Talk:Epoch/Archive 1

Expanding the Chronology section
Thanks Ike9898 for expanding the chronology section. I suggest expanding it further in NPOV spirit by adding the list of epoch dates for significant calendars in present use, which are currently hidden towards the end of the Common era article as Other calendars currently in use. What do you think? JFG 10:07, 2 January 2006 (UTC)
 * Sure, maybe one more example would be helpful. There are many examples in calendar era.  The problem that I have with the current article is that it seems to imply that epoch date means something different in human chronology vs. computing.  As far as I can tell it is exactly the same concept.  I think that there should be a general definition, before the two sections.  The sections can talk about examples and details in each area. Ike9898
 * Totally agree, it is indeed the same concept, but it was not clearly stated until we started working on those pages; don't they look good now? ;-) JFG 00:12, 4 January 2006 (UTC)

change name of article
I would prefer to name this article something like "epoch (computers)" to clearly differentiate it from things like epoch (astronomy). That is, astronomy epochs are dates that are used as a reference point, so it's not necessarily clear from this article title that the article doesn't discuss astronomy. --Interiot 18:27, 6 December 2005 (UTC)
 * The top of the page points to the epoch article for disambiguation. JFG 10:07, 2 January 2006 (UTC)
 * Now that you added a nice astronomy summary in here too, this disambig pointer can go away, along with the wiktionary pointer which was hidden behind "reference date". I have also simplified the epoch page to make it clearer (and added this wiktionary link where it belongs), tell me what you think. JFG 00:12, 4 January 2006 (UTC)
 * Yeah, it's definitely better than it was, but it still feels a bit lopsided since the chronology and astronomy sections are basically pointers to more substantial articles and computers isn't. Perhaps eventually computers will get its own article, and this article will have a a couple paragraphs of summary on each topic to tie everything together nicely (the alternatives being to revert to a simple disambig page (which isn't preferable because these topics really do have strong connections), or for it to become increasingly lopsided over time).  --Interiot 00:33, 4 January 2006 (UTC)
 * IMHO astronomy can remain a small summary here because the term used by astronomers is just epoch, not epoch date, which is why I listed it separately in the epoch disambig. However I'm not a native English speaker so please correct me if I'm wrong.
 * As Ike9898 noted, the concept of an epoch date is very similar in chronology and computing: defining an origin of time, instant 0 (or 1). In contrast, the astronomical epoch is a slightly different idea: picking a particular instant as a photograph of the Universe from which celestial positions at other instants can be computed.
 * As for the computer section being longer than the chronology section, it's just because I happen to know more about computers (and the original stub started with this meaning) -- I sure hope that people will expand the human side of the concept. JFG 01:54, 4 January 2006 (UTC)
 * Ultimately though, there's only a couple of uses of the word "epoch" that use them as instants in time as a reference point for other measurements, as compared to epoch as a large span of time. And each use has its wrinkles and competing reference points.  That's why I think astronomy goes here.
 * Somebody just improved the epoch page by explicitly classifying uses of the term to define moments in time vs periods of time. This ambiguity in the English language must have been the root of confusion about this term. Now "epoch date" can be seen as a specialized term for the "moment in time" acception of the "epoch" word, and it's clearer. -- JFG 14:33, 5 January 2006 (UTC)
 * No no, I didn't mean to imply that the computing section was remotely bad, it's quite good IMHO. I just think that at some point, it will be forced to split off as its own article, at which point people will ask why the current article can't just be merged with epoch, as they'll both be disambig-ish at that point.  I was arguing that they shouldn't be merged down the road, because this article serves to note the tie-ins between human chronology / computers / astronomy more than simple one-liners could, and that's a good thing.
 * As for "epoch" versus "epoch date", I'm not sure that "epoch date" is necessarily the perfect term. These google searches show that "epoch date" is rarely used:  --Interiot 02:40, 4 January 2006 (UTC)
 * {| border="1" cellspacing="0" cellpadding="2"

! ! epoch ! epoch date
 * Excel
 * 303,000
 * 230
 * Unix
 * 428,000
 * 535
 * GPS
 * 363,000
 * 506
 * astronomy
 * 637,000
 * 1,090
 * J2000.0
 * 76,000
 * 448
 * Gregorian
 * 88,000
 * 170
 * }
 * 448
 * Gregorian
 * 88,000
 * 170
 * }
 * }


 * Well, maybe we could move this article to Epoch (date), Epoch (instant in time), or something like that.  What say you?  ike9898 16:15, 5 January 2006 (UTC)


 * (added more to the table that makes the case even stronger) Yeah, it should probably be changed to Epoch (x).  Maybe Epoch (reference point) would work?  Or (instant in time) could work.  Anybody else have any thoughts?  --Interiot 17:11, 5 January 2006 (UTC)


