RE: [sca-j] ISSUE 4 - Proposed spec text for dependency reinjection

From
Simon Nash <>
Date
2008-02-11T10:58:58+00:00
ID
Thread
RE: [sca-j] ISSUE 4 - Proposed spec text for dependency reinjection
Mike,

The JIRA comments record already includes
this statement about callback references and composite-scoped components.
 The JIRA record was referenced in the motion to resolve, so I don't
think any further formal action is needed to agree this aspect of the resolution.
 It does need to be added to the agreed text by the editors.

    Simon 

Simon C. Nash, IBM Distinguished Engineer

Member of the IBM Academy of Technology

Tel. +44-1962-815156  Fax +44-1962-818999

Mike Edwards/UK/IBM@IBMGB

10/02/2008 22:39

To

"OASIS Java" <>

cc

Subject

RE: [sca-j] ISSUE 4 - Proposed spec
text  for dependency reinjection

Folks, 

I'm ok with what Simon proposes to add. 

There is one final addition worthy of consideration.  I think we need
to state that callback references will never 

be injected or reinjected for composite scoped components.  This is
necessary since for such long-lived 

components the only way to get the callback reference is to use the RequestContext
API.  (The alternative 

would be the complete opposite of (re)injecting on every invocation of
an operation, which I don't advocate). 

Yours,  Mike.

Strategist - Emerging Technologies, SCA & SDO.

Co Chair OASIS SCA Assembly TC.

IBM Hursley Park, Mail Point 146, Winchester, SO21 2JN, Great Britain.

Phone & FAX: +44-1962-818014    Mobile: +44-7802-467431  

Email:   

Simon Nash/UK/IBM@IBMGB

09/02/2008 17:08

To

"OASIS Java" <>

cc

Subject

RE: [sca-j] ISSUE 4 - Proposed spec
text  for dependency reinjection

Unfortunately something was omitted from the resolution of this issue.
 In our call of 1/31/08, we discussed the email thread in

   http://www.oasis-open.org/apps/org/workgroup/sca-j/email/archives/200801/msg00055.html

We talked abut the various options and agreed (I thought) to adopt option
3.  Unfortunately this decision was not recorded in the minutes and
therefore did not appear in the formal final resolution of this issue.
 Here's the relevant section of the minutes for that discussion:

..... 

anish: topic: issue 4 

anish: simon: 2 things to talk about one is on the agenda and the other
isn't 

anish: ... we have to consider injection and reinjection for conversations

anish: simon: does anyone think that the scenario in my email can happen

anish: mikeR: i do 

Dave Booz: 

http://www.oasis-open.org/apps/org/workgroup/sca-j/email/archives/200801/msg00081.html

..... 

To correct this omission, I propose that the following text be added to
the end of the text in 

   http://www.oasis-open.org/apps/org/workgroup/sca-j/download.php/27104/Issue_4_Resolution_01.doc

that was adopted as the resolution of issue 4:

Callback references MUST NOT be reinjected.  If a conversational-scoped
service implementation is invoked by multiple clients using the same application-generated
conversationID, any injected callback reference will refer to the client
that made the initial invocation.  A reference to the client that
made the current invocation can be obtained by using the RequestContext
API. 

   Simon 

Simon C. Nash, IBM Distinguished Engineer

Member of the IBM Academy of Technology

Tel. +44-1962-815156  Fax +44-1962-818999 

"Michael Rowley"
<> 

07/02/2008 19:05

To

"OASIS Java" <>

cc

Subject

RE: [sca-j] ISSUE 4 - Proposed spec
text  for dependency reinjection

 

For the record, this issue was resolved on today’s call.  Anish will
be updating the spec based on this proposal
wording and this
spreadsheet
table.

 

Michael 

  

From: Mike Edwards [mailto:] 

Sent: Friday, February 01, 2008 5:45 AM

To: OASIS Java

Subject: Re: [sca-j] ISSUE 4 - Proposed spec text for dependency reinjection

 

Folks, 

I've taken Simon's text and cast it into a Word document so that edits
and comments can be more 

easily tracked.  I've then made a few edits and comments, which I
hope improve the text and deal

with one significant problem relating to the handling or reinjected arrays/collections:

Yours,  Mike.

