"John S. Gage" wrote:
> Thomas Beale wrote:
>
> >
> > XML is not a database system - it makes no proposal at all about databases.
>
> My impression is that it is trivial to translate from XML into SQL or an
> object database, so that while technically XML is "not a database
> system", it is closely aligned to just about any database one wants.
Hm. I would not say trivial. At ths very moment, we are engaged in conversion
between XML-schema XML and objects in an ODB. It appears as though it will work
ok, although XML-schema's object-orientation is not real strong.
To SQL? Well, I suppose if you are expressing queries in the XML....
> Let me state where I am coming from, so that my premises can be
> attacked. I believe that open source in medicine, in order to succeed,
> must offer clinicians/domain experts a "lingua franca" similar to what
> the language C offered the Linux and the Apache project. That will not
> be easy, at all. I think that something very like XML will be that
Ah, I see where you are coming from. Agree with your aim..
> "lingua franca" (how many clinicians have dabbled in HTML versus
> dabbling in Java, for example?). And I think that the "proof of the
> pudding" is the HL7/PRA project, which is very close to the domain.
> Specifically, I don't think that any of the object languages, such as
> Java, Eiffel, etc., and the common object facilities such as CORBA, will
These all have a different purpose. Any system will be built from one or more of
these technologies, including HL7 systems. But like HL7, GEHR is using XML to
express its archetypes, which are its clinical semantics. So if this is the level
of lingua franca you mean, I agree, XML is a pretty reasonable way to go.
> attract clinicians/domain experts to this process. These are enormously
> controversial statements, I realize, but they are only spoken with a
> desire to move the EMR into reality.
Not controversial at all. I think clinicians should be interested in the clinical
semantics of EHR systems, and not waste time on the underlying technologies too
much. So I support clinicians concentrating on the clinical information model &
semantics level of systems.
Want to develop some GEHR archetypes?! (Have a look at
http://www.gehr.org/gpcg/Archetypes/Archetypes.htm for examples of clinical
archetypes under development. What you see there is the precursor design by
clinicians before they are expressed in XML-schema.)
- thomas beale
