On Tue, Feb 22, 2011 at 3:05 AM, Edward C. Zimmermann <[log in to unmask]> wrote:
> On Mon, 21 Feb 2011 16:31:27 -0500, Bruce D'Arcus wrote
>> To me, seasons are important, but indicating hemisphere is a level of
>> complexity I don't need. So I don't care if this isn't in the spec.
>> That said, I would just point out that such a thing would not need
>> anything so specific as location or country: it would only need
>> something that says "north" or "south."
> Not really. We have several competing semantics for seasonal names
> (Meteorlogical, Astronomical, Cultural, Historical..). Its, for example,
> right now Spring in Ireland (1 Feb) but Winter in England (where 1 March is
> traditionally used to mark the start of Spring).
> A university press, however, might use the predicates Fall, Winter and
> Spring to refer to academic quarters. At the University of California, for
> example, Spring starts on 28 March. Some campuses on the semester system use
> Fall and Spring to denote the two semesters but others use Winter and Spring
> (quite common in Germany) and other still ...
> In the Jewish calendar we have a distinction driven by a Lunar cyle. The
> prayer for rain starts on Shemini Atzeret/Simchat Torah--- the 8th day of
> the Fall harvest festival of Succot--- and ends on Passover. Tu B担hevat,
> the 15th of Shevat (this year it was 20 Jan), marks the the beginning of the
> New Year for trees--- or what we could call Spring.
> North and South only are relevant to determine what season staddeles two
> years but not when the season starts or ends.
Fair enough; which is why I think we shouldn't go there. If some use
case needs that more complex information, there are other ways to
encode it outside an EDTF value.
If there's no consensus that agrees with my position (and I think
Ray's), then drop it altogether.