RE: Instructions and team assignments for getting to Last Call

Hi Tim.

 

Some thoughts

 

Gregg

 

 

 

"A good test is:
*       Mappable to the specification (you must know what portion of the
specification it tests) 

 

Well. The tests don't test the specification. They only test the technique.
Since they are attached to the bottom of the technique, yes they are
directly mapped to (and connected to) the item that they test.


*       Atomic (tests a single feature rather than multiple features) 

 

Yes - the test is atomic.  It tests only the technique above it. 


*       Self-documenting (explains what it is testing and what output it
expects) 

 

Yes - what the test is testing is immediately above it.   the output of the
test is specified in the test section (right after the test process)


*       Focused on the technology under test rather than on ancillary
technologies 

 

Yes - it tests the technology under test (the technique).   


*       Correct "

 

Correct?   Do you mean the test is correct?     I presume all of our tests
will be correct. 

 

 

Note that there are no tests of the success criteria.    There are any
number of ways to satisfy most of the success criteria.   We always give at
least one way of meeting the success criterion and a way to test that you
have done that technique.  

 

RE 14 test metadata elements 

 

identifier,   --   Covered by Technique Number?

title,   --   Covered by Technique Name

purpose,  --  not needed.  All tests are to test the technique above them

description,  --  provided by process section.

status,   --   Covered by Status of the technique doc that contains the test

specref,  --  not needed.  All tests refer to the technique above them.

preconditions,   -- covered by process?  (what would a precondition be for
us?) (Lets watch this one)

inputs,   --   Same comment as preconditions

expected results,  -- Covered by 
version,   --  Hmmmmm.   Should we add??  Covered by release date of
collection? 

contributor,  --   covered by copyright/credits at bottom of techniques


rights,  --  covered by copyright statement at bottom of techniques (to be
added later - as soon as we get formal format)

grouping, and   -- no grouping of tests or techniques other than by
collection (HTML,CSS, etc) and that is handled by 

seealso.  --   We are looking at whether this should be done or not.  May
put ties to related advisory techniques+tests.

 

 

 
Gregg

 -- ------------------------------ 
Gregg C Vanderheiden Ph.D. 
Professor - Ind. Engr. & BioMed Engr.
Director - Trace R & D Center 
University of Wisconsin-Madison 

  _____  

From: w3c-wai-gl-request@w3.org [mailto:w3c-wai-gl-request@w3.org] On Behalf
Of Tim Boland
Sent: Tuesday, January 03, 2006 7:56 AM
To: Gregg Vanderheiden
Cc: w3c-wai-gl@w3.org
Subject: RE: Instructions and team assignments for getting to Last Call

 

Thanks for the update for the TEST portion!

>From QA Test FAQ Question #7 ("What makes a good test?") [1], 

"A good test is:
*       Mappable to the specification (you must know what portion of the
specification it tests) 
*       Atomic (tests a single feature rather than multiple features) 
*       Self-documenting (explains what it is testing and what output it
expects) 
*       Focused on the technology under test rather than on ancillary
technologies 
*       Correct "

Would any of these points apply/be relevant to the TEST portion mentioned in
the excerpted message following
(as possible additional guidance for the TEST portion)?

Also, the QA Note Test Metadata [2] mentions 14 test metadata elements as
follows:
identifier, title, purpose, description, status, specref, preconditions,
inputs, expected results,
version, contributor, rights, grouping, and seealso.

Some of these are already included in the TEST portion, but would any of
these additional terms be
 useful for the TEST portion mentioned in excerpted message following? 

Thanks and best wishes
Tim Boland NIST

[1]: http://www.w3.org/QA/WG/2005/01/test-faq#good
[2]: http://www.w3.org/TR/2005/NOTE-test-metadata-20050914/



At 04:41 PM 12/30/2005 -0600, you wrote:



Hi Tim,

I updated the wiki page for techniques to include the information for the
TEST portion.

http://trace.wisc.edu/wcag_wiki/index.php?title=Tips_for_editing_techniques 

Thanks
 
Gregg

 -- ------------------------------ 
Gregg C Vanderheiden Ph.D. 
Professor - Ind. Engr. & BioMed Engr.
Director - Trace R & D Center 
University of Wisconsin-Madison 


-----Original Message-----
From: w3c-wai-gl-request@w3.org [mailto:w3c-wai-gl-request@w3.org] On Behalf
Of Tim Boland
Sent: Wednesday, December 28, 2005 3:24 PM
To: w3c-wai-gl@w3.org
Subject: Re: Instructions and team assignments for getting to Last Call


Will there be any consideration of a template for tests, in addition to the
Techniques
template mentioned in the excerpted message following?   Is a test template 
appropriate or necessary?
I think that possibly including such a test template, or at least some
additional specific instructions for test creators, might serve to enhance
the consistency of test format across the various SCs, as well as simply the
process of test description and maintenance.  Since tests are to evaluate
the implementation of related techniques (which have a template), perhaps
tests should also have a template?

Just a thought..

Thanks and happy new year!
Tim Boland NIST

     At 03:05 PM 12/21/2005 -0600, you wrote:


>4.      Draft new techniques and edit existing ones as required, using
>the Techniques template in the WIKI <http://tinyurl.com/dgcd7>. Refer 
>to Tips for editing techniques <http://tinyurl.com/78v78> for detailed 
>instructions.
>5.      Include both pass and fail tests in the tests section of each
>technique. ("Pass" shows correct implementation of the technique; "Fail"
>shows incorrect implementation.)
>6.