Print

Print


I had suggested that we replace "open" (in an interval)with double-dot,
e.g.
20040101/..

And Ed suggests instead:   20040101/

Does anyone see a problem with Ed's suggestion?

--Ray

> -----Original Message-----
> From: Discussion of the Developing Date/Time Standards
> [mailto:[log in to unmask]] On Behalf Of Edward C. Zimmermann
> Sent: Tuesday, March 15, 2011 2:34 PM
> To: [log in to unmask]
> Subject: Re: [DATETIME] before/after revisited
> 
> My first reaction: Yes... but I'm not quite convinced on the
> appropriateness of the use of ".." for open in an interval using the /
> notation.
> 
> Instead of 20040101/..
> and        ../20040101
> 
> why not
>  20040101/
> and
>  /20040101
> 
> Even if we adopt // for comments it still works  20040101///After
> /20040101//Before
> 
> Is there any reason to need ".." or any other symbols ?
> It seems (at least this moment) linguistically redundant..
> 
> On Tue, 15 Mar 2011 14:06:03 -0400, Denenberg, Ray wrote
> > I think I've changed my view on this (in the past few hours), I've
> > thought
> it through with hopefully more clarity than before.
> >
> > I DON'T think we need to justify the before/after feature, because I
> > think
> it fits naturally with our syntax (as a few of you have hinted).
> >
> > We represent list using comma and double-dot, enclosing the list with
> square or curly braces to mean "one of" or "all of",
> > so
> > [1667,1668, 1670..1672]
> > means one of the years 1667, 1668, 1670, 1671, 1672
> >
> > and
> > {1667,1668, 1670..1672}
> > means all of the years.
> >
> > More simply,
> >
> > [1670..1672]
> > means one of the years  1670, 1671, 1672
> >
> > and
> > {1670..1672}
> > means all three years.
> >
> > It seems a natural extension that
> >
> > [..1672]
> > would mean some year preceeding (or equal to) 1672
> >
> > and
> > {..1672}
> > all years preceeding or equal to 1672.
> >
> > And that
> >
> > [1672..]
> > means some year following (or equal to) 1672
> >
> > and
> > {1672..}
> > All years including or following 1672
> >
> > And I would go further to suggest that we replace "open" with
> > double-dot,
> so
> >
> > 20040101/..
> > means the interval beginning January 1 2004 with no end date
> >
> > I have more to say about intervals but will save it for a later
> message.
> >
> > --Ray
> 
> 
> --
> 
> Edward C. Zimmermann, NONMONOTONIC LAB
> http://www.nonmonotonic.net
> Umsatz-St-ID: DE130492967