Minutes from 2008-02-07 call

["Michael Rowley" <mrowley@bea.com>]

 

SCA-BPEL Weekly meeting has been modified by Mr. Sanjay Patil

 

Date:  Thursday, 07 February 2008

Time:  08:00am - 09:00am PT

 

SUMMARY:

Opened Issue 18

 

Resolved Issue 2: Does the spec allow a componentType side file

Accepted text:

A BPEL process (which is identified by a QName) can be matched with any ComponentType document that has an <implementation.bpel> that refers to the same process QName, wherever it might be within the contribution. For example, if a componentType document has an <implementation.bpel process="foo:bar">, it will be considered a match for a BPEL process identified by "foo:bar" QName (where the prefix "foo" in both documents are resolved to the same namespace).  While the assembly specification allows <implementation.bpel> to be absent, for SCA BPEL, the component type MUST include an <implementation.bpel> element.  There MUST be only one ComponentType file corresponding to a BPEL QName within a contribution.

 

DETAILS:

 

Agenda:

1. Roll Call

   http://www.oasis-open.org/apps/org/workgroup/sca-bpel/members/roster.php

 

2. Appointment of scribe: Michael Rowley

 

3. Agenda bashing

- Moved issue 1 earlier

 

4. Approval of Jan 31, 2008 meeting minutes

   http://lists.oasis-open.org/archives/sca-bpel/200801/msg00042.html

 

Approved without objection.

 

5. Action items review

   http://www.oasis-open.org/apps/org/workgroup/sca-bpel/members/action_items.php

  

   #0021 chairs to follow up with Mary to post the CD docs to OASIS repository

Not done

 

   #0022 chairs to confirm the May f2f date with Jane

Confirmed as May 1 & 2 (Thur & Fri)

 

   #0023 Khanderao: file a new issue for re-writing specs with RFC-2119

Done.

 

   #0024 Martin: to write up a proposal for issue 15 (Define Conformance Targets)

Not Done.

 

6. New Issues

 

New issue: http://lists.oasis-open.org/archives/sca-bpel/200801/msg00048.html

Anish moves, Ashok seconds, to open issue 18: Need to rewrite the SCA-BPEL specifications with RFC-2119 keywords/statements

No objections.

 

7. Issue Discussion

 

Issue 1 http://www.osoa.org/jira/browse/BPEL-1

Deferred

 

Issue 2 http://www.osoa.org/jira/browse/BPEL-2

TITLE: Does the spec allow a componentType side file

SUBMITTED BY: Anish Karmarkar

 

Michael Rowley: Rowley recalls that Mike E didn't agree with ability to override information from the component type side file.

Michael Rowley: Anish recalls that Mike E also objected to having bpel-specific information in the component type

anish: http://lists.oasis-open.org/archives/sca-bpel/200801/msg00006.html

anish: another email from MikeR: http://lists.oasis-open.org/archives/sca-bpel/200801/msg00031.html

Sanjay: In section 2, line 194 add --

It is also possible to have a component type file that is associated

with the BPEL component implementation. When such a component type file

is present, as specified in [SCA-Assembly], it provides component type

information in addition to what is found by introspection.

Michael Rowley: SCA finds the ComponentType

For a BPEL process named "foo:bar" by finding any ComponentType document

that has <implementation.bpel qname="foo:bar">, wherever it might be

within the contribution.

Sanjay: Related assembly issue: http://osoa.org/jira/browse/ASSEMBLY-36

Michael Rowley: We should concentrate on how to find the component type side file for a process, since assembly-36 will handle what can be in it.

Michael Rowley: Unanimous agreement.

Michael Rowley: Anish believes that runtimes will not guarantee that there is only one component type file for a process, even if we require it with a "MUST".

Michael Rowley: ... also noted that this problem is true for everywhere SCA resolves QNames.

Michael Rowley: ... Isn't worried about having BPEL specific information in that side file.

Michael Rowley: ... Suggests allowing them to be anywhere, but perhaps we can also have a means of pointing to one explicitly.

Michael Rowley: ... But there again, that could be done for all QName resolution.

Sanjay: Related assembly issue for artifact resolution: http://osoa.org/jira/browse/ASSEMBLY-8

Michael Rowley: ... On a previous call, there was a suggestion that there may be a default location (same URI but with .componentType extension) and that the component type that is there doesn't have to have <implementation.bpel>

Michael Rowley: ... Anish is suggesting that we could maintain the requirement that it always have implementation.bpel?

Michael Rowley: Sanjay - what would be the value of the uri in that case?

