[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]
Subject: Re: [wsbpel] Issue 103 - Good Idea!
Clearly not: a non-existent node cannot be the destination. But if
anything this would only help the case for issue 11. My thinking goes like
this:
1) keeping WSDL 2.0 compliance in mind (and the issue 103 suggestions),
we want to align ourselves with the idea of the message being typed by a
schema type or element (this is implied by allowing usage like
$msgVar/part/foo/bar )
2) until 2.0 comes out we define a "BPEL implicit schema" for WSDL1.1
messages (we have to do this to allow queries like $msgVar/part/foo/bar)
3) we banish the notion of parts from bpel (eliminate part attribute from
bpel); to access a part we query the corresponding element in the implied
message schema (we can do this because we can apply xpaths to message
variables and the implied schema for message variables is set in stone by
the BPEL specification, per #2)
4) when WSDL 2.0 comes out there there is nothing to change; WSDL 2.0
messages will have the schema element/type indicated in the WSDL document,
WSDL 1.1 message will continue to have our "implied schema"
5) Blissful co-existence of WSDL versions is achieved.
The problem is this: how, starting with an empty message variable do we
insert the message parts? That is, how do we do:
<copy>
<to variable="foo" part="bar">
<from ....>
</copy>
if we have no "part" attribute? Answer seems to be that we need insert
operations.
-maciej
On Fri, 12 Mar 2004 10:54:40 -0800, Danny van der Rijn <dannyv@tibco.com>
wrote:
> maybe i just don't know XPATH well enough, but if $var/a/b/c doesn't
> exist,
> can it be the target of an assign?
>
> ----- Original Message -----
> From: "Maciej Szefler" <mbs@fivesight.com>
> To: "Danny van der Rijn" <dannyv@tibco.com>; "wsbpeltc"
> <wsbpel@lists.oasis-open.org>
> Sent: Friday, March 12, 2004 10:10 AM
> Subject: Re: [wsbpel] Issue 103 - Good Idea!
>
>
>> Can you elaborate? I don't the connection.
>>
>> -maciej
>>
>> On Fri, 12 Mar 2004 09:56:51 -0800, Danny van der Rijn
>> <dannyv@tibco.com>
>> wrote:
>>
>> > even so, if we add the ability to have assign create new nodes (issue
>> > 11),
>> > then this syntax wouldn't work.
>> >
>> > ----- Original Message -----
>> > From: "Maciej Szefler" <mbs@fivesight.com>
>> > To: "Danny van der Rijn" <dannyv@tibco.com>; "wsbpeltc"
>> > <wsbpel@lists.oasis-open.org>
>> > Sent: Friday, March 12, 2004 9:52 AM
>> > Subject: Re: [wsbpel] Issue 103 - Good Idea!
>> >
>> >
>> >> This might not be a problem if it were possible to define copy
>> >> assignment
>> >> in terms of node-replacement operations; then any expression that
>> >> returns
>> >> a node could act as an l-value to an assignment. Such a scheme could
>> >> rely
>> >> exclusively on schema-based type checking (with an implied schema for
>> >> message variables).
>> >> -maciej
>> >>
>> >>
>> >> On Fri, 12 Mar 2004 09:12:04 -0800, Danny van der Rijn
>> >> <dannyv@tibco.com>
>> >> wrote:
>> >>
>> >> > i wouldn't think so, since $var refers to the value, not the
> location.
>> >> > but
>> >> > being just a mere mortal, i am open to being corrected.
>> >> >
>> >> > ----- Original Message -----
>> >> > From: "Satish Thatte" <satisht@microsoft.com>
>> >> > To: "Maciej Szefler" <mbs@fivesight.com>; "Dieter Roller"
>> >> > <ROL@de.ibm.com>
>> >> > Cc: "Alex Yiu" <alex.yiu@oracle.com>; "Assaf Arkin"
>> >> <arkin@intalio.com>;
>> >> > "wsbpeltc" <wsbpel@lists.oasis-open.org>; <ygoland@bea.com>
>> >> > Sent: Friday, March 12, 2004 8:52 AM
>> >> > Subject: RE: [wsbpel] Issue 103 - Good Idea!
