DomainKey-Signature: a=rsa-sha1; q=dns; c=nofws; s=s1024; d=yahoo.com;
	h=Message-ID:X-YMail-OSG:Received:X-Mailer:Date:From:Subject:To:In-Reply-To:MIME-Version:Content-Type:Content-Transfer-Encoding;
	b=4JPYOADHH7PePkdp/3sSANKuWLae7iDajgl+sEP0UgVx/mcwSgGGeYvKSi4SkgTxr/y4xWgVfBIgEcA2WMzkaCStrsXGb3bpxyi7UQeoGPRmgySpBdA7izqQiehH8hUCgZg9ZmO2LrDC6ZlZADlRpMy/4goPu+1JOvuHlFfKeOI=;
Message-ID: <489318.10800.qm@web110704.mail.gq1.yahoo.com>
Date: Wed, 29 Dec 2010 09:38:36 -0800 (PST)
From: Leslie S Satenstein <lsatenstein@yahoo.com>
Subject: Re: Some comments about Julian Dates and possible bug. Please provide
	feedback.
To: Robert Haas <robertmhaas@gmail.com>,
	Postgres documentation <pgsql-docs@postgresql.org>
In-Reply-To: <AANLkTi=2kVZCqEVH1vDFoqt0-vRZApnmg1MHy7Z6e4=U@mail.gmail.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=iso-8859-1
Content-Transfer-Encoding: quoted-printable

Good morning Robert=0A=0AI presume that you are the acting project manager,=
 in the sense of vetting changes or commentary.  The pdf guide has only one=
 statement concerning julian dates.  I feel that more is required.=0A=0AJul=
ian dates are used as follows.=0A=0AWorking days between two dates is (the =
difference of the day numbers * 5/7) , with minor adjustment for rounding a=
nd weekend start-end considerations. =0A=0AJulian date conversion is used f=
or this type of calculation.  It would be nice to see more in the manual ab=
out Julian dates, such as tests if a day is a weekend day, or to determine =
the day of the week for a specific Gregorian calendar date and vice-versa.=
=0A=0ARegarding the julian bug(let), I will raise it as a comment on the co=
de buglist.=0A=0A =0A=0A=0A------------------=0A=0ARegards  =0A=A0Leslie=0A=
 Mr. Leslie Satenstein=0A=0A=A0=0Amailto:lsatenstein@yahoo.com=0Amailto les=
lie.satenstein@itbms.biz / leslies@itbms.biz=0Awww.itbms.biz=0A =0A=0A=0A--=
- On Wed, 12/29/10, Robert Haas <robertmhaas@gmail.com> wrote:=0A=0A> From:=
 Robert Haas <robertmhaas@gmail.com>=0A> Subject: Re: [DOCS] Some comments =
about Julian Dates and possible bug. Please provide feedback.=0A> To: "Lesl=
ie S Satenstein" <lsatenstein@yahoo.com>=0A> Cc: "pgsql-docs" <pgsql-docs@p=
ostgresql.org>=0A> Date: Wednesday, December 29, 2010, 6:58 AM=0A> On Mon, =
Dec 27, 2010 at 9:23 PM,=0A> Leslie S Satenstein=0A> <lsatenstein@yahoo.com=
>=0A> wrote:=0A> > I found the Julian date code that is programmed in=0A> P=
ostgres, to be accurate and fast except for one situation.=0A> But, I do ha=
ve one or two questions. 1) Which calendar is=0A> being used?=0A> >=0A> > I=
n the Gregorian Calendar, =A0January 1, 0001 is lowest=0A> positive date.=
=0A> > The day before 1/1/0001, according to the Gregorian=0A> calendar is =
December 31,-0001 in which the year has a value=0A> of negative one --- the=
re is no zero year in the Gregorian=0A> Calendar. (zero year is an error)=
=0A> >=0A> > I tested and found the algorithm in Postgres to have=0A> the d=
ay before January 1,0001 calculating as December=0A> 31,0000 even though th=
e world calculates the day before=0A> January 1,0001 as December 31,-0001.=
=0A> >=0A> > 2) Is the algorithm in Postgres correct? =A0I think=0A> not, a=
s the calculations for the difference in days between=0A> January 1, 0001 a=
nd December 31,-0001 is not 367 days, but=0A> just the value 1.=0A> >=0A> >=
 To convert the code to work with the Gregorian=0A> calendar takes two fixe=
s to two sub-routines. Each fix is=0A> two lines of C code.=0A> >=0A> > I h=
ave tested the PG Date C language routines=0A> with/without my fix, startin=
g with the year around -4713 to=0A> several centuries into the future. As l=
ong as both versions=0A> calculate Julian dates that are later than 1/1/000=
1, both=0A> the PG and my fixed versions respecting Gregorian algorithms=0A=
> produce identical results.=0A> >=0A> > 3) Does PG want to fully follow th=
e Gregorian Calendar=0A> rule? =A0If so,=0A> > 4) do they want my one patch=
 with two fixes6=0A> =0A> This seems like it'd be more appropriate for pgsq=
l-bugs or=0A> pgsql-hackers rather than here.=A0 I can't really figure=0A> =
out exactly=0A> what change you're proposing, so I'm not entirely sure=0A> =
whether it's=0A> right or wrong.=A0 Perhaps you could post your patch,=0A> =
and some examples.=0A> =0A> -- =0A> Robert Haas=0A> EnterpriseDB: http://ww=
w.enterprisedb.com=0A> The Enterprise PostgreSQL Company=0A> =0A> -- =0A> S=
ent via pgsql-docs mailing list (pgsql-docs@postgresql.org)=0A> To make cha=
nges to your subscription:=0A> http://www.postgresql.org/mailpref/pgsql-doc=
s=0A>