An Old Man Needs Help

[Jay R. Ashworth]

On Wed, Mar 14, 2007 at 03:52:14PM -0400, Kenneth Brody wrote:
> Quoting Jay R. Ashworth (Wed, 14 Mar 2007 13:41:45 -0500):
> > On Tue, Mar 13, 2007 at 10:08:41AM -0400, Boaz Bezborodko wrote:
> > > I'm still using 4.8 and what we do is create a field that contains a
> > > date derived from "12/31/50" - "the date being indexed".  The result
> > > is a field with a descending index on a date field.  Just be sure to
> > > update this field whenever the date is changed.
> >
> > And note that this depends on the setting of your PFCMARK.
> 
> Not really.  A sort on delta-T doesn't matter if it's relative to
> 1950 or 2050, as long as PFCMARK doesn't change.

I had to read that twice to be sure it was an answer to the problem I
*was* posing, and not one I wasn't.

And I'm *still* not sure.

12/31/1950 - 12/31/2006 = -56 years
12/31/1950 - 12/31/2007 = -57 years
12/31/2050 - 12/31/2006 =  44 years
12/31/2050 - 12/31/2007 =  43 years

Nope, Ken; I believe you're wrong.  If the date-to-subtract-from is
later than the dates you care about, then they go in opposite ...

Oh, of course.  This is an absolute-value problem.  They go in the same
direction, but it's not apparent because -57 is "larger" than -56 just
as 2007 is "larger" than 2006... but you're on the opposite side of the
number line, so you're still moving *to the left*.

Wow, that's esoteric.

Cheers,
-- jr 'once I figured it out, I figured it would be useful to post anyway' a
-- 
Jay R. Ashworth                                                jra at baylink.com
Designer                          Baylink                             RFC 2100
Ashworth & Associates        The Things I Think                        '87 e24
St Petersburg FL USA      http://baylink.pitas.com             +1 727 647 1274