 * These Googlefights look quite compelling, but I think they are misleading. Through recent refactoring, we have noticed that the English word epoch has the dominant meaning of a period and the secondary meaning of an instant. Consequently we have clearly separated those meanings on the epoch page and we have refined the epoch date page to encompass all the nuances of the epoch as an instant concept. The Google results for "epoch" include all meanings of the word, whereas "epoch date" narrows results radically to pages (such as this Wikipedia entry) that need to explicitly make the difference between the two main meanings of epoch. Therefore I vote to leave both pages as they are. In addition, just renaming "epoch date" to "epoch (date)" would seem less elegant to me. If we decide to change the name because "epoch date" is indeed rarely used, I would prefer "Epoch (reference date)". -- JFG 02:16, 6 January 2006 (UTC)
 * I don't follow. For instance, take a look at J2000.0.  It's not as if it's frequently used to refer to a span of time (or used by history bufs or geology bufs, or...).  Alternatively, take a look through the google results for J2000.0 epoch yourself.  It seems somewhat clear to me that epoch is frequently used on its own, and that Epoch (astronomy) is the best title, and not Epoch date (astronomy).  Since "epoch" is widely used to refer to both, and "epoch date" is rarely used (and somewhat misleading, since epoch (astronomy) refers to a precise instant in time, and not any span involving hours), I think the article should be renamed to "epoch (xx)".  --Interiot 02:30, 6 January 2006 (UTC)
 * I see your point. Now I'm ready to vote for renaming the page to "Epoch (reference date)" or some such. I was swayed by results of "What links here" that also point overwhelmingly to epoch, mostly from geology and astronomy articles; the few that point to epoch date can be easily changed. Then we'll have to take a serious look at bypassing the epoch disambig in many of those pages. In passing, I've just reordered the epoch meanings according to apparent usage frequency. -- JFG 03:05, 6 January 2006 (UTC)
 * I join JFG and support "Epoch (reference date). Interiot?  ike9898 15:02, 6 January 2006 (UTC)
 * "Epoc (reference date)" sounds great. "Reference instant" or "reference moment" are a little more precise, but don't imply all the things that "reference date" does IMHO.  --Interiot 15:48, 6 January 2006 (UTC)
 * Done! I've rephrased the article accordingly, but can you help with fixing the myriad pages that link here, expecially double redirects? and fixed the double redirects in referring articles. Glad we managed to reach consensus so fast on this thorny subject, thanks! -- JFG 17:04, 6 January 2006 (UTC)
 * Epoch (reference time)? Epoch (time reference)? ike9898 18:30, 5 January 2006 (UTC)

AD/CE
Per Manual_of_Style, we don't change back and forth between two acceptable styles. Also, "if in doubt, defer to the style used by the first major contributor". Since there does seem to be some contention over it, and since there haven't been substantial reasons put forth for why one should be prefered, the article should defer to the style first used. As far as I could tell, the reference was firsted added with this edit, and it was A.D. Please do not continue to change it unless substantial reasons are given. --Interiot 22:40, 14 May 2006 (UTC)

Computing section needs rollover years
It's already listing Unix, MS-DOS, and NFS but is not listing some of the other major ones in use, perhaps most noticably the Windows one that started in 1600. Jon 21:21, 22 August 2007 (UTC)

Exactly 2005 years?
Until I removed it a moment ago, the article contained this sentence:

Thus, the first instant of January 1, 2006 should be exactly 2005 years since the epoch.

I see several problems with this statement. We all agree the epoch was established by Dionysius Exiguus, but that does not mean that the first moment of the first of January, 1, is the epoch. The surviving works of Dionysius do make it clear he intended the Incarnation (Christianity) to be the epoch, but it isn't clear on what date he estimated that occurred. Although today the Incarnation is considered to be distinct from the Nativity of Jesus, some early scholars, such as Bede, didn't make the distinction. Also, we don't know if Dionysius used the dates the Catholic Church celebrated the Annunciation or Nativity, or if he had estimates of the real date(s) which did not survive.

In view of these problems, I think it is better to remove the statement. --Gerry Ashton (talk) 07:35, 28 January 2008 (UTC)

Hebrew Calendar stated incorrectly?
In the article it currently says the following:

"In Israel, the traditional Hebrew calendar, using an era dating from Creation, is the official calendar. However, the Gregorian calendar is the de facto calendar and is commonly used. Government documents usually display a dual date. The beginning of year 1 of the Hebrew calendar occurred in the autumn of 3761 BC. Therefore, "Rosh Hashanah, the Jewish New Year, in September 2003 marked the transition from 5763 to 5764"."

