Re: [COAS-List] tentative model of CEN PT27

[Angelo Rossi Mori <rossi@color.irmkant.rm.cnr.it>]

At 10.03 02/02/99 -0700, Tom Culpepper wrote:

>>>My concern about COAS is that its model is very general,
>>>whereas the models in the HL7 RIM (or the equivalent in CEN)
>>>are specific.
>
>The COAS model is a logical view. 
>The logical view of a system describes the existence 
>and meaning of the key abstractions that form the design.
>
>What we are providing with COAS is an access service 
>to medical information. These services allow you to ask questions 
>of some Server that will then return to you 
>the pertinent information you requested. 
>In an object-oriented world what I believe CEN and HL7 
>are providing are the attributes and not the methods or behaviors. 
>COAS is providing the behaviors.

what I cannot understand (I apologize for my ignorance 
about the actual mechanisms in the object-oriented world)
is how it will be possible to distinguish the behavior
of a whole section from the behavior of a little semantic network
expressing a phrase, if these constructs are not yet
standardized.

In other words, I must know very well each server,
in order to ask them the questions in the suitable way.
If I change the server, I believe that I have no means to 
reuse the previous ones 
or to make automatic transformations to my queries,
because the detailed models within the server are not specified.