OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.

 


Help: OASIS Mailing Lists Help | MarkMail Help

soa-rm message

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]


Subject: RE: [soa-rm] [issue:07-23 and 07-24] Tighten definitions of "data resource" and "processing resource"


Sure, that's actually what I mean: as editor for terminology, the proposals
I make will be encapsulating only those issues where there seems to be
consensus or, if not, propose a way forward. In both cases, the issue
proposal should merely provide a focus for a group discussion and
decision...

-Peter 

-----Original Message-----
From: Chiusano Joseph [mailto:chiusano_joseph@bah.com] 
Sent: 26 May 2005 12:54
To: SOA-RM
Subject: RE: [soa-rm] [issue:07-23 and 07-24] Tighten definitions of "data
resource" and "processing resource"

Agreed - but we should also note case in which there is no commonly accepted
terminology (or perhaps no terminology at all). I see part of our effort as
creating terminology where it makes sense, and also adopting existing
terminology where it makes sense.

Joe

Joseph Chiusano
Booz Allen Hamilton
Visit us online@ http://www.boozallen.com
 

> -----Original Message-----
> From: John Harby [mailto:jharby@gmail.com]
> Sent: Thursday, May 26, 2005 8:15 AM
> To: SOA-RM
> Subject: Re: [soa-rm] [issue:07-23 and 07-24] Tighten definitions of 
> "data resource" and "processing resource"
> 
> IMHO, it would be a mistake to introduce additional terminology and 
> classifications while we are still working on those which are commonly 
> accepted.
> 
> On 5/26/05, Peter F Brown <peter@justbrown.net> wrote:
> > Joe:
> >  
> > 1. I agree that the terminology needs tightening up, and I'll be 
> > proposing specific terminology revisions as issues later this week, 
> > taking on board your and others comments. I'm really
> grateful for your
> > extensive comments that highlight many of the inevitable
> inconsistencies.
> > 2. Your clarifications on these issues would be compatible
> with both
> > definitions of a service that I intend to submit as issues
> to the list
> > later this week [1] ...but...
> > 3. I'm a little nervous about the idea of introducing
> "-oriented" as a
> > descriptor, as in "data-oriented service" and "process-oriented 
> > service" If a "data-oriented service" is only providing data read 
> > functionality why does that make it more data-oriented than a 
> > "process-oriented service" that may commit/update/delete
> data? I think
> > the issues of data and process are orthogonal.
> > I'd be happier talking about data objects and process
> objects that are
> > invoked and used by services.
> > 4. On a more fundamental level, I'm not sure that the whole 
> > distinction data/process is useful as implying two "families" of
> > service: both of the scenarios are services, one that reads and 
> > manipulates data from a data resource; the other which might 
> > ultimately commit/update/delete data. But surely these are
> just policy
> > issues regarding how the service can invoke and use the
> resource (the data) in the context of a particular contract?
> > 5. I do not agree, within your text on a data-oriented
> service, that
> > "it should be noted that such processing does not consitute a 
> > process-oriented service" Why not? If the service
> transparency is such
> > that such processing can be seen by the service requestor,
> it may be
> > that at some future stage, that this specific processing be
> invoked as
> > part of another orchestrated service. It is not the
> distinction data
> > or process oriented that is useful, but the distinction based on 
> > resource use that I find useful, and that I think is a policy issue.
> >  
> > -Peter
> >  
> > [1] Service
> > 
> > a) A behavior or set of behaviors [PFB1]  offered by one entity for 
> > use by another according to a policy and in line with a service
> > description.b) The use by one entity of a resource made
> accessible by
> > another entity [PFB2] ________________________________
> > 
> > 
> > 
> >  [PFB1]"Bahviour" not defined
> > 
> > 
> >  [PFB2]Alternative definition
> > ________________________________
> > From: Chiusano Joseph [mailto:chiusano_joseph@bah.com]
> > Sent: 15 May 2005 18:14
> > To: SOA-RM
> > Subject: [soa-rm] [issue:content] draft 07, sect2.1.3.1, line 338, 
> > Tighten definitions of "data resource" and "processing resource"
> > 
> > 
> > 
> > 
> > <current>
> > 
> > I don't believe that the definitions presented beginning
> with line 338
> > accurately capture the distinct differences between data-oriented 
> > services (using terminology suggested in earlier issue) and 
> > process-oriented services. For example, line 338 states that a data 
> > resource (data-oriented
> > service) "accepts a request and returns a value or set of values in 
> > response". A processing resource (process-oriented service)
> does the
> > same - it accepts a request to execute a task/process (a
> process being
> > comprised of multiple tasks), and - more often than not - returns a 
> > value or set of values in response. For example, consider a
> process in
> > which a manufacturer orders several widgets from a supplier - the 
> > supplier will accept a request from the manufacturer, and then 
> > eventually send a response that will either confirm that
> the order has
> > been placed, or that the inventory is not available at that time.
> > 
> > </current>
> > 
> >  
> > 
> > <suggested>
> > 
> > Recommend considering these definitions: 
> > 
> >  
> > 
> > A data-oriented service is a service whose sole function is
> to provide
> > a set of data upon request, given certain criteria that are
> passed as
> > part of the request. The primary processing that a data-oriented 
> > service performs is to locate the set of data requested,
> and to either
> > return that data or convey that the data could not be found or 
> > accessed. Therefore, the extent of processing for a data-oriented 
> > service is minimal compared to other types of services.A
> data-oriented
> > service may perform processing necessary to produce the requested 
> > data, such as calculations. In doing so, it may invoke one or more 
> > process-oriented services, or perform such processing itself (it 
> > should be noted that such processing does not consitute a 
> > process-oriented service). Invocation of a data-oriented service, 
> > unlike other types of services, never results in a change
> in the state
> > of the environment within the reach of the service. That is, 
> > invocation of a data-oriented service will not result in a
> decrease in
> > inventory in a factory, a transfer of money from a bank to
> a business,
> > etc.  An example of a data-oriented service would be a
> stock quote service that accepts a ticker symbol and returns a stock 
> quote.
> > 
> >  
> > 
> > A process-oriented service is a service that performs a task or a 
> > process upon request that may results in a change in the
> state of the
> > environment within the reach of the service. For example,
> invocation
> > of a process-oriented service may result in a decrease in
> inventory in
> > a factory, a transfer of money from a bank to a business, or a 
> > notification of subscribed parties of a certain message. In
> order to
> > carry out its request, a process-oriented service may invoke one or 
> > more other process-oriented services, and/or a
> data-oriented service
> > (for example, to invoke the data-oriented service to
> perform a calculation).
> > 
> > </suggested>
> > 
> >  
> > 
> > <notes>
> > 
> > I know that there is sometimes a fine line between these 2 types of 
> > services
> > - I endeavored to differentiate them as clearly as possible.
> > 
> > </notes>
> > 
> >  
> > Kind Regards,
> > Joseph Chiusano
> > Booz Allen Hamilton
> > Visit us online@ http://www.boozallen.com
> >
> 




[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]