W3C home > Mailing lists > Public > xmlschema-dev@w3.org > May 2001

RE: Time Zones and XML Schema date types

From: Ashok Malhotra <ashokma@microsoft.com>
Date: Tue, 8 May 2001 15:57:43 -0700
Message-ID: <9D884881F5E1F24FB845967851720FC31D1A76@red-msg-12.redmond.corp.microsoft.com>
To: "Jeff Rafter" <jeffrafter@definedweb.com>, <xmlschema-dev@w3.org>
Consider a date like May 8, 2001.  This is actually a period of 24 hours
(roughly).  The exact time the period starts and ends depends on the 
time zone.  So, I think, its clear why such types need a timezone.

What is more controversial is why we allow dates without a timezone.
This is to conform with SQL and also common usage.  Yes, it makes 
comparing dates with and without time zones complicated.

All the best, Ashok 
Ashok Malhotra              <mailto: ashokma@microsoft.com> 
Microsoft Corporation
212 Hessian Hills Road
Croton-On-Hudson, NY 10520 USA 
Redmond: 425-703-9462                New York: 914-271-6477 

-----Original Message-----
From: Jeff Rafter [mailto:jeffrafter@definedweb.com] 
Sent: Tuesday, May 08, 2001 1:53 PM
To: xmlschema-dev@w3.org
Subject: Time Zones and XML Schema date types

I am wondering why the spec allows time zone indicators for non time
such as date, gMonthYear, gYear... etc.  Though the examples don't
explicitly show dates I am assuming

1999-01-15 <> 1999-01-15Z

Because the starting instant of 1999-01-15 cannot be determined.  Is
wierd?  It seems as though the behavior that this represents should be
reserved for types which include time (not date only).  Therefore, to
accomplish the above one would have to use dateTimes (as in the ordering

1999-01-15T00:00:00 <> 1999-01-15T00:00:00Z

Can anyone shed some light on how the decision was reached?

Jeff Rafter
Defined Systems
XML Development and Developer Web Hosting
Received on Wednesday, 9 May 2001 01:42:24 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 7 January 2015 14:55:51 UTC