>> >> >
>> >> >
>> >> > I imagine that expression= is as relevant to "to" as to "from" ..
>> >> >
>> >> > -----Original Message-----
>> >> > From: Maciej Szefler [mailto:mbs@fivesight.com]
>> >> > Sent: Friday, March 12, 2004 8:47 AM
>> >> > To: Dieter Roller; Satish Thatte
>> >> > Cc: Alex Yiu; Assaf Arkin; wsbpeltc; ygoland@bea.com
>> >> > Subject: Re: [wsbpel] Issue 103 - Good Idea!
>> >> >
>> >> > I haven't followed this discussion too closely, but am I correct in
>> >> > inferring that if these suggestion were adopted, we could eliminate
>> >> most
>> >> >
>> >> > of the from-specs in the assign activity? Could we normalize
>> >> assignment
>> >> > to
>> >> > the following
>> >> >
>> >> > <copy>
>> >> > <to variable="msgVar"/>
>> >> > <from expression="$otherMsgVar"/>
>> >> > </copy>
>> >> >
>> >> > could be used to assign a message variable.
>> >> >
>> >> > <copy>
>> >> > <to variable="simpleTypeVar"/>
>> >> > <from expression="$msgVar/msg/simplePart"/>
>> >> > </copy>
>> >> >
>> >> > <copy>
>> >> > <to variable="simpleTypeVar"/>
>> >> > <from expression="$otherSimpleTypeVar"/>
>> >> > </copy>
>> >> >
>> >> > could be used to assign a simple variable
>> >> >
>> >> > <copy>
>> >> > <to variable="elementVar"/>
>> >> > <from expression="$msgVar/msg/elementPartType"/>
>> >> > </copy>
>> >> > <copy>
>> >> > <to variable="elementVar"/>
>> >> > <from expression="$otherElementVar"/>
>> >> > </copy>
>> >> >
>> >> > could be used to assign an element variable.
>> >> >
>> >> > -maciej
>> >> >
>> >> >
>> >> >
>> >> >
>> >> > On Fri, 12 Mar 2004 07:23:35 +0100, Dieter Roller <ROL@de.ibm.com>
>> >> > wrote:
>> >> >
>> >> >>
>> >> >>
>> >> >>
>> >> >>
>> >> >> +1
>> >> >>
>> >> >> Cheers,
>> >> >>
>> >> >> dieter
>> >> >>
>> >> >>
>> >> >>
>> >> >>
>> >> >>
>> >> >>
>> >> >>
>> >> >>
>> >> >> "Satish Thatte"
>> >> >> <satisht@microsof
>> >> >> t.com>
>> >> >
>> >> >> To
>> >> >> "Assaf Arkin"
>> >> > <arkin@intalio.com>,
>> >> >> 03/12/2004 05:14 "Alex Yiu"
>> >> > <alex.yiu@oracle.com>
>> >> >> AM
>> >> >
>> >> >> cc
>> >> >> "wsbpeltc"
>> >> >>
> <wsbpel@lists.oasis-open.org>,
>> >> >> <ygoland@bea.com>
>> >> >>
>> >> > Subject
>> >> >> RE: [wsbpel] Issue 103 -
>> Good
>> >> >> Idea!
>> >> >>
>> >> >>
>> >> >>
>> >> >>
>> >> >>
>> >> >>
>> >> >>
>> >> >>
>> >> >>
>> >> >>
>> >> >> I am having trouble keeping up with this fast moving discussion.
>> I
>> >> am
>> >> >> hoping that you will reach an agreement and then educate the mere
>> >> >> mortals among us on what the consensus proposal is ..
