Talk:Epoch

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia
(Redirected from Talk:Epoch (reference date))

No citations for January 1, 0[edit]

I have removed the line in the table for January 1, 0, because there are no citations. Details follow.

Symbian appears to be effectively defunct, which makes it hard to obtain information about it. I was not able to locate any documentation that seemed likely to address the epoch.

"Turbo DB" is difficult to identify. A web search returns results for seemingly different products.

MATLAB has a few different ways to represent dates. http://www.mathworks.com/help/matlab/ref/datenum.html states "The datenum function creates a numeric array that represents each point in time as the number of days from January 0, 0000. The numeric values also can represent elapsed time in units of days. However, the best way to represent points in time is by using the datetime data type." This seems to contradict the information in the article, and also makes it doubtful whether a date associated with the datenum function should be mentioned in the article when MathWorks seems to consider the datetime data type superior. Jc3s5h (talk) 17:32, 8 April 2015 (UTC)[reply]

Datetime was only introduced in R2014b; before that version MATLAB only used serial dates, so the entry was accurate when it was added. The document by Moler (http://www.mathworks.com/moler/intro.pdf formerly cited on the page) still uses the datenum function, and was last updated 2013. Given that MATLAB serial dates are still important in legacy applications (and in its open-source clone Octave) I would recommend restoring it. Arcorann (talk) 04:57, 25 April 2015 (UTC)[reply]
Documentation has been located for Symbian. Turns out the correct date should be 1 January, 1 CE (not 0 CE). The page appears to be a copy of the official documentation; the original source is unknown. At any rate, the page gives enough info to locate a source suitable for Wikipedia. Arcorann (talk) 05:13, 25 April 2015 (UTC)[reply]
So I found a more recent version of the same documentation and it turns out it's contradicting itself. This page on date and time handling claims 0 CE, while the page corresponding to the earlier link still states 1 CE. At this point I no longer have the motivation to research further. Arcorann (talk) 05:33, 25 April 2015 (UTC)[reply]

Rationale cell for January 1, 1[edit]

In the last column of the table, the line for January 1, 1, said the rational was "Common Era, ISO 8601 date 0001-01-01". Each computing system or program that uses this has a citation. Only one of the sources provide a rationale for choosing January 1, 1: Dershowitz and Reingold. They write on pages 10

The date Monday, January 1 (Gregorian), though arbitrarily chosen as our starting point, has a desireable characteristinc: It is early enough that almost all the dates of interest are represented by positive integers of moderate size.

Rather than inventing rationales that are not stated in the sources, let the readers draw their own conclusions. For dates that have characteristics that the reader has probably never noticed, it's helpful to mention the characteristic (like 1904 being the first leap year of the 20th century). But the readers are just as capable of inferring the reason for January 1, 1, as an editor.

Also, many of the other comments in the column are probably observations of the editor who added them, rather than rationale explicity stated by the creators of the systems that use them. So it might be better to relabel the column to something like "Characteristics" or "Comment". Jc3s5h (talk) 22:09, 24 April 2015 (UTC)[reply]

Rata Die has a block quote about the REXX date() function, the same REXX reference as here. This date() function can handle days in various formats, one of these formats is known as "base" format and can handle non-negative whole numbers up to at least 999,999,999. The date() function can convert to other formats, one known as "S" (standard) and written as yyyymmdd (8 digits). The rationale for picking base 0001-01-01 is therefore inherently "because this is supposed to work" for anything plausible in the standard format starting at 00010101 (base 0) and forward. With a proleptic Gregorian calendar this matches exactly ISO 8601 0001-01-01, and roughly Common Era. The REXX author picked KISS and POLA—he was one of the POLA inventors—and "any base date modulo 7 is the number of the day of the week" is a nice feature. –Be..anyone (talk) 22:45, 24 April 2015 (UTC)[reply]
REXX added with ISO 2014 and the two relevant page numbers, date() standard format for ISO 2014, the predecessor of ISO 8601, and date() base format replacing an older century format, as explained above. –Be..anyone (talk) 23:16, 24 April 2015 (UTC)[reply]
It is an interesting feature that Rata Die modulo 7 is the day of the week, if you choose to assign 1 to Monday. Certainly if the REXX source says nice things about this feature, we could include it. (However, the citation for REXX in Rata Die does not seem to contain the quote which you attribute to it: "any base date modulo 7 is the number of the day of the week".) Perhaps we could add it ourselves as an interesting observation, but the column heading should be changed. It still has nothing to do with ISO 8601; January 1, 1 would be just as natural, and Rata Die would still have the modulo 7 property, if ISO 8601 had never been written. Jc3s5h (talk) 23:28, 24 April 2015 (UTC)[reply]
Complete quote from the online source:
Base
returns the number of complete days (that is, not including the current
day) since and including the base date, 1 January 0001, in the format:
dddddd (no leading zeros or blanks). The expression DATE('B')//7
returns a number in the range 0–6 that corresponds to the current day
of the week, where 0 is Monday and 6 is Sunday.
Note: The base date of 1 January 0001 is determined by extending
the current Gregorian calendar backward (365 days each year, with
an extra day every year that is divisible by 4 except century years that
are not divisible by 400). It does not take into account any errors in
the calendar system that created the Gregorian calendar originally.
–Be..anyone (talk) 00:36, 25 April 2015 (UTC)[reply]

