Re: EARL: representing uncertainty

The confidence property that was in earlier versions of EARL is probably a
useful thing - as I recall it allowed for numeric values, and it should be
possible to name some stages along that.

It is also useful to make EARL statements about types of EARL results - and
to compare results from different tools to know whether we can end up relying
on a statement that conflicts with a different one.

One approach is to do these smaller tests against specified test cases, and
then be able to add up the results of conforming to some of those as a basis
for working out conformance to a larger checkpoint.

Chaals

On Wed, 7 Nov 2001, Nick Kew wrote:

  It is in the nature of such evidence that it cannot always be
  dealt with by a simple binary yes/no or pass/fail representation.
  There will often be uncertainty.