Re: Updated version of the test survey (AI-20040126-1 done)

Ref:  http://lists.w3.org/Archives/Public/www-qa-wg/2004Jan/0074.html 
(Dom's original draft)

At 11:30 AM 1/28/04 -0500, Lynne Rosenthal wrote:
>My comments.  [...]
>
>Suggest adding a question, (maybe before 5)
>- Are the test cases conformance tests?

I'm thinking that this is a little lacking in context as stated.  If we do 
add such a new question, I would suggest:

** Move #5 up to follow #2 and put this new one with them.

** Generalize a bit, something like:  "What are the principal goals and 
purposes of the test materials:  conformance testing? or other (specify)?"

Or the wording could go further and mention some other types of test 
materials (interop, ...).


>Suggestion for Test Reporting question
>- Are test results captured and made available to WG members or the 
>public? (if test reports are available on the Web, please give us a URI)

Dom said, "I think it would be useful, but I haven't found a way to 
formulate it in a way that I thought would be understandable by enough 
people".  As I look at Lynne's wording, it occurs to me ... if people have 
a test-results-reporting scheme, they will understand the question.  If 
they don't understand the question, the answer is probably "no".

Starting with Lynne's suggestion, I wonder if we ought to play off of the 
preceding question about automation?  E.g., "Do you have an automated or 
semi-automated process by which test results are gathered and made 
available (examples: EARL, RDF, ...)?  If test reports are available on the 
Web, please give us a URI."

Or another alternatively:  "Is the process of gathering and reporting test 
results automated (examples: EARL, RDF, ...), or manual?  If test reports 
are ...blah...blah..."

-Lofton.