 27 October 2018 (UTC)

Without access to Nathaft, I cannot comment on Dionysius's false attribution of the 19-year cycle to the council, especially its later influence, although I note that the entire computus of Dionysius, including its 19-year cycle (but not its lunar cycle), was already being used by Alexandria at the time of the council, 200 years earlier, the only difference being that all dates were given in the Alexandrian calendar instead of the Julian calendar. The Alexandrian tables at the time of the council are preserved as Ethiopian tables using the Ethiopian calendar, which Otto Neugebauer published in Ethiopic Astronomy and Computus, recently reprinted by Red Sea Press. You note that Mosshammer states that the council endorsed whatever methods were in use by Rome and Alexandria, which means the council endorsed Alexandria's use of the 19-year cycle. — Joe Kress (talk) 05:19, 28 October 2018 (UTC)

The Council apparently agreed

Is it possible to explain or expand the wording in History "The council apparently agreed to two rules without explicitly stating them,"? Is that all the citation says? Does Mosshammer 2008 really say "I can't find any evidence that the Council agreed to anything, let alone exactly what it was they agreed"? And if it is really the case then the basis of this article, in which we claims that the foundation for Computus is the First Council, is simply false! Or have I missed something? --John Maynard Friedman (talk) 00:29, 29 October 2018 (UTC)

Here's a relevant quotation from Mosshammer's p. 52: "This denial of any Paschal rule may have gone to far.… By deferring to Rome and Alexandria the council endorsed whatever methods were in use by those churches and trusted the bishops of Rome and Alexandria to resolve any disagreements. It is therefore fair to say that the Council 'apparently' or 'implicitly' endorsed the rule of the equinox, even if it published no rule as such." --SteveMcCluskey (talk) 02:41, 29 October 2018 (UTC)

Synod of Mag Léne: the "insular" Easter dispute

As it is well above my pay grade, could someone who understands this kind of material take a look at https://books.google.co.uk/books?id=7ld9DAAAQBAJ&pg=PA93&lpg=PA93&dq=%22synod+of+mag+l%C3%A9ne%22&source=bl&ots=JKMpfYe48-&sig=ACfU3U0VLYNb0wywx6cPWXg8MiyttiG9ng&hl=en&sa=X&ved=2ahUKEwjsgPjrkJnhAhWSRBUIHVRRAlsQ6AEwAXoECAUQAQ#v=onepage&q=%22synod%20of%20mag%20l%C3%A9ne%22&f=false and extract the appropriate citations, please? Note that the author gives the reference for Bede as well as for the Synod.^[1] --John Maynard Friedman (talk) 21:03, 23 March 2019 (UTC)

I see that there is some more material at Celtic Christianity#Easter calculation which may be useful. Or then again maybe not because the text that has been added recently to this article (about the practice of the Church in GB&I) seems to stray into wp:FORK territory? --John Maynard Friedman (talk) 21:09, 23 March 2019 (UTC)

Besides Bede's Ecclesiastical History of England^[2] which is probably the principal source, the "Introduction" in Annals of Ulster by MacCarthy^[3] in Volume 4 has a lot of info but dated, and still in copyright but indispensable is the "Development of the Latin Ecclesiastical Calendar" by Charles W. Jones in his Bedae Opera de Temporibus, 1943, which has a long chapter entitled "Easter in the British Isles". I'll be checking these for useful citations. Several other more modern journal citations and books are also useful. — Joe Kress (talk) 19:15, 24 March 2019 (UTC)

MacCarthy mentions Magh-Lene on pages cxxxviii and cxlii but is hard to understand. Jones is concise on page 90, stating "The letter (of Cummian) is at once a report and an apology or justification to Abbot Seghine at Iona of a synod held at Campus Lenis (Magh-Lene), where the Easter question was considered. The direct result of the synod was an alteration in the observance among the southern Irish and the adoption of the Alexandrian reckoning." — Joe Kress (talk) 21:44, 26 March 2019 (UTC)

For further sources on the Synod of Mag Léne see Maura Walsh and Dáibhí Ó Cróinín in their 1988 edition of Cummian's letter De controversia paschalis. Cummian called the Synod at Mag Léne and in his letter mentioned ten different sources of Easter cycles available in Ireland when he wrote.