Pronunciation[edit]

how do you say this in english?

  • like epic: IPA ɛpɪk
  • like eepock: IPA iːpɔːk

please, someone who knows, edit this in as in this article: Heidi — Preceding unsigned comment added by OsamaBinLogin (talk • contribs) 05:42, 6 February 2016 (UTC)[reply]

Dubious source for Oracle database epoch[edit]

In these edits User:Kevinthenerd added an epoch for the Oracle Database of January 1, 4712 BC. There are several problems with this edit.

The source says "A Julian day number is the number of days since January 1, 4712 BC. This is false. A reliable source, Richards (see footnote 4 in article) states "The Julian Day Number ... of Tuesday −4712 January 1 is 1." (p. 592) (By the way, Richards is only correct if it is understood that January 1, &minus4712, is stated in the Julian calendar, not the Gregorian calendar).

The Oracle source also gives this example:

Example This statement returns the Julian equivalent of January 1, 1997:
SELECT TO_CHAR(TO_DATE('01-01-1997', 'MM-DD-YYYY'),'J')
FROM DUAL;
TO_CHAR
--------
2450450

But for the same date, the Multiyear Interactive Computer Almanac from the US Naval Observatory provides this result: 

                           CALENDAR                            
     
   Date        Time       Day      Julian Date      Day-of-Year
         (TT)                         (TT)             (TT)
          d  h  m   s                    d              d
1997 Jan 01 00:00:00.0    Wed     2450449.500000       1.000000

So I conclude the source is too poor to figure out what the epoch is for the Oracle Database and a better source is required. Accordingly, I have reverted the edits. Jc3s5h (talk) 21:41, 12 May 2016 (UTC)[reply]



The source by user:Kevinthenerd is the official source on Oracle Database, a commercial relational database software suite. There is no better source of information on the design of this software. The discrepancy noted by User:Jc3s5h is explained therein; both of the two ways of counting are described in the source material, and the source listed by User:Jc3s5h happens to use the other. kevinthenerd 20:21, 14 September 2020 (UTC)[reply]


Here is some original research, with commands entered into the Oracle Database software and their resulting responses. Even if Oracle Corporation is wrong about their description of their epoch, this still the way their software works. This is the best-selling database software on the market, so the relevance would be difficult to dismiss. 

   SQL> SELECT TO_CHAR(TO_DATE('1997-01-01', 'YYYY-MM-DD'), 'J') AS dt FROM DUAL;
   
   DT
   ----------------------------
   2450450
   
   SQL> SELECT TO_CHAR(TO_DATE('1997-01-02', 'YYYY-MM-DD'), 'J') AS dt FROM DUAL;
   
   DT
   ----------------------------
   2450451
   
   SQL>

kevinthenerd 20:34, 14 September 2020 (UTC)

Does the Oracle database treat these dates as integers, floating point, or what? Could it be a matter of rounding? What time zone does it use, if any? Jc3s5h (talk) 22:16, 14 September 2020 (UTC)[reply]
Also, in the years since the edits in question were made, computer-related epochs have been moved to Epoch (computing). Jc3s5h (talk) 22:22, 14 September 2020 (UTC)[reply]
As per the documentation (version 20 here), the DATE data type of the Oracle database stores dates internally as a tuple of (year, month, day, hour, minute, second). Storing time zone information requires TIMESTAMP WITH TIME ZONE, but this data type is also a tuple. Thus, the Oracle database does not have an epoch. The above commands use the database software's conversion capabilities to convert to/from a date format ("J", representing Julian days). Arcorann (talk) 15:12, 15 September 2020 (UTC)[reply]

External links modified[edit]

Hello fellow Wikipedians,

I have just modified 3 external links on Epoch (reference date). Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:

When you have finished reviewing my changes, you may follow the instructions on the template below to fix any issues with the URLs.

