At 04:01 PM 12/28/99 -0500, Gunther Schadow wrote:
>"Alvin B. Marcelo" wrote:
> > VistA, Java, Eiffel, GEHR, USAM, C++, C, Python.....
> >
> > I see CORBAmed/CORBA as the least common denominator in all these.......it
> > does not preclude any of the above and falls within our definition of "open
> > source" (or does it? -- Dave F?)
> >
> > Is it alright for us therefore to focus our sights on CORBAmed since it is
> > the most neutral of all the "models" set forward.
>
>I disagree very much. I do not want to cause any unnecessary violence
>here, but the conclusion CORBAmed is the neutral way to go is not
>founded. CORBAmed's COAS is not independent or any more general than
>GEHR or HL7 v3/USAM. CORBAmed comes with a specific model on observations
>etc., just like HL7 and GEHR. So, deciding on anything between these
>three is preemptive, and, I think pretty useless.
The COAS has an example of a data model from HL7 but should be easily used
also with GEHR.
So it in no way comes down on the side of a GEHR model or an HL7 model. It
is different than HL7 by providing a functional viewpoint rather than an
information viewpoint. It is very important that we understand the importance
of a functional viewpoint in addition to the information view point as
emphasized by the RM-ODP model which I've mentioned earlier. This has
nothing to do with CORBAmed, per se, but helps one understand how these
various models fit together. If we don't adopt a CORBAmed approach, we
will simply have to reproduce it in some other form.
>The fact that you can half-way map between any models is
>always touted by CORBAmed people as the big advantage of CORBAmed
>specs. But it is a pointless half-truth, and I resent this strategy
>of CORBAmed advertizing for creating exactly the notion that Alvin
>recites here. I have been silent to David Forslund's continuing blowing
>the marketing horn for CORBAmed here since I don't see a way to stop
>this by argument. I guess I don't know what this list or the OSHCA
>thing will do anyway, but I'm not gonna waste time to cheer CORBAmed
>and COAS.
We need to adopt a modeling strategy and come up with some cross-platform
specifications which bring out the requirements of an information model and
a computational model independent of the technology. It isn't that
CORBAmed is the greatest thing since sliced bread, but it is the only
distributed object model that is out there. It doesn't prescribe any
language or implementation strategy. It is based on an ISO standard and
fully supports the HL7 requirements. We need to follow some kind of strategy.
>For the 100th time I'd like to point out that CORBA and CORBAmed are
>*not* the same thing, and that I like CORBA technology very much, but
>I disagree on a lot of the healthcare domain work that CORBAmed has been
>done. The CORBAmed model would belong on the same examination table
>that we put GEHR and HL7 v3 on, but CORBAmed folks continuously step
>aside from this examination by claiming that CORBAmed is independent
>from the information model, it is not!
I've also said that CORBA and CORBAmed are not the same. What IDL would
you like to settle on if not some of the IDL specified by CORBAmed? It is
fully consistent with HL7 and is adaptable to GEHR. CORBAmed COAS has an
information model, yes, but it is basically that of XML DOM, to which HL7
is being mapped, too. If we want to move to a different IDL, what is the
process by which we do so, and why?
Best wishes for the new millennium, too.
Dave
>regards and happy millenium!
>-Gunther
>
>--
>Gunther_Schadow-------------------------------http://aurora.rg.iupui.edu
>Regenstrief Institute for Health Care
>1050 Wishard Blvd., Indianapolis IN 46202, Phone: (317) 630 7960
>[EMAIL PROTECTED]#include <usual/disclaimer>
