W3C home > Mailing lists > Public > public-p3p-spec@w3.org > July 2005

RE: Trouble with data schema

From: Giles Hogben <giles.hogben@jrc.it>
Date: Thu, 07 Jul 2005 20:00:39 +0200
To: "'Rigo Wenning'" <rigo@w3.org>, "'Lorrie Cranor'" <lorrie+@cs.cmu.edu>
Cc: "'public-p3p-spec'" <public-p3p-spec@w3.org>
Message-id: <000c01c5831d$c889c710$5f2abf8b@cs.jrc.it>
I'm inclined to think we should just stick with the backwards compatible
format XSD, but ban categories in future except by reference (as discussed).
I think converting all categories to data types in backward compatible way
is probably too much of a can of worms at this stage.

The other important point is the errors brought up yesterday by Mark - we
checked it with a schema validator and it didn't have any errors - can Mark
tell us which validation too he is using.

I attach the latest version in case the problem is that it's an older
version (hopefully that's it)...



-----Original Message-----
From: public-p3p-spec-request@w3.org [mailto:public-p3p-spec-request@w3.org]
On Behalf Of Rigo Wenning
Sent: Thursday, July 07, 2005 19:29
To: Lorrie Cranor
Cc: 'public-p3p-spec'; Giles Hogben
Subject: Re: Trouble with data schema


Didn't we agree, that a transform would be released as a WG Note? The 
problematic things is the backwards compatibility requirement. At the 
moment it requires the backwards transform. That's why it is referenced 
in the Specification and so desperately missing. 

Problem is: What implementations do we break if there is no transform? 
The transform makes the data format incredibly complicated / nearly 
unusable. So this is the key question. Only if we can provide a service 
for automatic transform to 1.0 dataschema, it all makes sense. This 
said, the necessary (and not the 'nice to have') transforms should be 
annexed to the Specification as long as we require both formats. 

The advantage put forward to use just plain XML Schema (tools, ease of 
use etc) slowly disappears here, if the new format is even more 
constrained as the old format and requires difficult operations before 
having a valid policy.

Best, 

Rigo

Am Thursday 07 July 2005 18:08 verlautbarte Lorrie Cranor :
> I think we can go to last call without the updated transforms (we 
> would need to document what's wrong with the existing transforms). We 
> would definitely need this fixed before going to PR, which we are 
> aiming for some time in September. What do others think?
>


Received on Thursday, 7 July 2005 18:01:23 GMT

This archive was generated by hypermail 2.2.0 + w3c-0.30 : Thursday, 7 July 2005 18:01:23 GMT