This message was posted before February 2018. After February 2018, "External links modified" talk page sections are no longer generated or monitored by InternetArchiveBot. No special action is required regarding these talk page notices, other than regular verification using the archive tool instructions below. Editors have permission to delete these "External links modified" talk page sections if they want to de-clutter talk pages, but see the RfC before doing mass systematic removals. This message is updated dynamically through the template {{source check}} (last update: 18 January 2022).

  • If you have discovered URLs which were erroneously considered dead by the bot, you can report them with this tool.
  • If you found an error with any archives or the URLs themselves, you can fix them with this tool.

Cheers.—InternetArchiveBot (Report bug) 04:43, 25 December 2016 (UTC)[reply]

External links modified[edit]

Hello fellow Wikipedians,

I have just modified 2 external links on Epoch (reference date). Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:

When you have finished reviewing my changes, you may follow the instructions on the template below to fix any issues with the URLs.

This message was posted before February 2018. After February 2018, "External links modified" talk page sections are no longer generated or monitored by InternetArchiveBot. No special action is required regarding these talk page notices, other than regular verification using the archive tool instructions below. Editors have permission to delete these "External links modified" talk page sections if they want to de-clutter talk pages, but see the RfC before doing mass systematic removals. This message is updated dynamically through the template {{source check}} (last update: 18 January 2022).

  • If you have discovered URLs which were erroneously considered dead by the bot, you can report them with this tool.
  • If you found an error with any archives or the URLs themselves, you can fix them with this tool.

Cheers.—InternetArchiveBot (Report bug) 22:28, 26 July 2017 (UTC)[reply]

Requested move 20 February 2019[edit]

The following is a closed discussion of a requested move. Please do not modify it. Subsequent comments should be made in a new section on the talk page. Editors desiring to contest the closing decision should consider a move review after discussing it on the closer's talk page. No further edits should be made to this section.

The result of the move request was: moved to alternative (closed by non-admin page mover) SITH (talk) 13:14, 28 February 2019 (UTC)[reply]


Epoch (reference date) → Epoch (date reference) – Should the undiscussed page move from Epoch (reference date) to Epoch (date reference) by RTG (talk · contribs) be accepted or reversed? Jc3s5h (talk) 03:42, 18 February 2019 (UTC) Iffy★Chat -- 14:21, 20 February 2019 (UTC)[reply]

