Re: A real world example: Dutch registry of buildings and addresses

--------------------------------------------
On Wed, 5/28/14, Frans Knibbe | Geodan <frans.knibbe@geodan.nl> wrote:

  the term accrualPolicy could be used to express the fact
 that existing data will not be changed when the dataset is
 updated.  In SQL terms, I would say that the database can be expected
 to change only by INSERT, not by UPDATE or DELETE. I don't
 know if if an expression exists that captures this kind of dataset. A
 comment for accrualPolicy states "Recommended best
  practice is to use a value from a controlled
 vocabulary". Could it be that such a vocabulary exists? Otherwise,
 the dublin  core wiki provides an example of using a text literal. I guess I
 could always  resort to that. 

===================

http://www.rustprivacy.org/2014/balance/gts/

I would say (while furiously waving my arms) that the "text literal" I used to increment
the date ...

$dogy['Quarterly']    = 1461/16; // Days per (harmonic) Quarter = 91.3125
$sdogy['Quarterly']   = "+91 days 7 hours 30 minutes"; // Days per (harmonic) Quarter

fits both wishes.  I compute the whole set of "report dates" in the range (4 years), some dates have past (archives) some are in the future (rdf:List).  Together, the points are a dct:Collection of dates which increment by a "text literal" ("+91 days 7 hours 30 minutes").  The list cannot be updated or deleted because neither the step size of the increment or the duration of the range may change.

If you add time to the increment you add time for every period (Quarter, Week, whatever) and the Year grows imaginary time.  If you do that too often you owe yourself an imaginary holiday on an imaginary beach :-)  If one defines "too hard" as any processing except recursion, then imaginary holidays are all you get.

Best,
Gannon

--Gannon

Received on Wednesday, 28 May 2014 22:32:10 UTC