Re: Revised Guidelines and Checkpoints for TestGL

Le Dimanche, 17 août 2003, à 20:41 America/Montreal, Patrick Curran a 
écrit :
> Checkpoint 1.3. Analyze the structure of the specification, partition 
> it as appropriate, and determine and document the testing approach to 
> be used for each partition. [Priority 1]

This will make difficult for people who want to create tests before the 
specification like OWL WG. When the tests predate the specification, we 
should make it clear that the Test GL is the consolidation of the test 
effort.

WG can still develop tests for their technology before defining the 
specification and it could be even encouraged and when the tests are 
mature, they are writing the specification and at the same time create 
the test suite which is in accordance with the Test GL (Which should be 
easy if they have done like that).

So WGs, like OWL WG will not have the feeling that they don't comply to 
the guidelines, and a frustration or discouragement will not arise.

Their process will be:
	1. Requirement docs
	2. Tests
	3. Feature writing (go back to 2. XP dev.)
	4. Specification writing
	5. Consolidation of the TS in conformance with Test GL

> Checkpoint 3.2. Define test materials metadata [Priority 1]
>
> Conformance requirements: The Working Group must define the required 
> set of metadata that will be associated with test materials. This must 
> include at least the following data:

This checkpoint is a bit bizarre. It asks people to define the required 
set in the first sentence and in the second sentence, it makes a 
minimal set mandatory. I would redefine this checkpoint as:

	Conformance requirements: The Working Group must describe the required 
set of metadata that will be associated with test materials. The 
minimal required set of metadata is :

--
Karl Dubost - http://www.w3.org/People/karl/
W3C Conformance Manager
*** Be Strict To Be Cool ***

Received on Monday, 18 August 2003 17:52:52 UTC