W3C home > Mailing lists > Public > www-qa-wg@w3.org > December 2003

some SpecGL TA issues

From: Lofton Henderson <lofton@rockynet.com>
Date: Mon, 22 Dec 2003 08:55:09 -0700
Message-Id: <>
To: www-qa-wg@w3.org

A few comments on the TAs, posed as issues...


1.) Issue -- Ambiguous syntax in TA (hard to parse).  Ignoring the 
editorial oops (repetition of the "EITHER" clause), what is written is:

EITHER c1 OR c2 AND c3.  What are the parsing rules for this?  What is 
intended is:  (EITHER c1 OR c2) AND c3.  Without reference to the ConfReqs, 
how do we know that it's not supposed to be:  EITHER c1 OR (c2 AND c3)?

2.) Issue -- Why consolidate independent CRs into single TAs?  There are 
two fairly independent Conformance Requirements.  If "A specification..." 
were pre-pended to each one, then they would indeed be completely 
independent and standalone.  Why conjoin these into a single test 
assertion?  This comment applies to several other CR/TA sets (although the 
CRs are not always clearly separated in SpecGL, as in this case).


Issue -- cross-dependence of this CP on the results of another CP, "@@each 
identified CoP@@".  Could be replaced by "each CoP for which the 
specification defines conformance requirements".  Should that be done?


1.) Issue (trivial) -- "any" or "all"?

2.) Issue -- Is anything lost by omitting any mention of the 6 bullets in 
the TA?  (I'm unsure.)  If "no" then what does their presence in the 
ConfReqs mean?


Issue (trivial) -- to "each of the other DoV" add "used by the 
specification" (we discussed this before).


Issue -- in the TA (as opposed to the ConfReqs), is the "...not applicable" 
sentence needed, given the presence of the "If..." clause?  I think we 
discussed this, but I'm unsure of our conclusion.

Out of time,
Received on Monday, 22 December 2003 10:58:26 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:14:31 UTC