I think this may be wrong because the author either does not know or is forgetting that there is no year 0. From September 3761 BC to September 1 BC there are 3760 years. From September 1 BC to September 2003 AD there are 2003 years. Total years is 3760 + 2003 = 5763, not 5764. October of 2003 would be the first month of the 5763rd year given an epoch of 3761 B.C.

Its actually quite surprising how common this error is. You can find the same mistake even in highly scholarly works. Scaliger is famous for heaping insults (e.g., written in Latin "stultissimus") on the many famous historians and astronomers who made the same mistake. —Preceding unsigned comment added by 72.248.107.194 (talk) 17:58, 31 March 2008 (UTC)


 * The Astronomical Almanac for the year 2003, page B2, contains a table of the year number for various calendars, and what Gregorian date is considered the first day of the various years. It states that the Jewish (A.M.) year 5764 begins September 26, 2003, at sunset. The Explanatory Supplement to the Astronomical Almanac (1992) states on page 584 that the Era of Creation, or Era Mundi, corresponds to -3760 October 7 on the Julian proleptic calendar. The year -3760 in astronomical notation corresponds to -3761, so there does not seem to be any problem. --Gerry Ashton (talk) 20:24, 31 March 2008 (UTC)


 * Looking ath 72.248.107.194's reasoning further, the 1st year of the Hebrew calendar, which began in -3760. In the autumn of 2003, 5763 years were completed, thus it was the beginning of the 5764th year. --Gerry Ashton (talk) 20:37, 31 March 2008 (UTC)

Mac and Unix Epoch Clarification
The Unix-epoch should be clarified. The article says that the Unix epoch date is Jan. 1 1970, but 1904 on Macs. Is Mac OS X's epoch date 1970 because of its Unix base, or is 1904 like pre-OS X Macs? 68.174.116.251 05:56, 1 January 2006 (UTC)


 * Done, along with more clarifications, rewrite and fact-checking. JFG 05:41, 2 January 2006 (UTC)

In various Mac file formats (BinHex or Stuffit come to mind), the epoch date is Jan 1, 1984. —Preceding unsigned comment added by 92.235.217.205 (talk) 23:01, 28 March 2010 (UTC)

Geological epoch
I propose the complete removal of this section - the use of the word epoch in this context is quite different from that of the rest of the article. There are already various articles which cater better to the topic -see Geologic time scale or Epoch (geology) for example. Moreover this paragraph contains numerous inaccuracies in its present form. Geopersona (talk) 05:00, 22 September 2010 (UTC)
 * I have now removed it for the reasons outline above and added an appropriate DAB link at the top of the page. Geopersona (talk) 06:12, 1 December 2010 (UTC)

Microsoft SQL Server 2005 base date
Would appear to be 1900-01-01 as mentioned at http://msdn.microsoft.com/en-us/library/ms187819(v=SQL.90).aspx. You can verify this in the product itself by running select cast(0 as datetime).

This conflicts with the date of 1753-01-01 Microsoft SQL Server. I have not been able to find any reference to this date being the epoch date for the product or any evidence for the rationale which seems weak to me. Microsoft SQL Server (2005, 2008, 2008 R2) has a data type of datetime which is able to store dates as far back as 1753-01-01, but as mentioned on the link above the actual base date is 1900-01-01.

Ftlie (talk) 10:21, 21 October 2012 (UTC) 1753-01-01 is from Sybase (ie the code base used for first mssql) http://www.karaszi.com/SQLServer/info_datetime.asp#Why1753 Earlier dates than this would be disambigous anyway, they thought.

Mac OS X epoch rationale inaccurate
I have been informed that NSDate was introduced in 1994 with NeXT's "Enterprise Objects Framework" (which introduced the new Foundation framework that Mac OS X uses to this day). So, if it was added back then and used 2001 as the start of its epoch, how could it have been because of Mac OS X's release year? Uliwitness (talk) 18:05, 11 November 2014 (UTC)

Questionable anchors
In this edit User:Matthiaspaul added several anchors, including 0000-01-01 and 0001-01-01. Because these anchors appear to be in the ISO 8601 format, which always uses the Gregorian calendar, they create the implication that the dates January 1, 0 and January 1, 1 mentioned in the article are Gregorian calendar dates. But there are not sufficient citations in the article to establish whether the dates are Julian calendar (which would normally be used for such ancient dates) or Gregorian calendar. So the anchors should be reformatted or the table should be edited to state the dates are in the Gregorian calendar, together with sufficient citations to establish such a claim. Jc3s5h (talk) 14:21, 8 April 2015 (UTC)