Re: [oiic-formation-discuss] Charter and Deliverables: the "State of Interoperability" assessment activity

["Peter Dolding" <oiaohm@gmail.com>]

On Sun, Aug 10, 2008 at 10:07 AM, Dennis E. Hamilton
<dennis.hamilton@acm.org> wrote:
> RE: 2008-07-24 Robert Weir, Draft Interoperability and Conformance TC formation proposal (0.3)
> http://lists.oasis-open.org/archives/oiic-formation-discuss/200807/msg00215.html
>
> CONCERN:
>
> I am startled to see assessments of the state of interoperability to be defined as the earliest work to be carried out.  I don't understand how this showed up with such prominence.  I would think that there are methodological issues and that maybe there needs to be some precursor to establishment of this activity.
>
> Shouldn't we also discuss whether the same group should perform this activity?  It strikes me that separation might be helpful in assuring some sort of objectivity and balance in the process.
>
> ANALYSIS:
>
> I am not concerned with this the same way as Peter Dolding is on this thread:
> http://lists.oasis-open.org/archives/oiic-formation-discuss/200807/msg00195.html
> I can see this assessment being done without requiring test suites, at least in the near term.
>
Near term I can see the same.   The issue is how does reporter answer
challenge of bias or defective reading of standard.   As soon as you
accuse something of having a defect you need the means to back it up.
 Creating a test case that shows it containing methods used exactly
sort out any dispute.  It allows black and white sort out of issue.
Either the testers method is right or wrong.  Same with the
implementation.   Also in a case a company thinks one of the other
companies doing ODF have a defect submitting test cases can be used
too.

Now describing what you did in text two different people can read the
text 2 different ways at times.  So dispute.   Code logic is kinda
universal.  Even if you cannot read English well the logic of what is
contained in a test case does not change.

I am after a Zero room for long lasting dispute model.  No document
created disputes over nothing.

Min is really defect found and a test case created showing it so the
report writers text is miss read or human error its sorted out in a
quick process.

Now if you have a better method to make sure we don't have long term
disputes between report creater and implementers I am all ears.


Peter Dolding