Re: Conformance requirements extracted from Ops and Spec

At 04:37 PM 2/19/04 +0100, Dominique Hazaël-Massieux wrote:
>[...]
>As discussed during yesterday's teleconf, I built a quick XSLT that
>allows to extract the conformance requirements from our GL docs, relying
>on our markup conventions; the XSLT is at:
>http://www.w3.org/QA/WG/2004/02/extract-cr-from-gl.xsl

Nice stuff!  A future candidate for our toolbox (of things that are useful 
to QAWG/IG's audience)?

Looking at the results, it makes clear that we are using subtly different 
styles in Ops and Spec...


>and to see it applied to the latest version of SpecGL, see:
>http://www.w3.org/2000/06/webdata/xslt?xmlfile=http%3A%2F%2Fwww.w3.org%2FTR%2Fqaframe-spec%2Fguidelines-chapter&xslfile=http%3A%2F%2Fwww.w3.org%2FQA%2FWG%2F2004%2F02%2Fextract-cr-from-gl.xsl
>and to OpsGL, see:
>http://www.w3.org/2000/06/webdata/xslt?xmlfile=http%3A%2F%2Fwww.w3.org%2FTR%2Fqaframe-ops%2Fguidelines-chapter&xslfile=http%3A%2F%2Fwww.w3.org%2FQA%2FWG%2F2004%2F02%2Fextract-cr-from-gl.xsl

In Ops, each conformance requirement is a bullet, and it is self contained 
and standalone.  In Spec, single CRs are sentences or paragraphs, and 
multiple ones are groups of multiple bullets where the subject has been 
factored out and precedes the list.

Thinking back to our discussion of yesterday, about possible test materials 
for our GL documents, we looked at two options.  Condensing from a 2-CR 
example at [1], Option I looked like:

###
      *  CR4.4-1:  the Working Group's QA Process Document MUST specify at 
least one public archived mailing list for QA announcements and submission 
of public QA comments;

      TS4.4-1, determine the appropriate choice:
      ____ The WG does not satisfy this requirement.
      ____ The WG satisfies the requirement, as demonstrated at (URL) 
_______________ .

      *  CR4.4-2:  the Working Group's QA ...blah...blah...etc
###

We decided that this was an *adequate* test case (or two), but that we 
preferred the Option II style better.  Option I has the nice characteristic 
that a small upgrade to Dom's XSLT would allow it to be generated 
automatically from OpsGL.  But we didn't like the unspecialized, generic 
wording of the test-case (e.g., TS4.4-1) section.

I was thinking about that after the telecon.  It would be just as easy to 
automatically generate this variant:

###
      *  @@CR4.4-1@@  [Ed.note -- just a link to the spec.]

      TS4.4-1, determine the appropriate choice:
      ____ The WG satisfies the requirement that, "the Working Group's QA 
Process Document MUST specify at least one public archived mailing list for 
QA announcements and submission of public QA comments", as demonstrated at 
(URL) _______________ .
      ____ The WG does not satisfy this requirement.

      *  @@CR4.4-2@@:  [...ditto...]

      TS4.4-2, determine the appropriate choice:
      ____ The WG satisfies the requirement that, "the Working Group's QA 
...blah...blah...etc"
      ____ The WG does not satisfy this requirement.
###

Does that go at least part way towards the Option II style that we 
preferred more?

-Lofton.

[1] http://lists.w3.org/Archives/Public/www-qa-wg/2003Dec/0057.html

Received on Thursday, 19 February 2004 19:03:22 UTC