xml schema profile

[Rob Seaman]

Hi Arnold,

> And somewhere in that article is a link to an interesting (but  
> that's of course because it supports what I've been preaching ;-)  
> article on the use of defaults (Just say No): http://www.hr-xml.org/ 
> blog/?p=26

In all conscience I don't understand how this is possible.  The link  
points to a (somewhat tepid) example:

> <EducationalMeasure>
>      <MeasureSystem>GPA</MeasureSystem>
>      <MeasureValue>
>           <NumericValue>3.9</NumericValue>
>      </MeasureValue>
> </EducationalMeasure>
>
> The data that is missing from this xml is the default information.  
> In Xml Schema, the default values of data do not have to be  
> included in the xml instance. In this example, the default value  
> for the element ExcessiveValueIndicator is omitted. Here is the  
> full data piece:
>
> <EducationalMeasure>
>      <MeasureSystem>GPA</MeasureSystem>
>      <MeasureValue>
>           <NumericValue>3.9</NumericValue>
>      </MeasureValue>
>      <ExcessiveValueIndicator>false</ExcessiveValueIndicator>
> </EducationalMeasure>

Now I don't know what "ExcessiveValueIndicator" is – maybe something  
they flash on the screen of the Home Shopping Network – but there are  
any number of other bits of semantic DNA missing from this example.   
Whether you call them defaults or not, they simply aren't here.  For  
just one example, what numerical scale is used to represent  
particular letter grades?  One might infer that an A is a 4.0, but  
this is not stated.  Isn't that the same thing as a default?  An  
application has to have specific knowledge of the context of this  
data structure to be able to make productive use of it.  Software  
always has to have some backstory furnished to it to use any API, any  
external data format.

Perhaps I misunderstand the issue.  I'm probably not alone :–)

Rob
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.ivoa.net/pipermail/dm/attachments/20060921/7037ad2c/attachment-0003.html>