On 25/04/13 11:41, Torsten Schassan wrote:
> I've been using the temporal from/to (respectively
> notBefore/notAfter) in this very same way: If I know the year, I
> supplied the same value.
> On the other hand I think that the use of two attributes seem to
> signify a range rather than a point in time (or space, as in
> biblScope). In this sense I really appreciate the @when on
> date/origDate as this is the place to indicate the point.
> (Appreciation especially as the processing of datasets is
> slightly easier.)
> Thus I recently switched in practice from @from/to to @when in
> order to express the point in time.
> (Some value combinations might offer room for interpretation though:
> Does <date from="1900-01-01" to="1900-12-31"> equal <date
> when="1900">? This kind of problems should not appear with pages?)
I agree with what Kevin has posted later.
In addition I tend to muse theoretically on the view of this as
points vs ranges slightly. (But only slightly!) When I say:
when="1900" I'm being fairly imprecise if I mean a point inside
1900 that I don't know specifically when it is. (As that is what
@notBefore/@notAfter is for after all.) No, when I say that I
mean 'for the entirety of 1900' so it is a range, not a point.
Because all dating attributes are about 'time' really this will
always be the case. When I say from="1999-03-17" to="1999-03-17"
this is not a single point in time, it is a range of a day. Even
if I extend this to from="1999-03-17T13:31:25Z" and
to="1999-03-17T13:31:25Z" then it is still a 'range' of a whole
second. Is this different if I had used @when since we are
limited in W3C to single second granularity? I both cases I'm
saying from the beginning of that second to the end of that
second. But we tend to treat these things like points because it
is convenient for the way we talk about them.
I could argue the same is true of page numbers. When I say
<citedRange from="34" to="34">p. 34</citedRange> am I really
saying from the 'beginning of page 34 to the end of page 34'?
Kevin's example of complex page layout makes the case even
better. If we have 4 columns on page 34 (a,b,c,d) and are citing
columns b and c then from="34b" to="34c" makes sense to me. If
it is just column be then from="34b" to="34b" likewise seems to
Kevin suggests continuing the discussion on the same ticket, and
that also makes sense. (My preference for separate tickets stems
from wanting to keep topics that are truly separate as separate
objects to be dealt with.)
Dr James Cummings, [log in to unmask]
Academic IT Services, University of Oxford