- From: Arnold, Curt <Curt.Arnold@hyprotech.com>
- Date: Mon, 22 Oct 2001 09:44:29 -0600
- To: "'www-qa@w3.org'" <www-qa@w3.org>
Mark Skall wrote: >Ok, let me interpret my own point. It' really pretty simple. All I am >saying is that you write tests to test exactly what is required by the >spec (that is the bible) - no more and no less. If the spec is clear, >no interpretation is needed. If the spec is ambiguous, someone >needs to interpret and change the spec to make it unambiguous. But you may not find out that the spec is ambiguous until after you write a particular test and then find out that a particular implementation did not share your interpretation. >No test should be used until the spec (or some other normative >reference like an errata) is changed to remove the ambiguity. I would agree that no test should be ratified as part of a conformance suite until the ambiguity is removed, but running the test would be very useful to determine what implementations agreed (or were at least compatible) with a particular interpretation and to provide guidance to the WG for determining the appropriate resolution.
Received on Monday, 22 October 2001 12:04:00 UTC