Re: Duration format

On Wed, Nov 12, 2014 at 9:16 AM, Léonie Watson <LWatson@paciellogroup.com>
wrote:

> I have a question about the time format used in some vocabularies, for
> example the cookTime itemprop in the recipe vocabulary [1]. It requires a
> duration, which must use the ISO8641 format (HH:MM:SS).
>
>
> ["ISO 8641:2008
<http://www.iso.org/iso/catalogue_detail.htm?csnumber=38543> specifies the
required characteristics for metric self-locking nuts, with MJ thread, for
use in aerospace construction at a maximum temperature greater than 425
°C."]

I see Charles got the sensible answer out already, but I found this
amusing, and aerospace hasn't been a bundle of laughs recently.

It would possibly be better if schema:Duration was defined in terms of the
xml schema restrictions on ISO 8601;  xsd:dayTimeDuration, which excludes
months and years, and xsd:yearMonthDuration, which is restricted to only
years and months, have more useful properties.

This does bring up a problem with OWL 2. Although duration types were fixed
for XML Schema 1.1, and are thus added to  RDF 1.1, and although the OWL 2
working group was kept in being specifically to to await the publication of
XSD 1.1,  W3 process would not allow the now-usable types to be added. Bijan
Parsia was not impressed
<http://lists.w3.org/Archives/Public/public-owl-wg/2012May/0010.html>.
This is on the wish  list for WHO-WG <http://who-wg.github.io/>

[OWLAPI accepts  literals with these types, and some reasoners and
databases will work with them (e.g. Pellet/Stardog). Hermit checks all
datatype restrictions strictly. ]

 Whereas this makes complete sense from a programmatic point of view, it’s
> not particularly user friendly from a content point of view. To a human
> “00:15:00” doesn’t easily translate into 15 minutes.
>

To me, as a duration it translates as... 15 seconds
<https://www.google.com/search?q=smpte+time+code+display&tbm=isch> (tail
end of an SMTPE Timecode).

Simon