Well I just noticed in B.1.4 of 8601 (the "examples" section) the example:
"a time interval starting at 12 April 1985 and ending on June 25, 1985:
It seems only reasonable to infer from the inclusion of this example that it
is intended to be allowed. Even though the examples are non-normative, I'm
tired of playing interpretation with 8601 and I take the position that given
such a glaring ambiguity we can intepret it as we please. So I am now
inclined to agree with Ed's interpretation, and to cast this as a level 0
And, it is interesting to note that the year is omitted on the end
component of the interval, implying that it is the same as the start year.
I would deduce from this that
is legal, and means the same as
> -----Original Message-----
> From: Discussion of the Developing Date/Time Standards
> [mailto:[log in to unmask]] On Behalf Of Ray Denenberg
> Sent: Monday, June 13, 2011 2:37 PM
> To: [log in to unmask]
> Subject: [DATETIME] Interval - compiance with 8601
> It seems there is a discrepancy about whether it is legal in 8601 for
> the endpoints (start and end) of an interval to be other than "complete
> representation". For example year/year , year-month/year-month, even
> year-month-day/year-month-day would be less that complete
> representation because the time is not included. Complete represention
> has to be year-month-day-time/year-month-day-time.
> By a literal interpretation of 8601 this seems to be true, Ed
> conjectures that it is an unintentional oversight. I don't think it
> matters much as I'm sure we all agree that there is no reason to
> disallow these "incomplete representations" in our spec.
> I suggest that we designate these as level 1 features. That is, they
> are not part of the formal 8601 profile (level 0) but rather part of
> the first level extensions.
> > -----Original Message-----
> > From: Discussion of the Developing Date/Time Standards
> > [mailto:[log in to unmask]] On Behalf Of Edward C. Zimmermann
> > Sent: Tuesday, June 07, 2011 2:44 PM
> > To: [log in to unmask]
> > Subject: Re: [DATETIME] A three level suggestion // ISO 8601
> > Hermeneutics
> > On Tue, 7 Jun 2011 12:44:13 -0400, Ray Denenberg, Library of Congress
> > wrote
> > > From: Edward C. Zimmermann
> > > > I only have a copy of the 2000 draft.. but..
> > > >
> > > > 126.96.36.199 defines these truncated dates.
> > > > 5.5.1 gives pattern a) as an interval: start/end where start and
> > end
> > > > are dates.
> > >
> > > Ed - It appears to me that ISO 8601-2004 (the current version) is
> > > freely available, since it is easily googled, for example:
> > >
> > > http://dotat.at/tmp/ISO_8601-2004_E.pdf
> > In that version
> > 4.4.1 here specifies start and send with / as a).
> > In the verbage of 188.8.131.52 it asks for expressions compliant to 4.3.2
> > Again, I think, this was an unintentional logical oversight in the
> > 184.108.40.206 defines the representations with reduced acuracy. There is no
> > reason, I think, to disallow these. Note that they explicitly allow
> > for reduced representations for durations so I don't see why they
> > >
> > > And my printed copy seems to be the same as this version. It
> > > doesn't have any of the sections you cite; there is a section 5 but
> > > it is a half-page with no subsections. Could you have a look at
> > > newer
> > one
> > > and determine if what you saw in the older version is there?
> > >
> > > Thanks.
> > >
> > > --Ray
> > --
> > Edward C. Zimmermann, NONMONOTONIC LAB http://www.nonmonotonic.net
> > Umsatz-St-ID: DE130492967