RE: [ubl] Minutes of Atlantic UBL TC call 9 November 2005

[Grimley Michael J NPRI <GrimleyMJ@Npt.NUWC.Navy.Mil>]

> Of course the TC will have to explain at some point why it reserves the right to use redefines but does not give it to others...

Keeping in mind that redefine requires both schemas to be in the same namespace, I hope customizers understand that their customizations shouldn't reside in the UBL namespace. The only reason we are now able to use xsd:redefine is that the group agreed that the namespace name of a minor version will be the same as its associated major version.


-----Original Message-----
From: Eduardo Gutentag [mailto:eduardo.gutentag@sun.com] 
Sent: Monday, 21 November 2005 1433
To: Grimley Michael J NPRI
Cc: ubl@lists.oasis-open.org
Subject: Re: [ubl] Minutes of Atlantic UBL TC call 9 November 2005



On 11/21/2005 06:06 AM, Grimley Michael J NPRI wrote:
> All,
> 
> Thank you, Eduardo, for the input. Although a continuation of the dialog would be very welcome, I realize that you may not have the time, so I address this to the entire group:
> 
>> I see from the minutes that the TC has decided to go for redefine. I 
>> don't want to reopen the long discussion that we had on this in 
>> London (when Mike H. hosted our meeting), at which point we decided 
>> not to go that route mostly because of software concerns on the one hand and because redefine does not leave traces of it having been used in the document instances, only in the schemas, on the other.
> 
> I believe the issue about the instance was in relation to customization rather than our own minor versioning. 

The original understanding was that we were going to do our own minor versioning as customization. In my mind there is still no difference between them ;) though of course it seems like one has now been introduced. If redefine is going to be used silently only in the production of minor versions, with no implication that customization should be also done through redefines, then all my comments are invalid and do not apply.

Of course the TC will have to explain at some point why it reserves the right to use redefines but does not give it to others...

> At the time, redefine was being considered for customization and I had pointed out that it would require customizers to work within the UBL namespace and, therefore, there would be no way to determine what was a customization and was 'real' UBL. In the case of minor versioning it is all still 'real' UBL.
> 
> 
>> a) the Customization Guide we wrote for v1.0 should be deleted for 
>> 2.0 and replaced with something completely new. Nothing in the old one applies now, so it would be a disservice to UBL users to try to re-use it.
> 
> I'm not quite sure why this is so. Can someone please point out which parts of it would no longer be applicable?
> 
> 
>> The use of redefine completely eliminates the possibility of using a 
>> previous schema version to validate a new version document, as it provides none of the guarantees that XSD extensions and restrictions do.
> 
> First, in the examples given so far, we have limited our redefines to using extension:
> 
>   <xsd:redefine schemaLocation="UBL-CommonAggregateComponents-1.0.xsd">
>     <xsd:complexType name="PartyType">
>       <xsd:complexContent>
>         <xsd:extension base="PartyType">
>           <xsd:sequence>
>             <xsd:element ref="cbc:SomeNewElement" minOccurs="0"/>
>           </xsd:sequence>
>         </xsd:extension>
>       </xsd:complexContent>
>     </xsd:complexType>
>   </xsd:redefine>
> 
> Does this not give us the same guarantees? Also, the issue of what we mean by 'backwards compatibility' has reared its ugly head again. Just to clarify, we have agreed with Eduardo's example of 18 October:
> 
> "In other words, if all documents that validate against MySchema v1.0 also validate against MySchema v1.1, then MySchema v1.1 is said to be backwards compatible; but there is no expectation that any document that validates against MySchema v1.1 must also by necessity validate against MySchema v1.0"
> 
> Any assistance will be appreciated as we continue to evaluate the efficacy of redefine.
> 
> Thank You,
> Mike

-- 
Eduardo Gutentag               |         e-mail: eduardo.gutentag@Sun.COM
Corporate Standards            |         Phone:  +1 510 550 4616 (internal x31442)
Sun Microsystems Inc.          |         W3C AC Rep / W3C AB / OASIS BoD