>> >> >>
>> >> >> Satish
>> >> >>
>> >> >>
>> >> >>
>> >> >> -----Original Message-----
>> >> >> From: Assaf Arkin [mailto:arkin@intalio.com]
>> >> >> Sent: Thursday, March 11, 2004 7:59 PM
>> >> >> To: Alex Yiu
>> >> >> Cc: wsbpeltc; Satish Thatte; ygoland@bea.com
>> >> >> Subject: Re: [wsbpel] Issue 103 - Good Idea!
>> >> >>
>> >> >>
>> >> >>> (4)
>> >> >>> Assaf suggested:
>> >> >>> In WSDL 2.0,
>> >> >>> $variable/ns:element[/ns:subElement]
>> >> >>> In WSDL 1.1,
>> >> >>> $variable/partName/ns:element
>> >> >>>
>> >> >>> I was wondering whether it make sense to add a WSDL QNAME
>> >> >>> (ns:wsdlMsgName) like the following for WSDL 1.1:
>> >> >>>
>> >> >>> $variable/ns:wsdlMsgName/partName/...
>> >> >>>
>> >> >>> then the syntax would be more symmetrically between WSDL 1.1 and
> 2.0
>> >> >>> ns:wsdlMsgName => ns:element
>> >> >>> partname => subElement
>> >> >>>
>> >> >>> The BPEL code migration may be easier from WSDL 1.1 to 2.0
>> >> >>
>> >> >> If anyone has a good handle on where WSDL 2.0 is heading with
>> their
>> >> >> message definition, would be great to throw some ideas around.
>> >> Ideally
>> >> >> if you have a WSDL 2.0 interface that's backward compatible with
> WSDL
>> >> >> 1.1, you could use the BPEL process with both 1.1 and 2.0 without
>> >> >> change.
>> >> >>
>> >> >> Assaf
>> >> >>
>> >> >>
>> >> >>
>> >> >> To unsubscribe from this mailing list (and be removed from the
> roster
>> >> > of
>> >> >> the OASIS TC), go to
>> >> >>
>> >> >
>> >>
> http://www.oasis-open.org/apps/org/workgroup/wsbpel/members/leave_workgr
>> >> > oup.php
>> >> >> .
>> >> >>
>> >> >>
>> >> >>
>> >> >>
>> >> >> To unsubscribe from this mailing list (and be removed from the
> roster
>> >> > of
>> >> >> the OASIS TC), go to
>> >> >>
>> >> >
>> >>
> http://www.oasis-open.org/apps/org/workgroup/wsbpel/members/leave_workgr
>> >> > oup.php.
>> >> >>
>> >> >
>> >> >
>> >> >
>> >>
>> >>
>> >>
>> >> --
>> >> Maciej Szefler [mbs(a)fivesight.com] [+1-312-432-0556x226]
>> >>
>> >> To unsubscribe from this mailing list (and be removed from the roster
> of
>> > the OASIS TC), go to
>> >
> http://www.oasis-open.org/apps/org/workgroup/wsbpel/members/leave_workgroup.php.
>> >>
>> >
>> >
>> > To unsubscribe from this mailing list (and be removed from the roster
>> of
>> > the OASIS TC), go to
>> >
> http://www.oasis-open.org/apps/org/workgroup/wsbpel/members/leave_workgroup.php.
>> >
>>
>>
>>
>> --
>> Maciej Szefler [mbs(a)fivesight.com] [+1-312-432-0556x226]
>>
>> To unsubscribe from this mailing list (and be removed from the roster of
> the OASIS TC), go to
> http://www.oasis-open.org/apps/org/workgroup/wsbpel/members/leave_workgroup.php.
>>
>
>
> To unsubscribe from this mailing list (and be removed from the roster of
> the OASIS TC), go to
> http://www.oasis-open.org/apps/org/workgroup/wsbpel/members/leave_workgroup.php.
>
--
Maciej Szefler [mbs(a)fivesight.com] [+1-312-432-0556x226]
[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]