[news.eclipse.technology.ohf] Re: Limit the scope?

[Grahame Grieve <grahame@xxxxxxxxxxxxxxx>]

Hi Don

> Does Eclipse formally define the terms "framework" and "platform" anywhere?
> Is this project building a framework or a platform or both?

I don't know. I guess we're going to find out

> The initial scope of this project should be limited to the implementation of
> a clearly specifed core framework in direct support of healthcare
> workstation and embedded applications.

After some thought, I can't agree. The initial scope of the project should 
be driven be what people want to contribute. Limiting the scope of the 
project seems like a bad idea to me. I also wonder quite how we will know
how to scope the core framework without having real functionality provided,
and why it would actually be useful or why anyone would use it

As an example of what I mean:

> This core framework should provide
> the functionality and define the extension points available to application
> developers and those supplying additional framework and platform extensions
> down the road. Extensions that handle specific processing requirements (e.g.
> HL7 messages, HL7 CDA, DICOM or Terminology) should be developed outside
> this project.

how do we know what common functions these widely varying pieces of work
share? How do we know how to reduce this to what's common?

I think that we should regard the core framework as an inevitable and
desirable outcome of our work, but not the actual work itself.

It also bothers me, that there is a huge variation in the way functionality
is spread between client and server - who would we deal with this?

> Essential healthcare infrastructure services of the core framework should
> include security/privacy, transport and semantic interoperability,
> structured document/message processing, workstation-centric process support
> and domain resource management--but only in the "framework" sense. Any
> specific realizations should be basic or exemplary in nature.  Basic tool
> support for the framework should remain in scope.

but what tools could there be in the absense of actualy implementations of 
problem solving code?

Grahame