OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.

 


Help: OASIS Mailing Lists Help | MarkMail Help

sca-bindings message

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]


Subject: Re: [sca-bindings] ISSUE 55: WSDL 2.0


Eric Johnson wrote:
> Hmmm - I suppose I should put more explanation behind my response:
> 
> We've changed the specification such that the @wsdlElement attribute 
> MUST point at a WSDL 1.1 construct.  As a consequence, any way to 
> "extend" the binding.ws element to support WSDL 2.0 means either:
> 
>     * You've effectively side-stepped many of the existing elements and
>       attributes of binding.ws by defining alternates that apply to WSDL
>       2.0 - and you might as well have defined a new element.
>     * You've actually made the extension explicitly incompatible with
>       default binding.ws, an approach of dubious merit
>
Is it "incompatible" if something that was previously prohibited
becomes allowed?  Surely the essence of an upwards compatible
extension is that it makes documents that would not have been
valid/meaningful previously become valid/meaningful now, while
all documents that would have been valid/meaningful previously
continue to be valid and have the same meaning?

> Practically, I think that it does mean that any future binding that 
> supports WSDL 2.0 almost certainly is a new binding element.  That extra 
> phrase highlights that we do not think you can extend the binding.ws 
> element in a way that supports WSDL 2.0, and remain conformant to the spec.
> 
The above logic would seem to imply that any vendor could not
extend binding.ws to support WSDL 2.0 and remain conformant to
the spec.  I don't think this should be the case.

I believe it should be possible for the @wsdlElement attribute
to be extended to point to WSDL 2.0 artifacts either by vendors
now or by the binding.ws spec in the future, without becoming
"incompatible" with the present binding.ws specification.

   Simon

> -Eric.
> 
> Eric Johnson wrote:
>> Hi David,
>>
>> David Booz wrote:
>>>
>>> I may have missed some discussion. I was surprised to see this:
>>> "This specification only defines a binding using WSDL 1.1, but does 
>>> not prohibit a future specification defining a binding for WSDL 2.0."
>>>
>>> I was expecting to see this:
>>>
>>> This specification only defines a binding using WSDL 1.1, but does 
>>> not prohibit a vendor from extending the binding to support WSDL 2.0.
>>>
>>>
>>> The first statement doesn't really say anything since future specs 
>>> are free to change whatever they want anyway. I would have thought 
>>> that the important point to address was around concerns that the 
>>> binding was limited to 1.1 in the presence of a WSDL 2.0 spec, and 
>>> therefore we need to leave the door open for vendors to go there if 
>>> they want to.
>>>
>> The way I read the first statement is that we expect any future WSDL 
>> 2.0 binding to be a separate binding element, that is, not just an 
>> extension to the current binding.
>>
>> -Eric.
>> --------------------------------------------------------------------- 
>> To unsubscribe from this mail list, you must leave the OASIS TC that 
>> generates this mail. Follow this link to all your TCs in OASIS at: 
>> https://www.oasis-open.org/apps/org/workgroup/portal/my_workgroups.php 
> --------------------------------------------------------------------- To 
> unsubscribe from this mail list, you must leave the OASIS TC that 
> generates this mail. Follow this link to all your TCs in OASIS at: 
> https://www.oasis-open.org/apps/org/workgroup/portal/my_workgroups.php




[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]