My take on this discussion is that there will be some users (e.g. Bruce)
happy with nothing more than 'yyyy-ss' for example '2000-21'; and other
users (e.g. Ed) who will want an additional designator, e.g. '2000-21-x',
and that we just need to nail down the designators, and designate one as the
default so that the default form (without the designator) can be used. Do
I don't agree that it is necessarily true that "2000 is not before or after
2000-10". Bruce, in his system, is free to design his sorting algorithm such
that the value of 2000, for purposes of sorting, is the first day of that
period, or the last day, i.e. either 2000-01-01 or 2000-12-31. Of course,
the standard is not going to dictate any such rules but any individual
system's sorting algorithm certainly would be free to adopt whatever rules
> -----Original Message-----
> From: Discussion of the Developing Date/Time Standards
> [mailto:[log in to unmask]] On Behalf Of Edward C. Zimmermann
> Sent: Friday, November 19, 2010 8:49 AM
> To: [log in to unmask]
> Subject: Re: [DATETIME] seasons
> On Fri, 19 Nov 2010 08:21:35 -0500, Bruce D'Arcus wrote
> > But ...
> > > Its all about precision:
> > > 2000 is not before or after 2000-10
> > > 2000-10 is not before or after 2000-10-12
> > > 2000-24 would, similarly, not be before or after 2000-10-12.
> > > 2000-24 would also not be before or after 2000-10-11
> > Yes, but in a bibliography, you can mix those dates (an author with
> > three cited items: a book with a year, and two journal articles: one
> > of them with a month cover date, and the other with a season cover
> > date), and you need to sort them. So saying something is "not before
> > or after" isn't practical.
> Hugh? No No. That's how one defines a sort.
> I'm pointing out how one compares items of different precision here..
> In my date class which I use in my engine I have all kinds of precision
> down to seconds and in the other direction decade (e.g. 1950s) and
> century. The year 1852, for example, belongs to the 1850s and also to
> the 19th century.
> 1852 is not before or after the 19th century
> 1852 is not before or after the 1950s
> 1853 is not before or after the 19th century
> 1853 is not before or after the 1950s
> 1852 < (is before) 1853
> in C++ with a class called, for example, xDATE
> friend int operator < (const xDATE &dt1, const xDATE &dt2);
> friend int operator <= (const xDATE &dt1, const xDATE &dt2);
> friend int operator > (const xDATE &dt1, const xDATE &dt2);
> friend int operator >= (const xDATE &dt1, const xDATE &dt2);
> friend int operator == (const xDATE &dt1, const xDATE &dt2);
> friend int operator != (const xDATE &dt1, const xDATE &dt2);
> from which a simple sort can be derived.
> My implementation currently does not have quarter or seasons but once
> we develop a standard for encoding them I'd naturally extend things..
> For precision--- in my implementation--- I think I'd probably take a
> pragmatic approach and align all the types as equivalent precisions
> given the feeling that they tend to be used in vaguer manners than the
> definition sets. This would let me handle them as quarters with some
> fuzziness.. But.. I have to give it some more thought.
> Edward C. Zimmermann, NONMONOTONIC LAB
> Basis Systeme netzwerk, Munich Ges. des buergerl. Rechts Office Leo
> Leopoldstrasse 53-55, D-80802 Munich,
> Federal Republic of Germany
> Umsatz-St-ID: DE130492967