Will Ross wrote:
> On Mar 24, 2006, at 9:44 PM, Rod Roark wrote:
>
> > I repeat: NOBODY will pay thousands for certification of Free
> > Software. They will use it because they already believe in it.
>
> Rod,
>
> I have been following the CCHIT process. I do not consider CCHIT to
> be biased against open source. I think competing on a level playing
> field for a fair, tough, industry standard certification is good
> idea. The cost is trivial. If an open source project cannot
> produce a coherent release candidate and collectively finance its
> certification by CCHIT, then that open source project has not scaled
> up to be a credible package for real clinical situations where lives
> may hang in the balance.
>
> I also think it is a disservice to the open source definition to
> propose a dumbed down parallel open source certification process. I
> have no plans to pitch physicians on on software they can "believe"
> in. I want solid open source code that can be equally certified by
> CCHIT alongside NextGen, Centricity, Allscripts, et. al. I see a
> huge marketing advantage for open source to stand up, get certified,
> and start taking business away from the proprietary vendors.
I agree, but certifying authorities such as CCHIT (is CCHIT a govt
certification authority or is it a certifying business set up to make
money or is it an "industry" non-profit set up to perform a service?)
need to to be told to concentrate on making their validation criteria as
automatable and repeatable as possible. Not all criteria can be
automated, but those that can't still need to be specified in a manner
which maximises re-usability - in particular, the form of the vendor's
response. Having a human validator with a clipboard and a tick list,
laboriously working through each test script on each occasion of testing
is not the way it ought to be done. Automated testing tools are the way
to go, and organisations like CCHIT should be prepared to accept scripts
for such testing tools as evidence that criteria are met. Sure they
still need to validate that the test script actually demonstrates that
the criterion in question is met - but they need only do that once.
Repeat testing is then a) check the script is the same as last time b)
click a button. The mantra for testing of all types - unit tests,
regression tests, functional and integration tests - is automate and
then do it often. I can't see why certification criteria can't be guided
by that same principle.
Tim C
Yahoo! Groups Links
<*> To visit your group on the web, go to:
http://groups.yahoo.com/group/openhealth/
<*> To unsubscribe from this group, send an email to:
[EMAIL PROTECTED]
<*> Your use of Yahoo! Groups is subject to:
http://docs.yahoo.com/info/terms/
