RE: [wsrf] Alternative GetAllResourceProperties proposal

["David W. Levine" <dwl@watson.ibm.com>]

At 05:19 PM 10/6/2004, Steve Graham wrote:

> I am struggling to
> understand that the use cases are the same.  With
> SetResourceProperties, the idea was that the designer of the request
> would have access to the "right" level of granularity to set
> only those properties that were "settable" and that the
> requestor actually wished to set. It seems to me that with PutRPDoc, we
> are suggesting that there are a non-trivial set of situations wherein the
> entire set of properties is in fact "settable" and that
> requestors will normally want to set all of the properties at the same
> time.  Have you seen this as a situation that is perceived as
> "common" from WSDM?  It strikes me that the more common
> situation will be one in which several (often the majority) of a
> Ws-Resource's properties are read-only, therefore rendering the use of
> PutRPDoc to be marginal at best. 
>
> sgg 

And suggestive of complex and messy fault semantics, in all the cases
where the client is wrong/confused about what is
settable. 

- David

++++++++
Steve Graham
(919)254-0615 (T/L 444)
STSM, On Demand Architecture
Member, IBM Academy of Technology
<Soli Deo Gloria/>
++++++++



"Sedukhin, Igor S" <Igor.Sedukhin@ca.com>

10/06/2004 04:42 PM

To

Steve Graham/Raleigh/IBM@IBMUS

cc

"Springer, Ian P." <ian.springer@hp.com>, <wsrf@lists.oasis-open.org>

Subject

RE: [wsrf] Alternative GetAllResourceProperties proposal




I guess the use case is the same as the use case for update of multiple properties. If I know that I'm updating all of them in a small record such as a cusomer record exposed as a WS endpoint. Why would I necessarily have to assemble the fairly intricate update multiple request instead of a simple put RP doc request?
 

-- Igor Sedukhin .. (igor.sedukhin@ca.com)
-- (631) 342-4325 .. 1 CA Plaza, Islandia, NY 11788

 

---

From: Steve Graham [mailto:sggraham@us.ibm.com]
Sent: Wednesday, October 06, 2004 4:20 PM
To: Sedukhin, Igor S
Cc: Springer, Ian P.; wsrf@lists.oasis-open.org
Subject: RE: [wsrf] Alternative GetAllResourceProperties proposal


Consistency is nice, but valid use cases are even better. We should not be driven by architectural considerations that are *not* backed up by use.

So, are we serious to consider PutRPDoc?  Do we seriously consider that wholesale change/update of RPs on mass with a single put is reasonable?  Would anyone want this? Does WSDM?  What is the semantics of doing a PutRPDoc that contains properties that are normally read-only.  I see a can of worms here that I don't think are worth opening for "architectural purity" reasons alone.

sgg

++++++++
Steve Graham
(919)254-0615 (T/L 444)
STSM, On Demand Architecture
Member, IBM Academy of Technology
<Soli Deo Gloria/>
++++++++


"Sedukhin, Igor S" <Igor.Sedukhin@ca.com>

10/06/2004 03:58 PM

To

"Springer, Ian P." <ian.springer@hp.com>, <wsrf@lists.oasis-open.org>

cc
Subject

RE: [wsrf] Alternative GetAllResourceProperties proposal






+1 to "GetResourcePropertiesDocument"
 
Now, this also, somehow leads me to think that we'd need a "PutResourcePropertiesDocument" to mirror this functionality and be logically consistent.

-- Igor Sedukhin .. (igor.sedukhin@ca.com)
-- (631) 342-4325 .. 1 CA Plaza, Islandia, NY 11788

---

From: Springer, Ian P. [mailto:ian.springer@hp.com]
Sent: Wednesday, October 06, 2004 10:24 AM
To: wsrf@lists.oasis-open.org
Subject: RE: [wsrf] Alternative GetAllResourceProperties proposal

Steve,
 
I just checked the spec and the term used is actually "resource properties document" (properties plural), so for consistency, the operation name should be "GetResourcePropertiesDocument".
 
-Ian

---

From: Steve Graham [mailto:sggraham@us.ibm.com]
Sent: Wednesday, October 06, 2004 9:48 AM
To: Springer, Ian P.
Cc: wsrf@lists.oasis-open.org
Subject: RE: [wsrf] Alternative GetAllResourceProperties proposal


I am fine with the name change to GetResourcePropertyDocument.

++++++++
Steve Graham
(919)254-0615 (T/L 444)
STSM, On Demand Architecture
Member, IBM Academy of Technology
<Soli Deo Gloria/>
++++++++

"Springer, Ian P." <ian.springer@hp.com>

10/06/2004 09:10 AM

To

<wsrf@lists.oasis-open.org>

cc
Subject

RE: [wsrf] Alternative GetAllResourceProperties proposal








Steve,

I like the old way better, because it is consistent with the response formats of GetResourceProperty and GetMultipleResourceProperties - i.e. the children of the response element are the property elements. Also, I'm not sure what value including the property document element adds.

However, if you do decide to include the document element, I think a better name for the operation would be GetResourcePropertyDocument.

Regards,
Ian

---

From: Steve Graham [mailto:sggraham@us.ibm.com]
Sent: Tuesday, October 05, 2004 8:01 AM
To: wsrf@lists.oasis-open.org
Subject: [wsrf] Alternative GetAllResourceProperties proposal


WSRFers:
I posted a modified GetAllResourceProperties proposal [1], with WSDL [2]
This is a slight modification to the original proposal, in that the response to the request wrappers the resource property values with a root element (which is the root element of the resource properties document.

Please comment

sgg

[1]http://www.oasis-open.org/apps/org/workgroup/wsrf/download.php/9574/GetAllResourceProperties.b.doc
[2]http://www.oasis-open.org/apps/org/workgroup/wsrf/download.php/9575/GetAllResourceProperties.wsdl
++++++++
Steve Graham
(919)254-0615 (T/L 444)
STSM, On Demand Architecture
Member, IBM Academy of Technology
<Soli Deo Gloria/>
++++++++