W3C home > Mailing lists > Public > www-qa-wg@w3.org > June 2002

Re: Proposed outline of the evolution of the QA Framework

From: Lofton Henderson <lofton@rockynet.com>
Date: Sun, 16 Jun 2002 09:16:49 -0600
Message-Id: <5.1.0.14.2.20020616084837.0361b6d0@rockynet.com>
To: Dominique Haza螔-Massieux <dom@w3.org>
Cc: www-qa-wg@w3.org

(All QAWG members -- note "QUESTION." below, and please respond.)

Dom,

Thanks for this excellent summary/proposal.   In particular, I think that 
the two-step approach to "mandatoriness" is a very sensible way 
forward.  To recap your proposal:

1.)  Once we declare "CR" phase, it would be mandatory that the WGs 
use/apply the GL documents (Ops, Spec, Test) and report the results to the 
QAWG.

2.) After CR phase, the GL documents are published as W3C Notes and the WGs 
are required to conform (e.g., at least to Level-A).

I would like to present this as WG consensus in the Thursday (6/20) Project 
Review (QA chairs with W3C Team).

QUESTION.  Does anyone object to presenting this as consensus QAWG position?

I do have a question on some details.  You write, "we push a document in WD 
in Last Call when we feel it's stable enough for review by the other 
working groups."  It is (was?) our target to reach last call on all parts 
in 1/2003.  So Last Call period would happen just before Tech Plenary, and 
we would enter CR "implementation" phase after Tech Plenary.

Given that we can't really expect serious attention to documents before 
Last Call (usual W3C practice), how would you suggest to try to get some WG 
input in the interim?  As we said, the evolution and maturation of the 
documents requires external input to us.  I'd like to figure out how to 
stimulate some input before January.  (For example, Ops might be ready for 
some review now;  Spec might be clean enough to benefit from review in the 
August WG-only draft, else in the October publication; etc)

-Lofton.

P.S.  This relates to two issues marked "Closed":

http://www.w3.org/QA/WG/qawg-issues-html#16
http://www.w3.org/QA/WG/qawg-issues-html#18

If we have a new consensus, we'll document the new discussions and modify 
the resolutions of the two issues.

At 08:58 AM 6/14/2002 -0400, Dominique Haza螔-Massieux wrote:

>Summary of the issue:
>It wouldn't make much sense to have the QA Framework be promoted to a
>W3C Recommendation, since it only or mostly concerns W3C internal
>functioning. Provided that it will finish as a set of Notes, how do we
>want these notes to be used and especially enforced.
>
>Summary of yesterday discussions:
>After some discussion, we agreed that in their final state, the various
>specifications should define required level of commitments to the
>objects they apply, namely:
>- QA-OPS: should probably be an addendum to the Process Document (maybe
>like the pubrules)
>- QA-SPEC: probably referenced by the pubrules (like the WCAG)
>- QA-Test: ? Hard to say, no document yet.
>
>The level of commitment for each of these spec has not yet be defined
>(probably level A). The process to have them be required is yet to be
>explored, probably by the staff contacts.
>
>We agreed also that we could not enforce requirement on non measurable
>guidelines, which means that we should aim to have at least all our
>priority 1 check points be easily quantifiable.
>
>Outline of the various status of the QA Framework status:
>While we don't aim to produce Recommendations, It is expected that our
>document will follow the Recommendation track progress as follow:
>- we push a document in WD in Last Call when we feel it's stable enough
>for review by the other working groups. Note that the operational
>guidelines might require a review by the Advisory Board (owner of the
>process document) and/or by the W3C Management Team.
>- after integrating last call comments, we would use the stage of
>Candidate Recommendation as an implementation phase: the idea would be
>to require that each WG applies the GL and sends us the resulting
>report; note that the requirement would *not* to comply with the GL,
>only to use them. This would have the double benefit to allow them to
>have a QA approach to their work and us to have some real life feedback
>on ours. It has yet to be defined how such a requirement could be
>obtained and applied.
>- depending on how the candidate recommendation is successful, we would
>go back to Last Call WD or finalize the document as a Note. Once a Note,
>the document would define a minimum level of compliance as explained
>above.
>
>This would be a 2 stages process: first stage, the WGs would be required
>to use the GL; second stage, the WGs would be required to comply with
>them. This is not a very orthodox evolution in the W3C Process, but
>then, our documents aren't very classic either.
>
>Whatever the consensus we reach on this issue, the resulting idea should
>probably be explained in the introduction document and detailed for each
>specific part of the framework.
>
>Comments and corrections are welcome.
>
>Dom
>--
>Dominique Haza螔-Massieux - http://www.w3.org/People/Dom/
>W3C/INRIA
>mailto:dom@w3.org
Received on Sunday, 16 June 2002 16:41:27 GMT

This archive was generated by hypermail 2.2.0 + w3c-0.30 : Thursday, 9 June 2005 12:13:10 GMT