Re: RE: XML Turing test

On 2/14/2023 9:56 AM, Roger L Costello wrote:
> Hi Folks,
> 
> An engineer mapped this XML form:
> 
> <Document>
>      <Pencil_Manufacturer>______</Pencil_Manufacturer>
> </Document>
> 
> to this XML form:
> 
> <Document>
>      <Umbrella_Manufacturer>______</Umbrella_Manufacturer>
> </Document>
> 
> That’s a bizarre mapping, right?
> 
> Let’s not be so quick to judge.
> 
> The allowable values of Pencil_Manufacturer are: Staedtler, Faber, and 
> Camlin.
> 
> The allowable values of Umbrella_Manufacturer are: Totes, Pogessi, and 
> Dynateck.
> 
> The engineer mapped the values as follows:
> 
> Staedtler --> Totes
> Faber --> Pogessi
> Camlin --> Dynateck
> 
> Question: Is that a correct mapping? Are the two forms equivalent?
> 
> It seems preposterous to even consider the two forms as equivalent. 
> After all, how can a document containing data about pencil manufacturers 
> be equivalent to a document containing data about umbrella manufacturers?
> 
> Possibly it’s not so preposterous.
> 
> What does it mean for two forms to be equivalent? Certainly they are not 
> equivalent with regard to string comparison:
> 
> “pencil-manufacterer” != “umbrella-manufacturer”
> 
> How about semantic equivalence? Intuitively we all know that pencils are 
> not the same as umbrellas.
> 
> And yet, the applications that process the two forms produce the same 
> output. In my example I said that both applications output 1, 2, 3, but 
> the output could be something far more complex, such as outputs that 
> control the flight of an aircraft.
> 
> If this form:
> 
> <Document>
>      <Pencil_Manufacturer>______</Pencil_Manufacturer>
> </Document>
> 
> and this form:
> 
> <Document>
>      <Umbrella_Manufacturer>______</Umbrella_Manufacturer>
> </Document>
> 
> are input into an aircraft’s Flight Management System (FMS) and both 
> result in the aircraft flying the same way, are the two forms equivalent?
> 
>  From the perspective of how they influence the application (aircraft 
> FMS) they are the same.
> 
>  From the perspective of semantics they are different.
> 
>  From the perspective of syntax they are different.
> 
> Conclusion: it doesn’t matter how you map one XML to another. If they 
> both elicit the same response in applications, then the mapping is 
> correct/equivalent. By definition.
> 
> Do you agree?

You keep mixing up names, strings, and semantics.  Usually what matters 
is the relationship between names and their values.  You seem to assume 
that your names carry a lot of meaning, but you don't say what that 
meaning is, nor the relationship between the names and string values.

If I have an element in an XML document whose value is "north", and 
another - maybe in another document - whose value is "nord", do you 
imagine you can feed those strings into a flight controller as is?  No, 
there must be some means to connect those values with a direction in a 
form that the controller can make use of.  Who cares if they are called 
a "pencil" or an "umbrella" or a "direction"?  It's the mapping to their 
meaning, the semantics, that matters.

For that matter, from your examples we can't tell if e.g., "north" means 
"fly at 0 deg true direction", "fly at 0 deg magnetic direction", "fly 
to some destination that is currently to our north", "you are currently 
flying without turning at 0 deg true", or something else.

So yes, if the relationships between the elements and values of two 
documents are the same, and if you have a mapping for each to the 
desired semantics of the consuming system, then fine, the two can be 
considered to be equivalent, *for that purpose*.

Things aren't usually specified that completely, though.


_______________________________________________________________________

XML-DEV is a publicly archived, unmoderated list hosted by OASIS
to support XML implementation and development. To minimize
spam in the archives, you must subscribe before posting.

[Un]Subscribe/change address: http://www.oasis-open.org/mlmanage/
Or unsubscribe: [email protected]
subscribe: [email protected]
List archive: http://lists.xml.org/archives/xml-dev/
List Guidelines: http://www.oasis-open.org/maillists/guidelines.php