[tei-council] <when>/@absolute and dateTime values

[Martin Holmes]

That makes sense, and I'll make the relevant changes. This text, though:

"If it is likely that the value used is to be compared with another,
then a time zone indicator should always be included, and only the
dateTime representation should be used."

needs to be changed a bit to make it clear that what we mean here is 
comparison with an external datetime source, because in a TEI 
<timeline>, as shown in the example below, the @absolute value is 
arguably always being compared with other internal values such as those 
specified in @interval.

Cheers,
Martin

On 12-12-31 07:52 PM, Syd Bauman wrote:
> I think your analysis is pretty correct, although I reach a slightly
> different conclusion.
>
> * The "BST" is leftover from P4, when it was specified in the
>    absolute=. I don't know how it got dropped. Seems to me the value
>    of absolute= in that example in TS should have been 12:20:01+01:00.
>
> * While you're right, and in many cases the note in data.temporal.w3c
>    is right, in many (if not most) cases where we can imagine
>    <timeline> being used, dates (and for that matter time zones) are
>    simply irrelevant. In many (if not most) cases, although the times
>    might be compared, they're only going to be compared to other
>    <when> in the same timeline, or other <timeline>s in the same
>    corpus. Thus requiring a date (IMHO) would be silly.
>
> So my first reaction is that think we should drop the "BST", and
> re-iterate the warning that if you expect the values in your timeline
> to be compared to other values out there in the world, you should
> include a date and timezone.
>
>> This bit of Chapter 8:
>>
>> <http://www.tei-c.org/release/doc/tei-p5-doc/en/html/TS.html#TSSAPA>
>>
>> contains this example:
>>
>> <timeline unit="s" origin="#TS-P1">
>>    <when xml:id="TS-P1" absolute="12:20:01"/>
>>    <when xml:id="TS-P2" interval="4.5" since="#TS-P1"/>
>>    <when xml:id="TS-P6"/>
>>    <when xml:id="TS-P3" interval="1.5" since="#TS-P6"/>
>> </timeline>
>>
>> with the following commentary:
>>
>> "... TS-P1 is located absolutely, at 12:20:01:01 BST. TS-P2 is 4.5
>> seconds later than TS-P2 (i.e. at 12:20:46)..."
>>
>> I was first of all puzzled by the assertion that this is BST (British
>> Summer Time). Nothing in the example suggests that there is any
>> timezone offset from UTC, which I assumed was the default. Then I
>> looked at the datatype for @absolute, which is data.temporal.w3c,
>> about which we say in a note:
>>
>> "If it is likely that the value used is to be compared with another,
>> then a time zone indicator should always be included, and only the
>> dateTime representation should be used."
>>
>> <http://www.tei-c.org/release/doc/tei-p5-doc/en/html/ref-data.temporal.w3c.html>
>>
>> So in this context, the xsd:dateTime representation should be used,
>> and it should include a timezone. So I looked at the W3C spec:
>>
>> <http://www.w3.org/TR/xmlschema-2/#dateTime>
>>
>> which, as I read it, requires the presence of a date; whereas our
>> example has only a time.
>>
>> It seems to me, therefore, that this usage of @absolute is wrong on
>> two counts: first, it should include a date, and second, it ought to
>> have a timezone offset of 'Z', the canonical representation for an
>> offset of zero, i.e. UTC. Further, I think the claim that this value
>> is BST makes no sense (BST is UTC+1) given that the example has no
>> timezone offset in it.
>>
>> Am I missing anything here?