GNU bug report logs - #20583
calendar-absolute-from-gregorian

Previous Next

To add a comment to this bug, you must first unarchive it, by sending
a message to control AT debbugs.gnu.org, with unarchive 20583 in the body.
You can then email your comments to 20583 AT debbugs.gnu.org in the normal way.

Toggle the display of automated, internal messages from the tracker.

Message #5 received at submit <at> debbugs.gnu.org (full text, mbox):

From: Nicholas Strauss <nicholas.strauss <at> gmail.com>
To: bug-gnu-emacs <at> gnu.org
Cc: gjs <at> ai.mit.edu
Subject: calendar-absolute-from-gregorian
Date: Thu, 14 May 2015 21:20:21 -0700

Hi Bug Emacs,

Why does (calendar-absolute-from-gregorian '(10 10 1582))
return 577731?
(calendar-generate '(10 1582)) is another example of this
error.

There are no Gregorian dates from 5 - 15 October 1582.
Pope Gregory declared this.

Hence, ALL absolute day numbers in calendar before 1582
are probably wrong. So using emacs for historical computations
is bogus.

I will work on writing calendar-absolute-from-gregorian and
calendar-gregorian-from-absolute to patch this error.

Let me know if this sounds OK.

nick
ncs <at> alum.mit.edu

Message #8 received at 20583 <at> debbugs.gnu.org (full text, mbox):

From: Glenn Morris <rgm <at> gnu.org>
To: Nicholas Strauss <nicholas.strauss <at> gmail.com>
Cc: 20583 <at> debbugs.gnu.org, gjs <at> ai.mit.edu
Subject: Re: bug#20583: calendar-absolute-from-gregorian
Date: Fri, 15 May 2015 20:54:04 -0400

Nicholas Strauss wrote:

> Why does (calendar-absolute-from-gregorian '(10 10 1582))
> return 577731?
> (calendar-generate '(10 1582)) is another example of this
> error.

It's behaving as designed. It's a utility function that operates on the
basis that today's calendar system is valid for all time. It's not meant
as literally historically accurate, and I think trying to make it so (by
introducing discontinuities at certain dates) would be a mistake. See eg
thread at:

http://lists.gnu.org/archive/html/emacs-devel/2006-07/msg01008.html

Message #13 received at 20583 <at> debbugs.gnu.org (full text, mbox):

From: Nicholas Strauss <nicholas.strauss <at> gmail.com>
To: Glenn Morris <rgm <at> gnu.org>
Cc: 20583 <at> debbugs.gnu.org, gjs <at> ai.mit.edu
Subject: Re: bug#20583: calendar-absolute-from-gregorian
Date: Fri, 15 May 2015 18:48:02 -0700

Hi Glenn,

The real point here is that Calendar Systems are instruments and as
subject are subject to error (Think like an engineer!). The real world
is messy -- Pope Gregory abolished some days, and various countries
took their time to jump on the band wagon.

I think what Reingold is/was trying to do was to use "absolute" as an
internal reference format -- a primary key -- that would correlate all the
other calendars.

there are other "absolute" choices for internal references.
Peter Duffett-Smith took a different approach -- he counts 2415020
days from January 1, 4713 BC until the 1900 epoch, January 0.5, 1900.

This crops up in sqlite3 calculations for date() and also common
javascript formulas.

I am still researching this.

If our civilization lasts for any appreciable time, having a consistent
calendar will be crucial for keeping a consistent history.
I think this helps why I am stressing the astronomy. If an Aztec calendar
records an eclipse on a certain day, that's pretty much it.

Nick


On Fri, May 15, 2015 at 5:54 PM, Glenn Morris <rgm <at> gnu.org> wrote:
> Nicholas Strauss wrote:
>
>> Why does (calendar-absolute-from-gregorian '(10 10 1582))
>> return 577731?
>> (calendar-generate '(10 1582)) is another example of this
>> error.
>
> It's behaving as designed. It's a utility function that operates on the
> basis that today's calendar system is valid for all time. It's not meant
> as literally historically accurate, and I think trying to make it so (by
> introducing discontinuities at certain dates) would be a mistake. See eg
> thread at:
>
> http://lists.gnu.org/archive/html/emacs-devel/2006-07/msg01008.html

Message #16 received at 20583 <at> debbugs.gnu.org (full text, mbox):

From: Nicholas Strauss <nicholas.strauss <at> gmail.com>
To: Glenn Morris <rgm <at> gnu.org>
Cc: 20583 <at> debbugs.gnu.org, gjs <at> ai.mit.edu
Subject: Re: bug#20583: calendar-absolute-from-gregorian
Date: Fri, 22 May 2015 08:36:43 -0700

[Message part 1 (text/plain, inline)]

There are other day number systems more common than the  emacs
"absolute" day number --
the "astro" or Julian day number and the proleptic Gregorian system
used by sqlite3
e.g. select julianday("-4713-11-24").
cal-julian.el:alendar-astro-to-absolute refers to  "astro".
Proleptic Gregorian day 0 = November 24, 4714 BC while "astro" day 0 =
 January 1, 4713 BC.
I'm attaching julian.lisp which has calendar-correlate-from-date and
calendar-correlate-from-julianday
may help with days before 1582. These are based on Peter
Duffett-Smiths calculations.

[julian.lisp (application/octet-stream, attachment)]

Message #19 received at 20583 <at> debbugs.gnu.org (full text, mbox):

From: Nicholas Strauss <nicholas.strauss <at> gmail.com>
To: Glenn Morris <rgm <at> gnu.org>
Cc: 20583 <at> debbugs.gnu.org
Subject: Re: bug#20583: calendar-absolute-from-gregorian
Date: Sun, 24 May 2015 08:29:22 -0700

[Message part 1 (text/plain, inline)]

Corrected comments to calendar-julianday-from-proleptic-gregorian
and cond for year 1582.
File requires calendar.el

Nick

On Fri, May 22, 2015 at 8:36 AM, Nicholas Strauss
<nicholas.strauss <at> gmail.com> wrote:
> There are other day number systems more common than the  emacs
> "absolute" day number --
> the "astro" or Julian day number and the proleptic Gregorian system
> used by sqlite3
> e.g. select julianday("-4713-11-24").
> cal-julian.el:alendar-astro-to-absolute refers to  "astro".
> Proleptic Gregorian day 0 = November 24, 4714 BC while "astro" day 0 =
>  January 1, 4713 BC.
> I'm attaching julian.lisp which has calendar-correlate-from-date and
> calendar-correlate-from-julianday
> may help with days before 1582. These are based on Peter
> Duffett-Smiths calculations.

[julian.lisp (application/octet-stream, attachment)]

Previous Next

GNU bug tracking system
Copyright (C) 1999 Darren O. Benham, 1997,2003 nCipher Corporation Ltd, 1994-97 Ian Jackson.