On Sat, Mar 29, 2008 at 6:32 AM, Michael Poulin <m3poulin@...> wrote:
>
> I see your point, Anne. If I properly understood, you confirm the statement
> of my colleague - RIA is good with fine-grained services.

I'm not sure I'd make such a deterministic judgment. It very much
depends on the type of RIA system you're using. Many RIA technologies
work very effectively with coarse-grained, method-oriented services
(e.g., SOAP services). The "webbier" RIA systems and most mashup tools
are designed to exploit web resources. They can still access SOAP
services, but they interact more natively with web resources.

>
> However, I am confused a little (I am not an expert in REST but know
> basics). The confusion comes from this logic: if everything in REST is a
> resource and resource (in REST) has an "interface [the resources it exposes]
> is fine-grained", then what is the service in REST, especially,
> coarse-grained one?

You need to distinguish "resource" from "service". A service supports
a capability, such as account management. Service consumers interact
with the service through the set of resources it exposes. In other
worlds, the resource model *is* the interface to the service. Each
resource exposes a uniform interface (e.g., GET, PUT, POST, and
DELETE), but an individual resource is not the complete service.
>
> My understanding of REST is: REST is a concrete communication model between
> resources; one can read resources as applications.

Not quite. REST is a resource-oriented architectural style for
creating applications. The style is defined by a set of principles and
constraints. The fundamental unit in a RESTful application (or
service) is a resource, but an application typically comprises
multiple resources. All resources use a uniform means of
identification (.e.g, a URI). All resources expose a uniform interface
(e.g., GET, PUT, POST, DELETE). Client applications interact with the
resource by exchanging representations. A RESTful application uses
hypermedia as the engine of application state. (There are more
principles and constraints, but these are the basics.)

> It is a clear alternative
> to Web Services but we look like agreed that Web Services is not SOA, they
> might be one of the interface types of a SOA Service.

Again, it's not quite that black and white. A RESTful service can
support communications (i.e., exchange of representations) using SOAP.
Alternatively, a service can expose its capabilities through a variety
of interface types, such as:
- a method-oriented interface (i.e., a set of methods) via SOAP
- a resource-oriented interface (i.e., a set of resources) via HTTP
- an event-driven interface (i.e., a set of sinks) via JMS

Bear in mind that I'm suggesting a very wide separation of concerns
between a service's interfaces and its implementation.

>
> OASIS SOA RM and coming RA recognize that SOA Service may have a business
> behavior not necessary visible through the service interface. It is not
> prohibited for a resource having such behavior but with fine-grained
> interface it becomes less comprehensive and expected than with
> coarse-grained interface.

I reiterate that the service interface is composed of the resources
that it exposes. An individual resource is (typically) not a complete
service. The granularity of the resources exposed determines how
fine-grained the interface is.

>
> Thus, to fully understand your response, please, comment on what is a
> service (coarse-grained service) in REST.

Please let me know if I have not yet explained it sufficiently.

Anne
>
> - Michael
>
>
> ----- Original Message ----
> From: Anne Thomas Manes <atmanes@...>
> To: service-orientated-architecture@yahoogroups.com
>
> Sent: Friday, March 28, 2008 2:23:09 PM
> Subject: Re: [service-orientated-architecture] Re: Meehan on RIA meets SOA
>
>
>
>
> On Wed, Mar 26, 2008 at 3:44 PM, Michael Poulin <m3poulin@yahoo. com> wrote:
> >
> > One of lead architects around me said that SOA and RIA are almost
> orthogonal
> > because RIA demands fine-grained operations while SOA tends to
> > coarse-grained ones...
> >
> > - Michael
>
> Hence my assertion that RIA and mashups will become more intimately
> connected with SOA if/when
> REST becomes a predominant approach for building services.
>
> REST exposes capabilities through a resource interface. (see my recent
> post, REST is about Resources
> [http://apsblog. burtongroup. com/2008/ 03/rest-is- about-r.html]). The
> resource interface is fine-grained. Any "thing" that you want to
> interact with has a URL. RESTful services are significantly easier to
> interact with than SOAP APIs, particularly from the RIA and mashup
> tooling perspective.
>
> (Note that the RESTful service can still be coarse-grained -- but the
> interface [the resources it exposes] is fine-grained. )
>
> Anne
>
> >
> >
> > ----- Original Message ----
> > From: Rob Eamon <reamon@cableone. net>
> > To: service-orientated- architecture@ yahoogroups. com
> > Sent: Tuesday, March 25, 2008 2:37:30 PM
> > Subject: [service-orientated -architecture] Re: Meehan on RIA meets SOA
> >
> >
> >
> >
> > Why is that? Why would an RIA be "more connected" to services based on
> > the interaction style? Why would accessing services via REST vs. any
> > other mechanism be considered more connected? A service consumer is a
> > service consumer, regardless of the service interface, no?
> >
> > Or are you referring to the relative prevalence of an RIA's use of
> > services compared to other approaches?
> >
> > -Rob
> >
> > --- In service-orientated- architecture@ yahoogroups. com, "Anne Thomas
> > Manes" <atmanes@... > wrote:
> > >
> > > RIA and mashups will become more intimately connected with SOA
> > > if/when REST becomes a predominant approach for building services.
> > >
> > > Anne
> >
> >
> >
> > ____________ _________ _________ __
> > Be a better friend, newshound, and know-it-all with Yahoo! Mobile. Try it
> > now.
>
>
> ________________________________
> You rock. That's why Blockbuster's offering you one month of Blockbuster
> Total Access, No Cost.