Re: [Terminology]

[Konstantin Beznosov <beznosov@baptisthealth.net>]

> hi,
> 
> >>> ===
> >>> Target:
> >>> Let's borrow a term from Bob Blakley's picture and use the term
> >>> "target" as the
> >>> client of the HRAC.  Agreed?  
> >>
> >>OMA uses two main player names "client" and "target object". I suggest to
> >>stick to "target object" and use "target" as a short name for "target
> >>object." The meaning is the same -- client of the HRAC.
> >>
> >Sounds good to me...
> 
> I am not one to quibble over terminology. I'm happy to use whatever
> terminology everyone else wants. However, if the goal is to be
> consistent with OMA, then shouldn't the circle labled "target or target
> wrapper" in Bob's picture be labled "client" and the circle labeled
> "Access Decision" be referred to as the "target"? HRAC is an interface
> that becomes instantiated in an object, the ADO, the client of which is
> a "client" whose "target object" is the ADO.
> 
> jb


John is correct from one point of view. Each object in OMA can play a role of
client or target depending on an interaction it is involved at the particular
moment, i.e. th context.

We want to make sure that our terminology and names of role are consistent with
the context and that we do not have change the terminology when we put the
specification together.


What if we call the object, which calls ADO, an "application service" (AS) ?

In that case, we do not have to get confused who is a client and who is the
target. A phrase "Application service invokes methods on ADO" will be correct
no matter what role (client, target) the application service performs.

Konstantin

----------------
Broadcast message to hrac-rfp from Konstantin Beznosov <beznosov@baptisthealth.net>.
Go to http://cadse.cs.fiu.edu/omg/hrac-rfp to browse the mail list archive.