W3C home > Mailing lists > Public > www-xml-schema-comments@w3.org > July to September 2000

Re: Confused by recurring-duration

From: James Clark <jjc@jclark.com>
Date: Thu, 14 Sep 2000 13:02:19 +0700
Message-ID: <39C069EB.F42D4A88@jclark.com>
To: Ashok Malhotra/Watson/IBM <petsa@us.ibm.com>
CC: XML Schema Comments <www-xml-schema-comments@w3.org>
That doesn't seem to address my question of what precisely the value
space of recurringDuration is.

Ashok Malhotra/Watson/IBM wrote:
> 
> The recurringDuration datatype is intended to represent durations of time
> which
> recur with some frequency.  We use it to generate a number of the other
> types
> and that is its main usage.  By itself, it can be used to represent
> arbitrary durations that
> recur.  For example, the first week of every month is a duration of seven
> days that
> recurs every month starting at the first midnight of some month.
> 
> All the best, Ashok
> 
> James Clark <jjc@jclark.com>@w3.org on 09/12/2000 05:01:40 AM
> 
> Sent by:  www-xml-schema-comments-request@w3.org
> 
> To:   XML Schema Comments <www-xml-schema-comments@w3.org>
> cc:
> Subject:  Confused by recurring-duration
> 
> I find the recurringDuration datatype a little confusing.  What exactly
> is the value space?  Specifically, is the value space of
> recurringDuration different from the value space of timeInstant? If I
> have a datatype like recurringDay, which is derived from
> recurringDuration, then presumably any value in the value-space of
> recurringDay needs also to be in the value-space of recurringDuration.
> It's not obvious to me how to define the value space of
> recurringDuration to make this the case.  It seems like the value space
> is the same as a timeInstant except that some fields can optionally be
> null/missing. But if that's the case, then not all values in the
> recurringDuration value space can be represented by literals in the
> lexical space.
> 
> I'm trying to figure out how one might add a recurring-duration datatype
> and associated operations to XPath. timeDuration makes perfect sense to
> me, as do all the datatypes derived from recurringDuration, but
> recurringDuration itself feels rather artificial.
Received on Thursday, 14 September 2000 02:02:52 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 7 January 2015 14:49:53 UTC