RE: AGENDA: 23 June P3P Spec call

OK - here are some examples:
Note that any element from the base data schema will look the same in the
policy, but can be represented in parallel with an extension which can be
validated by XSD:

BSD Elements:
<DATA-GROUP>
	<EXTENSION optional="no">
		<DATATYPE>
			<DYNAMIC>
				<COOKIES/>
			</DYNAMIC>
		</DATATYPE>
	</EXTENSION>

	<DATA ref="#dynamic.cookies">
</DATA-GROUP>

These can then be parsed and validated by both old and new agents.
It can also be written with only the new format and then put through our
XSLT (Which will need some alteration to add the extension mechanism stuff
in) to produce the old format:



Non BDS Elements:

<DATA-GROUP>
	<EXTENSION optional="no">
		<DATATYPE>
			<X>
				<Y/>
			</X>
		</DATATYPE>
	</EXTENSION>

	<DATA ref="#x.y">
</DATA-GROUP>


In this case, this can be parsed, by both old and new agents but only the
Extension element can be validated against the custom XSD schema.



>**-----Original Message-----
>**From: Rigo Wenning [mailto:rigo@w3.org] 
>**Sent: 22 June 2004 11:51
>**To: public-p3p-spec@w3.org
>**Cc: Giles Hogben; massimo@w3.org; 'Lorrie Cranor'
>**Subject: Re: AGENDA: 23 June P3P Spec call
>**
>**
>**Giles, 
>**
>**can you come up with some example XML code for your 
>**assertions? I think we need more explanatory text if we want 
>**to convince people.
>**
>**Rigo
>**
>**Am Tuesday 22 June 2004 11:26 verlautbarte Giles Hogben :
>**> I think there is a slight misunderstanding of the extent of the 
>**> breaking of backward compatibility. It is only broken in a 
>**small way, 
>**> which should not affect anyone. This also does not break 
>**the backward 
>**> compatibility guidelines as the BDS is not part of the P3P 
>**1.0 Schema 
>**> anyway.
>**>
>**> 1. Old Base Data Schema data element syntax can still be used 2. 
>**> Extended elements can still be written in the old format.
>**>
>**> The only thing which breaks backward compatibility (with only some
>**> implementations) is that custom elements written with the 
>**new format 
>**> cannot be validated against the old style schema. They can 
>**still be 
>**> written in the old syntax, with a new style syntax alongside in an 
>**> extension element to provide validation where possible.
>**

Received on Tuesday, 22 June 2004 07:20:04 UTC