Michael Rowley: Anish - even if it is in the default location, you still need the implementation.bpel with the right QName.

ichael Rowley: SCA finds the ComponentType for a BPEL process named "foo:bar" by finding any ComponentType document that has <implementation.bpel process="foo:bar">, wherever it might be within the contribution.  While the assembly specification allows <implementation.bpel> to be absent, for SCA BPEL, the component type MUST include an <implementation.bpel> element.

Michael Rowley: Suggested amendment:

Michael Rowley: If <implementation.bpel> is not present, then the component type has to have a contribution URI that is identical to the contribution URI except that the extension has to be changed to ".componentType".

anish: here is what I would prefer: 1) A componentType document for a BPEL process definition can be found by using either "process QName matching" or "contribution URI matching" mechanisms.

 

If a componentType document cannot be found through either "process QName matching" or "contribution URI matching" mechanism, SCA infrastructure MUST implicitly generate a componentType document based on the introspection of the implementation BPEL process definition.

 

For "process QName matching", a BPEL process (which is identified by a QName) can be matched with any ComponentType document that has an <implementation.bpel> refers to the same process QName, wherever it might be within the contribution. For example, if a componentType document has an <implementation.bpel process="foo:bar">, it will be considered a match for a BPEL process identified by "foo:bar" QName (where the prefix "foo" in both documents are resolved to the same namespace.).

 

For "contribution URI matching", a BPEL process can be matched with a componentType document of which the contribution URI is identical to the contribution URI of the BPEL process except of that the extension is changed to ".componentType". For example, if "file:/scratch/dir1/compositeX/Foo.bpel" is the contribution URI of a BPEL process, then the contribution URI of a matching componentType document would be: "file:/scratch/dir1/compositeX/Foo.componentType"

 

If a non-one-to-one relationship is found during matching (e.g. one-to-many or many-to-one), SCA Infrastructure SHOULD signal an error to users.

anish: here is what I would prefer 1) CA finds the ComponentType for a BPEL process named "foo:bar" by finding any ComponentType document that has <implementation.bpel process="foo:bar">, wherever it might be within the contribution.  While the assembly specification allows <implementation.bpel> to be absent, for SCA BPEL, the component type MUST include an <implementation.bpel> element.

anish: 2) there MUST be only one CT file corresponding to a BPEL QName within a contribution

Michael Rowley: A BPEL process (which is identified by a QName) can be matched with any ComponentType document that has an <implementation.bpel> refers to the same process QName, wherever it might be within the contribution. For example, if a componentType document has an <implementation.bpel process="foo:bar">, it will be considered a match for a BPEL process identified by "foo:bar" QName (where the prefix "foo" in both documents are resolved to the same namespace.).

Michael Rowley: While the assembly specification allows <implementation.bpel> to be absent, for SCA BPEL, the component type MUST include an <implementation.bpel> element.

Michael Rowley: there MUST be only one CT file corresponding to a BPEL QName within a contribution

Michael Rowley: Rowley moves that the previous three entries be used to resolve issue 2.

Michael Rowley: Charlton seconds.

Michael Rowley: No objections -- issue 2 is resolved.

Michael Rowley: Meeting adjourned.

 

-----------------------------------------

Rotating scribe list:

* Genov, Mr. Genadi; SAP AG

* Guizar, Alejandro; Red Hat

* Kinner, Mr. Jason; Oracle Corporation

* Levinson, Rich; Oracle Corporation

* Little, Dr. Mark; Red Hat

* Mischkinsky, Jeff; Oracle Corporation

* Andrabi, Mr. Najeeb; TIBCO Software Inc. (10/4/07)

* Barreto, Dr. Charlton; Adobe Systems (10/11/07)

* Pellegrini, Mr Michael; Active Endpoints, Inc. (10/18/07)

* Durand, Mr Jacques; Fujitsu Limited (11/1/07)

* Ford, Mr Mark; Active Endpoints, Inc. (11/8/07)

* Moser, Mr Simon; IBM (11/15/07)

* Malhotra, Mr Ashok; Oracle Corporation (11/29/07)

* van der Rijn, Danny; TIBCO Software Inc. (12/06/07)

* Trickovic, Mrs Ivana; SAP AG (1/27/08)

* Chapman, Dr Martin; Oracle Corporation (9/7/07, 12/13/07)

* Koenig, Mr Dieter; IBM (10/25/07, 1/10/08)

* Kand, Khanderao; Oracle Corporation (9/20/07, 1/31/08)

* Rowley, Mr Michael; BEA Systems, Inc. (9/27/07, 2/7/08)