Re: [sca-assembly] NEW ISSUE: Need definition of compatible for propertytypes

[David Booz <booz@us.ibm.com>]

I do not want to lose the top down use cases.

Strict type equivalence is the easiest and safest solution.

The question I'm struggling with conceptually is the degree of coupling between a component def'n and a component implementation. We've gone to great lengths to enable a clean component vs. implementation separation, presumably with the goal of enabling top down development which in turn enables replace-ability, substitute-ability, etc. As an analogy, we reduced coupling between components by enabling subset and superset interface relationships, and a similar argument could be made for the component v. implementation relationship. It will serve us well to remember that this component model is intended primarily for implementing coarse grained services which inherently benefit from loose coupling of all kinds. I'll note that strict type equivalence increases the degree of coupling. Just throwing out some thoughts for discussion.....

Dave Booz
STSM, BPM and SCA Architecture
Co-Chair OASIS SCA-Policy TC and SCA-J TC
"Distributed objects first, then world hunger"
Poughkeepsie, NY (845)-435-6093 or 8-295-6093
e-mail:booz@us.ibm.com

Mike Edwards ---04/07/2009 04:23:15 AM---Folks, Comment inline...

From:	Mike Edwards <mike_edwards@uk.ibm.com>
To:	sca-assembly@lists.oasis-open.org
Date:	04/07/2009 04:23 AM
Subject:	Re: [sca-assembly] NEW ISSUE: Need definition of compatible for property types

---

Folks, 

Comment inline... 

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:  mike_edwards@uk.ibm.com

From:	Anish Karmarkar <Anish.Karmarkar@oracle.com>
To:	sca-assembly@lists.oasis-open.org
Date:	07/04/2009 06:28
Subject:	Re: [sca-assembly] NEW ISSUE: Need definition of compatible for property types

---

+1 to raising this issue.

The correct line number for [1] is 1045.

Do we really need this feature? Why allow @type or @element on component 
properties? 

<mje> 
I am somewhat surprised by Anish making this comment! 
One of the usecases of specifying the type of a property on a component is to ensure 
that whichever implementation is used for that component conforms to the needs of 
the component, when building using a top-down approach.  If the type cannot be 
specified, then the composite is forced to accept whatever type the implementation 
decides to provide and no error would get raised in the case of a mismatch. 
</mje> 


The implementation declares the type of the property, it is 
tricky to allow subtypes and hope that it would get mapped correctly and 
would be allowed by the implementation/implementation language 
(especially when we want to allow multiple C&I types). Do we lose 
anything by removing this? 

<mje> 
+1 to disallowing subtypes... 
</mje>

-Anish
--

David Booz wrote:
> TARGET: Assembly spec CD03 [1]
> 
> DESCRIPTION:
> Line 1036 of CD03 [1] says that if a component specifies a property 
> type, then that type must be compatible with the type of the same 
> property in the componentType. What does compatible mean? It might be as 
> simple as the same type of sub-type (i.e though some form of 
> inheritance), but it needs to be specified.
> 
> PROPOSAL:
> None
> 
> [1] 
> http://www.oasis-open.org/committees/download.php/31740/sca-assembly-1.1-spec-cd03.pdf
> 
> 
> 
> Dave Booz
> STSM, BPM and SCA Architecture
> Co-Chair OASIS SCA-Policy TC and SCA-J TC
> "Distributed objects first, then world hunger"
> Poughkeepsie, NY (845)-435-6093 or 8-295-6093
> e-mail:booz@us.ibm.com
> 

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

---

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