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

Re: per-Framework glossary sections

From: Lofton Henderson <lofton@rockynet.com>
Date: Fri, 03 Jan 2003 13:07:52 -0700
Message-Id: <>
To: David Marston/Cambridge/IBM <david_marston@us.ibm.com>
Cc: www-qa@w3.org

At 12:53 PM 1/3/03 -0500, David Marston/Cambridge/IBM wrote:
>Can you explain the normativity implications? If I'm on track, the
>QA Glossary is not a normative document, but the various QA Guidelines
>documents will be. Thus, any attempt to assess how well a testable
>item conforms to those guidelines will need normative definitions.

You raise an interesting side question.  The current SpecGL text (OpsGL as 
well) defines what parts of the document are normative, at [1].  The 
"Definitions" chapter does not identify itself as normative.  Therefore, it 
is by default informative.  By comparison, UAAG has this sentence at the 
start of its glossary section [2]:

"This glossary is normative. However, some terms (or parts of explanations 
of terms) may not have an impact on conformance."

So it looks like we have an oversight in SpecGL (OpsGL does not yet have a 
Definitions section) -- the normativity of the terminology.

>If a QA document had a Definitions section, and there was a term whose
>entry said just "see QA Glossary", would that promote the term from
>Informative to Normative for purposes of that document?

It would seem so.  I guess we could also make it explicit by moving "QA 
Glossary" from "Informative References" section to "Normative References" 
section in SpecGL (and OpsGL).


[1] http://www.w3.org/QA/WG/2002/12/qaframe-spec-20021220#b2ab3d475
[2] http://www.w3.org/TR/2002/REC-UAAG10-20021217/glossary.html#terms
Received on Friday, 3 January 2003 15:05:54 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 20:40:30 UTC