Strategist - Emerging Technologies, SCA & SDO.

Co Chair OASIS SCA Assembly TC.

IBM Hursley Park, Mail Point 146, Winchester, SO21 2JN, Great Britain.

Phone & FAX: +44-1962-818014    Mobile: +44-7802-467431  

Email:  

Simon Nash/UK/IBM@IBMGB

31/01/2008 15:24

To

"OASIS Java" <>

cc

  

Subject

[sca-j] ISSUE 4 - Proposed spec text
 for dependency reinjection

  

 

 

Here's my attempt to merge the contents of the v5 spreadsheet with 

Michael's original text.  There's bit more work still to do to capture
the 

decisions we made regarding callback injection and reinjection.

- - - - begin spec words - - - -

References MAY be reinjected after the initial creation of a component
if 

the reference target changes due to a change in wiring that has occurred

since the component was initialized.  In order for reinjection to
occur, 

the following MUST be true:

1) The component MUST NOT be STATELESS or REQUEST scoped.

2) The reference MUST use either field-based injection or setter 

injection. References that are injected through constructor injection MUST

NOT be changed.  Setter injection allows for code in the setter method
to 

perform processing in reaction to a change.

3) If the reference has a conversational interface, then reinjection MUST

NOT occur while the conversation is active.

If a reference target changes and the reference is not reinjected, it MUST

continue to work as if the reference target was not changed.

If an operation is called on a reference where the target of that 

reference has been undeployed or has become unavailable, the SCA runtime

SHOULD throw InvalidServiceException.  It the target of the reference
is 

changed, the reference MAY continue to work, depending on the runtime and

the type of change that was made.  If it doesn't work, the exception

thrown will depend on the runtime and the cause of the failure.

A ServiceReference that has been obtained from a reference by 

ComponentContext.cast() corresponds to the reference that is passed as
a 

parameter to cast().  If the reference is subsequently reinjected,
the 

ServiceReference MUST continue to work as if the reference target was not

changed.  If the target of a ServiceReference has been undeployed
or has 

become unavailable, the SCA runtime SHOULD throw InvalidServiceException.

If the target of a ServiceReference is changed, the reference MAY continue

to work, depending on the runtime and the type of change that was made.
If 

it doesn't work, the exception thrown will depend on the runtime and the

cause of the failure.

A reference or ServiceReference accessed through the component context
by 

calling getService() or getServiceReference() MUST correspond to the 

current configuration of the domain.  If the target has been undeployed
or 

has become unavailable, the result SHOULD be a reference to the undeployed

or unavailable service, and attempts to call business methods SHOULD throw

InvalidServiceException.  If the target has changed, the result SHOULD
be 

a reference to the changed service.

The rules for reference reinjection also apply to references with a 

multiplicity of 0..N or 1..N.  This means that in the cases listed
above 

where reference reinjection is not allowed, the array or Collection for

the reference MUST NOT change their contents.  In cases where the
contents 

of a reference collection MAY change, then for references that use setter

injection, the setter method MUST be called for any change to the 

contents.  The injected collection MAY be the same collection object
as is 

currently used by the component, but with some change to its contents.

- - - - end spec words - - - -

 Simon

Simon C. Nash, IBM Distinguished Engineer

Member of the IBM Academy of Technology

Tel. +44-1962-815156  Fax +44-1962-818999

Unless stated otherwise above:

IBM United Kingdom Limited - Registered in England and Wales with number

741598. 

Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire PO6
3AU

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

To unsubscribe from this mail list, you must leave the OASIS TC that

generates this mail.  You may a link to this group and all your TCs
in OASIS

at:

https://www.oasis-open.org/apps/org/workgroup/portal/my_workgroups.php

  

Unless stated otherwise above:

IBM United Kingdom Limited - Registered in England and Wales with number
741598. 

Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire PO6
3AU 

Unless stated otherwise above:

IBM United Kingdom Limited - Registered in England and Wales with number
741598. 

Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire PO6
3AU 

Unless stated otherwise above:

IBM United Kingdom Limited - Registered in England and Wales with number
741598. 

Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire PO6
3AU 

Unless stated otherwise above:

IBM United Kingdom Limited - Registered in England and Wales with number
741598. 

Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire PO6
3AU