Format of discussion converted from RFC to RM by me. Iffy★Chat -- 14:21, 20 February 2019 (UTC)[reply]
Hi, I disambiguated the topics to tidy them up. The search terms "epoch" +"reference date gets 21,000+ hits. +"date reference" is almost 46,000, and just slightly less ambiguous. Google books +"date reference" 6,650 hits +"reference date" 1,810. Now I'm not a calendar specialist or anything. I didn't read a book for this, I just learned what I needed for the purpose of disambiguating the lead sections. All the other topics epoch titles are defined by their field rather than "reference date", i.e. Epoch (astronomy), and they would all lay claim to "reference date" though some mean an instant and others an entire age. "Date reference" seems to be calendar specific. So it's less ambiguous on Epoch, therefore better dissemination. That's probably all I can think of about it o/ ~ R.T.G 04:16, 18 February 2019 (UTC)[reply]
When dealing with such massive numbers of ghits, there is no way to know if the terms mean the terms mean the same thing in the context in which they occurred. Plus the way they were searched for merely means that the word "epoch" occurred in the same document as either "date reference" or "reference date". In mind, an epoch is a reference date because it is a date that other dates are referred to. A date reference, on the other hand, is the mention of a date within a passage of text. Jc3s5h (talk) 14:46, 18 February 2019 (UTC)[reply]
An epoch is a reference date in all circumstances, however, the usage is widely different, sometimes meaning a moment, and at other times meaning an age. I suggest writing an article for Epoch (reference date) specifically toward the topic of the variations of usage, as the title demands, and letting each variation specifically described have its own article, as the content has been expanded to the stage where that is relevant. As it was, only three of the topics were covered, though there are at least six major interpretations. The hatnote pointed to Epoch (disambiguation) whereas the title is simply Epoch (suggesting the issue had not been deeply examined). The part about computing was certainly an article in its own right. Saying "No way to know" merely provides us nothing to go on, rather than information to guide us, so to consider that statement a waypoint does more for feeling than dissemination. There was a very technical and unintuitive feel to the article as it was. What got me started was the statement that epochs are decided by authoritative fiat. Fiat is authoritative by nature and means making a decision without true reasoning. I do not think that applies to the consideration of any epoch outside computing, by its very nature. An Epoch is a decisive division. It screamed to be wrenched apart. I have consoled it as best I could. ~ R.T.G 15:14, 18 February 2019 (UTC)[reply]
To quote a recent statement on the matter, " there is no way to know if the terms mean the terms mean the same thing in the context in which they occurred", which was my point exactly when editing the article. It definitely needed split owing to computing, and it definitely appeared as though structure was not considered, or the wider topic gone into. A person reading this article as it was may have believed only three major interpretations existed, whereas there are at least six. Searching up the terms does not provide us detail on every entry, but finding the two terms together at least provides us coincidence. Do you not believe the individual topics merit and provide enough material, already written, for separate articles? ~ R.T.G 15:26, 18 February 2019 (UTC)[reply]
I agree with removing the redundant material that is also covered by "Epoch (computing)", "Epoch (astronomy)", and "Epoch (geology)", although it might be better to have a paragraph explaining the scope of these articles rather than just putting them in the "See also" section. I do not believe this article should be titled "Epoch (date reference)" because an epoch is not a date reference, it is (in some senses) a reference date. Jc3s5h (talk) 17:19, 18 February 2019 (UTC)[reply]
No, epoch is not a date reference. But this article has become about the use of it as such with the other material moved. Maybe it should be moved to Epoch (calendar), and a further lead section written on the page Epoch beginning something like, "An epoch is a moment or period in time used as a reference point for dates and times in various fields of study."? ~ R.T.G 17:49, 18 February 2019 (UTC)[reply]
If you'll forgive a pun, I cannot see why an article couldn't be written on the article Epoch, specifically about the epochs of the words use in various fields of study. But there is no guarantee of finding suitable resource material. A good lead explaining the disambiguation is definitely possible though and would necessitate use of the term "reference date". ~ R.T.G 18:03, 18 February 2019 (UTC)[reply]
"Epoch" is a disambiguation page, and contains links to any article a reader might be thinking of when they type "Epoch" in the search box, including a DC Comics supervillain. Before your changes, this article was about any situation where time was counted from a certain starting point. I believe that should still be the scope of this article, but uses of the term where there is enough material for a full article should just be referred to briefly—example: "Epoch (computing)". Jc3s5h (talk) 20:38, 18 February 2019 (UTC)[reply]
Indeed. Giving a primary meaning for a term on a dab page generally requires the term to redirect to that topic and then the dab page to actually take the form Title (disambiguation). No great urgency to return to that arrangement, but something to bear in mind if you think this really is the primary meaning. I would like to suggest an alternative, perhaps less contentious, title. For example, the article itself describes the meaning as being in the field of chronology, so possibly Epoch (chronology). The term doesn't have to be widely used, so long as the disambiguator is clear to any readers about where they are going to end up. Or something simpler like Epoch (time), or even just Epoch (date) (currently a redirect to here), if we want this to be a generic description of the word epoch meaning a point or period in time. Again, the disambiguator doesn't have to be a complete dictionary definition, just an unambiguous and clear way to distinguish from other similar titles. Lithopsian (talk) 20:56, 18 February 2019 (UTC)[reply]
Well I am responsible for attempting this without discussion, and both your ideas seem reasonable to me, so if you can decide on a preferred structure, I will offer to attempt to finish up the work until it seems to be in order and natural, unless there are further objections. ~ R.T.G 21:25, 18 February 2019 (UTC)[reply]
Why is this an WP:RFC and not an WP:RM? Iffy★Chat -- 13:43, 20 February 2019 (UTC)[reply]
I am not familiar with that process. If you are, please feel free to make any necessary adjustments. Jc3s5h (talk) 14:05, 20 February 2019 (UTC)[reply]
Done. RMs resolve quicker than RFCs do, so I've moved the page back to the old title while the discussion is in progress. Iffy★Chat -- 14:21, 20 February 2019 (UTC)[reply]
The above discussion is preserved as an archive of a requested move. Please do not modify it. Subsequent comments should be made in a new section on this talk page or in a move review. No further edits should be made to this section.
Retrieved from "https://en.wikipedia.org/w/index.php?title=Talk:Epoch&oldid=1194750299"