Re: [xml-dev] ANN: a portable data component -- length

[Liam R E Quin <liam@w3.org>]

On Sat, 2011-04-09 at 15:55 -0400, John Cowan wrote:
> Liam R E Quin scripsit:
> 
> > (I'd use metric/SI units in a program as they're easier to deal with,
> > and mixed-based ambiguity like 3 feet 7 inches doesn't occur)
> 
> If you want to represent measurements, as distinct from counts, you need
> to provide the measurement in the units actually measured.

+1 if they are actually being measured; one also has to give tolerance.

On Pete Cordell's point, the name of the element - I used <measurement>
rather than <distance> for no good or bad reason - one can't know a
priori whether the fact it's a measurement is more or less important
than the fact it's a linear measurement through space, a distance. One
is not intrinsically more "semantic" than the other.  But we can know
that there's a common danger in using element names for field names that
come from some non-XML source, as you might end up with non-XML
chracters to deal with, or markup in the field name -- e.g. mathematical
notation, or annotations (Japanese ruby comes to mind).

Someone else said that Normal Form is no gold standard - it isn't, but
the principle of not duplicating data still holds.  If Costelloitis
takes hold, you should at least mark which of two values is primary and
which was derived.

<jug id="jug">
  <capacity units="imperial_UK">4 gallons</capacity>
  <capacity units="US" derivedfrom="imperial_UK">4.5 gallons</capacity>
  <size>little</size>
  <colour>brown</colour>
</jug>


-- 
Liam Quin - XML Activity Lead, W3C, http://www.w3.org/People/Quin/
Pictures from old books: http://www.fromoldbooks.org/